Attachment A - NSF Act of 1950 and ACA Reauthorization of 2010

SDR OMB Package for the 2017 Cycle

Attachment A
NSF Act of 1950
and
America COMPETES Reauthorization
Act of 2010

Survey of Doctorate Recipients

Page A-1

SDR OMB Package for the 2017 Cycle

NSF Act of 1950

Only includes the relevant sections of the legislation address the circumstances making the
collection of information on the Survey of Doctorate Recipients necessary.

Survey of Doctorate Recipients

Page A-2

SDR OMB Package for the 2017 Cycle

SECTION I
NATIONAL SCIENCE FOUNDATION ACT OF 1950

FUNCTIONS (42 U.S.C. §1862)
§ 1862. Functions
(a) Initiation and support of studies and programs; scholarships; current register of
scientific and engineering personnel
The Foundation is authorized and directed—

(1) to initiate and support basic scientific research and programs to strengthen scientific
research potential and science education programs at all levels in the mathematical,
physical, medical, biological, social, and other sciences, and to initiate and support
research fundamental to the engineering process and programs to strengthen engineering
research potential and engineering education programs at all levels in the various fields of
engineering, by making contracts or other arrangements (including grants, loans, and
other forms of assistance) to support such scientific, engineering, and educational
activities and to appraise the impact of research upon industrial development and upon
the general welfare;
(2) to award, as provided in section 1869 of this title, scholarships and graduate fellowships
for study and research in the sciences or in engineering;
(3) to foster the interchange of scientific and engineering information among scientists and
engineers in the United States and foreign countries;
(4) to foster and support the development and use of computer and other scientific and
engineering methods and technologies, primarily for research and education in the
sciences and engineering;
(5) to evaluate the status and needs of the various sciences and fields of engineering as
evidenced by programs, projects, and studies undertaken by agencies of the Federal
Government, by individuals, and by public and private research groups, employing by
grant or contract such consulting services as it may deem necessary for the purpose of
such evaluations; and to take into consideration the results of such evaluations in
correlating the research and educational programs undertaken or supported by the
Foundation with programs, projects, and studies undertaken by agencies of the Federal
Government, by individuals, and by public and private research groups;
(6) to provide a central clearinghouse for the collection, interpretation, and analysis of data
on scientific and engineering resources and to provide a source of information for policy
formulation by other agencies of the Federal Government;
(7) to initiate and maintain a program for the determination of the total amount of money for
scientific and engineering research, including money allocated for the construction of the
facilities wherein such research is conducted, received by each educational institution and
appropriate nonprofit organization in the United States, by grant, contract, or other
Survey of Doctorate Recipients

Page A-3

SDR OMB Package for the 2017 Cycle

arrangement from agencies of the Federal Government, and to report annually thereon to
the President and the Congress; and
(8) to take a leading role in fostering and supporting research and education activities to
improve the security of networked information systems.

BIENNIAL REPORT (42 U.S.C. §1885d)
§ 1885d. Biennial reports
(a) By January 30, 1982, and biennially thereafter, the Director shall simultaneously
transmit a report to the Congress, the Attorney General, the Director of the Office of
Science and Technology Policy, the Chairman of the Equal Employment Opportunity
Commission, the Director of the Office of Personnel Management, the Secretary of
Labor, the Secretary of Education, and the Secretary of Health and Human Services.
(b) The report required by subsection (a) of this section shall contain—
(1) an accounting and comparison, by sex, race, and ethnic group and by discipline, of the
participation of women and men in scientific and engineering positions, including—
(A) the number of individuals in permanent and temporary and in full-time and part-time
scientific and engineering positions by appropriate level or similar category;
(B) the average salary of individuals in such scientific and engineering positions;
(C) the number and type of promotional opportunities realized by individuals in such
scientific and engineering positions;
(D) the number of individuals serving as principal investigators in federally conducted or
federally supported research and development; and
(E) the unemployment rate of individuals seeking scientific and engineering positions;
(2) an assessment, including quantitative and other data, of the proportion of women and
minorities studying scientific and engineering fields, including mathematics and
computer skills, at all educational levels; and
(3) such other data, analyses, and evaluations as the Director, acting on the advice of the
Committee on Equal Opportunities in Science and Engineering, determines appropriate to
carry out the Foundation’s functions as well as the policies and programs of sections
1885 to 1885d of this title.

Survey of Doctorate Recipients

Page A-4

SDR OMB Package for the 2017 Cycle

America COMPETES Reauthorization
Act of 2010

Only includes the relevant sections of the legislation that address the circumstances making the
collection of information on the Survey of Doctorate Recipients necessary.

Survey of Doctorate Recipients

Page A-5

SDR OMB Package for the 2017 Cycle

America COMPETES Reauthorization Act of 2010
H. R. 5116—26
SEC. 505. NATIONAL CENTER FOR SCIENCE AND ENGINEERING STATISTICS.
(a) ESTABLISHMENT. – There is established within the Foundation a National Center for Science
and Engineering Statistics that shall serve as a central Federal clearinghouse for the collection,
interpretation, analysis, and dissemination of objective data on science, engineering, technology,
and research and development.
1.
(b) DUTIES. – In carrying out subsection (a) of this section, the Director, acting through the Center
shall –
2.
(1) collect, acquire, analyze, report, and disseminate statistical data related to the science and
engineering enterprise in the United States and other nations that is relevant and useful to
practitioners, researchers, policymakers, and the public, including statistical data on –
(A) research and development trends;
3.
(B) the science and engineering workforce;
(C) United States competitiveness in science, engineering, technology, and research and
development; and
(D) the condition and progress of United States STEM education;
(2) support research using the data it collects, and on methodologies in areas related to the work
of the Center; and
4.
(3) support the education and training of researchers in the use of large-scale, nationally
representative data sets.
(c) STATISTICAL REPORTS. – The Director or the National Science Board, acting through the
Center, shall issue regular, and as necessary, special statistical reports on topics related to the
national and international science and engineering enterprise such as the biennial report required
by section 4(j)(1) of the National Science Foundation Act of 1950 (42 U.S.C. 1863(j)(1)) on
indicators of the state of science and engineering in the United States.

Survey of Doctorate Recipients

Page A-6

SDR OMB Package for the 2017 Cycle

Attachment B
2015 Survey of Doctorate Recipients: Sample
Design and Implementation Report

Survey of Doctorate Recipients

Page B-1

SDR OMB Package for the 2017 Cycle

2015 SURVEY OF DOCTORATE
RECIPIENTS:

Sample Design and Implementation

PREPARED FOR:
Steve Proudfoot, SDR COTR
National Science Foundation
4201 Wilson Boulevard
Arlington, VA 22230
(703) 292-5111

PREPARED BY:
Michael Yang
Karen Grigorian
NORC at the
University of Chicago
55 East Monroe Street
Chicago, IL 60603
(312) 759-4000

DECEMBER 4, 2015

Survey of Doctorate Recpients

Page B-2

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

Table of Contents
1

Overview of the 2015 SDR Sample Design ................................................. 1

2

Sample Design Changes from the 2013 Cycle............................................ 4

3

Sample Frame Construction ........................................................................ 6

4

3.1

Identifying Frame Cases ................................................................................. 6

3.2

Construction of frame variables ...................................................................... 9

Sample Design ............................................................................................ 23
4.1 Precision Requirements ..................................................................................... 23
4.2 Design Approaches Considered ........................................................................ 24
Dual Frame Design ........................................................................................................... 24
Single Frame with a Fresh New Sample........................................................................... 26

4.3 Sample Allocation .............................................................................................. 26
Step 1 ...................................................................................................................... 27
Step 2 ...................................................................................................................... 28

5.

Sample Selection ........................................................................................ 30

6.

SDR 2017 and Beyond ................................................................................ 33

7.

Concluding Remarks .................................................................................. 34

Appendices
APPENDIX A: Requirements of Sample Expansion and Sample Redesign of the
Survey of Doctorate Recipients and Addendum ..................................................... A-1
Section A1. Requirements of Sample Expansion and Sample Redesign of the Survey of
Doctorate Recipients ........................................................................................................A-1
Section A2. Addendum to Requirements of Sample Expansion and Sample Redesign
of Survey of Doctorate Recipients ...................................................................................A-3

APPENDIX B: Descriptions of 2015 SDR Sample Design Options ...................... B-1
Overview ..........................................................................................................................B-1
Sampling Frame ...............................................................................................................B-1
Two Design Approaches ..................................................................................................B-3
Dual Frame Estimation ....................................................................................................B-4
Fresh New Sample...........................................................................................................B-5
Design Options Considered Under Dual Frame Estimation Approach ............................B-6
Design 1.1: Stratification by FFOD Only, No Oversampling ............................................B-6

Survey of Doctorate Recpients

FINAL REPORT | I

Page B-3

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

Design 1.2: Stratification by FFOD Only, Oversampling Underrepresented Minorities
(URM) and Women (Not Presented in Tables) ................................................................B-7
Design 1.3: Stratification by FFOD and Cohort, No Oversampling (Not Presented in
Tables) .............................................................................................................................B-8
Design 1.3B: Stratification by FFOD and Cohort, No Oversampling, Adjustment to New
Cohort Allocation ..............................................................................................................B-9
Design 1.4: Stratification by FFOD and Cohort, Oversampling URM and Women,
Adjustment to New Cohort Allocation ............................................................................B-10
Design Options Considered Under Fresh New Sample Approach ................................B-11
Design 2.1: Stratification by FFOD Only, No Oversampling ..........................................B-11
Design 2.1b: Stratification by FFOD Only, Two-Step Sample Allocation ......................B-12
Design 2.2: Stratification by FFOD Only, Two-Step Sample Allocation, Oversampling
URM and Women...........................................................................................................B-13
Design 2.3: Stratification by FFOD Only, Two-Step Sample Allocation, Oversampling
URM, Women, and Panel Cases ...................................................................................B-13
Designs 2.4a – 2.4d: Stratification by FFOD Only, Two-Step Sample Allocation,
Oversampling URM, Women, and Panel Cases ............................................................B-14
Allocation ........................................................................................................................B-14
Sampling ........................................................................................................................B-16
Design 2.5a: Stratification by FFOD Only, Two-Step Sample Allocation with Second Step
Favoring Large FFODs, Oversampling URM, Women, and Panel Cases .....................B-17
Design 2.6a and 2.6c: Stratification by FFOD Only, Two-Step Sample Allocation with
Second Step Favoring Large FFODs, Oversampling URM, Women, and Panel Cases ...B18
Concluding Remarks ......................................................................................................B-20
File Locations .................................................................................................................B-21

APPENDIX C: 2013 and 2015 SDR Comparison Tables ..................................... C-1
APPENDIX D: 2015 SDR Sample Frame Layout ................................................ D-1
APPENDIX E:

2015 SDR Field of Study Coding Taxonomies Crosswalk ............ E-1

APPENDIX F: Geographic Code Frame Information ........................................... F-1
APPENDIX G: 2015 SDR Frame and Sample Distribution Tables ....................... G-1

Survey of Doctorate Recpients

FINAL REPORT | II

Page B-4

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

Overview of the 2015 SDR Sample Design

1

Since its inception in 1950, the National Science Foundation (NSF) has been charged to provide a central
clearinghouse for the collection, interpretation and analysis of data on scientific and technical resources in
the United States, and provide a source of information for policy formulation by other federal agencies.
The Survey of Doctorate Recipients (SDR) has been an important means for the NSF to accomplish this
objective. Conducted biennially since 1973, the SDR follows a sample of U.S.-trained doctorates in
science, engineering, and health (SEH) fields throughout their careers, from shortly after degree award
through age 75. The SDR is widely used by the U.S. Congress and Federal agencies, universities and
professional societies, and other organizations and individuals interested in the nation’s education, supply,
and employment of doctorate recipients in SEH fields. Employers in universities, industry, and
government sectors also use the SDR to understand and predict trends in employment opportunities and
salaries for SEH doctorates.
The traditional target population of the SDR includes individuals who meet the following requirements:


Received a doctoral degree in an SEH field from a U.S. institution;



Age 75 years or younger on survey reference date; and



Living in a noninstitutionalized setting on the survey reference date.

The SDR has historically featured a stratified systematic sample design, where the strata are defined by
degree field, gender, race and ethnicity, citizenship, disability status, and other relevant demographic
variables. The SDR sample design has undergone some significant modifications over the years in
response to changes in its analytical objectives and budgetary constraints. For example, the number of
strata has been reduced from over 1,000 in the early cycles to 150 as a result of the 2003 redesign. The
target population of the SDR has also been redefined several times over the life course of the survey. For
example, doctorates awarded in humanities were once part of the target population. Furthermore, prior to
the 2003 survey cycle, the SDR restricted data collection to U.S. residents only. SEH doctorates who
resided outside the U.S. on the survey reference date were excluded from the target population of the
survey.

Survey of Doctorate Recpients

FINAL REPORT | 1

Page B-5

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

In addition to the sample redesign, the 2003 SDR included a methodological experiment which showed
that data collection from international residents is operationally feasible.1 From the 2006 cycle, the SDR
sample consisted of two relatively independent components: the national SDR (NSDR) and the
international SDR (ISDR).2 While the NSDR covers doctorates residing in the U.S., the ISDR targets
those residing outside of the U.S. For the 2010 SDR, the NSF decided to integrate the NSDR and ISDR
to create a unified survey of U.S. trained SEH doctorates regardless of residential location.3 The
integrated sample design developed for the 2010 SDR was maintained for the 2013 SDR.4
The 2013 SDR features a total of 194 strata, including 150 NSDR strata and 44 ISDR strata. The NSDR
strata are defined by degree field, gender, race and ethnicity, citizenship at birth, and disability status; the
ISDR strata are defined by degree field, gender, race and ethnicity, and citizenship at birth. These strata
were defined to align with the analytical domains used in official publications as well as those used by
SDR data users.
The 2015 SDR features a substantial sample size expansion and sample redesign in response to a set of
updated analytical objectives and requirements. The sample size is increased from 40,078 cases in 2013
to 120,000 cases in 2015. The main objective of this expansion is to support reliable estimates of
employment outcomes by the fine field of degree (FFOD) taxonomy used in the Survey of Earned
Doctorates (SED). With the marked increase in the overall sample size, the traditional SDR estimation
capabilities are also expected to increase. As directed by the NSF, the overarching 2015 SDR sample
design objectives are twofold:


First, the expanded SDR is required to produce reliable estimates of employment outcomes by the
fine field of degree taxonomy used in the SED;



Second, the expanded sample is expected to maintain the existing estimation capabilities associated
with analytical domains defined by various demographic characteristics and currently used in
National Center for Science and Engineering Statistics (NCSES) publications such as Science and

1

Grigorian, Karen and Tom Hoffer (2005). Non-U.S. Citizen Undercoverage Feasibility Study Report. Report
submitted to the National Science Foundation by the National Opinion Research Center at the University of
Chicago, Chicago, IL.
2
Cox, Brenda G., Karen Grigorian and Michael Yang (2006). The 2006 International Survey of Doctorate
Recipients (ISDR): Sample Design. Report submitted to the National Science Foundation by Battelle under
subcontract to the National Opinion Research Center at the University of Chicago, IL.
3
Cox, Brenda. G., Karen Grigorian, Fang Wang, and Rebecca Wang (2012b). 2010 Survey of Doctorate
Recipients: Sample Design and Implementation. Report submitted to the National Science Foundation by the
National Opinion Research Center at the University of Chicago, Chicago, IL.
4
Cox, B. G., K. Grigorian, Y.M. Yang, M. Sinclair, 2013. 2013 Survey of Doctorate Recipients: Sample Design and
Implementation. Prepared for the National Science Foundation, January 2013. Chicago, IL: NORC.

Survey of Doctorate Recpients

FINAL REPORT | 2

Page B-6

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

Engineering Indicators , Women, Minorities and People with Disabilities report, and detailed data
tables.
For more detailed discussions of the 2015 SDR sample design objectives, please refer to the attached

document “Requirements of Sample Expansion and Sample Redesign of the Survey of Doctorate
Recipients” and its addendum, both can be found in Appendix A.

The expansion of the SDR, along with its new estimation objectives, required a significant redesign of the
SDR sample. As specified by the NSF, the broad objectives of the SDR redesign include meeting the
newly defined estimation objectives, resolving any longstanding sampling issues to improve efficiency,
creating a more unified sample design which eliminates the NSDR and ISDR distinction, and constructing
a flexible and sustainable design for the growing demands of SDR data.
The NSF and NORC conducted extensive research in order to meet these new requirements. Many design
options were considered, simulated, and reviewed during the course of the research. Appendix B contains
comprehensive discussions of these options. Upon evaluating all these options, the NSF decided to select
a fresh new sample from a new sampling frame constructed from the original Doctorate Record File
(DRF). By so doing, the existing SDR panel will cease to exist after the 2013 cycle and a new panel will
start from the 2015 cycle. The main advantage of a fresh new sample is its unparalleled simplicity as it
eliminates all the sample frame, sample design, and database maintenance complexity accumulated over
the past 40 years under the old design. The drawback of a fresh new sample design, however, is the
interruption of a prominent longitudinal data series.
This report documents the 2015 SDR sample design and selection procedures. Section 2 summarizes the
major sample design changes from the previous cycle. The remaining sections discuss in detail the main
parameters of the 2015 SDR design. Section 3 describes the frame construction process which is
considerably different from the prior cycles because a completely new frame is required under the fresh
new sample approach. Section 4 presents the 2015 SDR sample design, including sample size, sample
stratification, and sample allocation procedures under the new sample design to meet the new analytical
objectives. Section 5 describes the sample selection procedures, including methodical oversampling of
the traditional SDR analysis domains under the new design. Section 6 briefly discusses how the SDR
design will be maintained in 2017 and beyond. Finally, Section 7 provides some concluding remarks
regarding data processing procedures under the 2015 design.

Survey of Doctorate Recpients

FINAL REPORT | 3

Page B-7

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

Sample Design Changes from the 2013 Cycle

2

The changes between the 2013 and 2015 SDR sample designs are substantial. Design changes in a
longitudinal study such as the SDR must be documented so that data users can properly analyze the data
and interpret their findings, especially when they employ SDR data from multiple cycles with different
sample designs. Before presenting the 2015 SDR sample design and implementation in detail, this
section highlights the most significant sample design changes from the 2013 SDR and their implications
for analysts, as follows.


Through the 2013 cycle, the SDR had been a longitudinal survey with a significant panel component.
Well over 90 percent of the previous cycle sample is retained in the current cycle sample. With a
fresh new sample selected from a newly constructed sample frame, the 2015 SDR no longer retains
the previous panel, i.e., no explicit longitudinal panel is automatically carried forward from the 2013
sample. Through oversampling, about one third of the 2013 SDR panel cases is included in the 2015
SDR sample. In general, however, the 2015 sample does not support longitudinal analyses. It
practically represents the starting point of a new panel sample.



The 2015 SDR sample frame is constructed afresh from the most recent version of the DRF. In the
past, the SDR sample frame at each cycle consists of two components: the panel (old cohort frame)
from the previous cycle and the new cohort doctorates awarded after the previous cycle (new cohort
frame). Note that the old cohort frame is a secondary frame because it is a sample itself.
Conceptually, the 2015 SDR frame contains three components: (1) the 2013 SDR sample that remains
eligible for the 2015 SDR (n=45,936); (2) the new cohort cases from the 2012 and 2013 SEDs, and
(3) the 2015 “expansion cohort” (n = 979,526) constructed from the 2013 DRF. The expansion cohort
is made up of the following:
►

Those that were selected into the SDR sample but later dropped from the panel due to ineligibility
discovered during subsequent SDR data collection, including the deceased, no degree earned, and
maintenance cut, i.e., deselection from the sample during the 1995-2013 cycles;



►

Those that were eligible for sample selection but were never selected during the past cycles;

►

Those that had been ineligible for selection based on previous target population definitions.

The 2015 sample size is increased to 120,000 cases from a sample size of 47,078 for the 2013 SDR
cycle.

Survey of Doctorate Recpients

FINAL REPORT | 4

Page B-8

SDR OMB Package for the 2017 Cycle

NORC


| SDR 2015 Sample Design and Implementation Report

Instead of defining the sampling strata by degree field and demographics, as had been the case in the
past, the 2015 SDR strata are defined by fine field of degree alone, reflecting the emphasis on the new
analytical objectives at the fine field level.



The 2013 SDR sample allocation is mostly proportional, with additional allocation to small domains
to guarantee a minimum sample size for these domains. The 2015 SDR sample allocation involves a
two-step process to achieve a compromise between the two sets of analytical goals: the first step
allocation to the fine fields is intended to meet the analytical goals at the fine field level; the second
step allocation by the broad field of degree is designed to maintain and improve the existing
analytical capabilities by the traditional analysis domains. The result is a much more disproportional
allocation across the explicit sampling strata. Based on the variation of the base weight, the overall
2015 SDR design effect is 1.59, while the 2013 overall design effect is 1.09.



Under the 2015 design, the traditional analytical capabilities are maintained through oversampling
women and underrepresented minorities (URM). The 2013 panel cases were also oversampled to
support limited longitudinal analysis. For the key traditional SDR domains, a series of tables in
Appendix C compares the coefficient of variation for a typical sample estimate between the 2013 and
2015 SDRs. With rare exceptions, the 2015 SDR is projected to achieve better precision than the
2013 SDR.

Survey of Doctorate Recpients

FINAL REPORT | 5

Page B-9

SDR OMB Package for the 2017 Cycle

NORC

3

| SDR 2015 Sample Design and Implementation Report

Sample Frame Construction

The 2015 SDR employed a completely new original sample frame constructed from the DRF. This
section discusses the frame construction procedures in detail. The goals of frame construction are
twofold: one is to include all doctorates in the target population so they all have a non-zero probability of
being selected into the sample; the other is to define auxiliary frame variables to support sample design
and survey operations. Subsection 3.1 discusses the identification of frame cases; subsection 3.2
discusses the construction of key frame variables. The layout of the frame file is presented in Appendix
D.
3.1 Identifying Frame Cases
Prior to the current expansion, the SDR sample of each cycle consists of two components: an old cohort
sample and a new cohort sample. While the new cohort sample is selected from the new cohort portion of
the frame, the old cohort sample is selected from the old cohort frame that is composed of the previous
cycle’s sample. That is, the old cohort frame is a so-called secondary frame rather than an original frame
constructed from the DRF. The old cohort frame represents the old cohort population through the base
weight, and the old cohort sample represents the longitudinal panel that gets updated at each cycle
through maintenance cut. Given the sample expansion, however, the 2015 SDR needs to redefine its
sampling frame from the original DRF because a fresh new sample requires a fresh new sampling frame.
The DRF is a database that contains educational information for all doctorate recipients from U.S.
universities since 1920. The DRF is updated annually based on the SED which collects information
annually from all doctorates awarded by U.S. institutions about their educational history, funding sources,
and post-doctoral plans.
The target population for the 2015 SDR remains the same as the previous cycle except for the addition of
the new cohort doctorates awarded in academic years 2012 and 2013. Specifically, it includes individuals
who meet the following requirements regardless of residency location:


Received a doctoral degree in an SEH field from a U.S. institution;



Seventy five years of age or younger on 1 February 2015; and



Living in a noninstitutionalized setting on 1 February 2015.

Survey of Doctorate Recpients

FINAL REPORT | 6

Page B-10

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

The final 2015 SDR sampling frame includes 1,102,985 cases, consisting of six groups of doctorates, as
shown in Table 3.1, based on their historical relationship with the existing SDR design. These six groups
may be combined into three broad categories: the panel, the new cohort, and the expansion cohort, as
described below.

TABLE 3.1 The Six Groups of 2015 SDR Frame
Cohort

Frame
Group

Panel

1
2

Expansion
Cohort

3
4
5

New Cohort

6

Description
2013 SDR sample cases that remain eligible for 2015 SDR
Permanently ineligible cases determined in past cycles of
the SDR accrued since 1973 forward (i.e., deceased, no
degree earned, non-U.S. citizens located abroad 2 cycles in
a row)
Maintenance cut cases removed from the sample during
1995-2013 sample selection (proportionally deselected
regardless of response outcome)
Eligible for primary selection from SED 1960-2011, but not
selected
Not eligible for primary selection from SED 1975-2000 that
are now considered eligible (i.e., new graduates with plans
to leave the U.S. after degree award)
New cohort cases from SED 2012 and 2013
Total

SED
Academic
Years (AY)
1960-2011

Number of
Cases
45,936

1964-2011

2,292

1960-2009

64,532

1959-2011

859,891

1975-2000

52,811

2012-2013

77,523
1,102,985

The panel portion of the frame is identified from the 2013 SDR sample of 47,078 doctorates. Of these
cases, 45,936 meet the target population definition and are included in the 2015 sample frame. There are
893 cases determined to be out of scope for the 2015 frame based on information available in the DRF;
887 cases determined to be out of scope due to age ineligibility and 6 cases classified as double
doctorates. There are an additional 249 cases known to be out of scope for the SDR based on information
available from the 2013 SDR. These 249 cases are transferred to the eligible expansion frame case set
and give a chance at selection. Unlike past cycles, the panel cases on the 2015 frame no longer carry a
base weight; they represent no other cases other than themselves on the frame.
The expansion cohort is constructed from the 2013 DRF. These doctorates are needed on the 2015 frame
because they are no longer represented by the panel cases through the base weight, as it was the case in
prior SDR cycles. The expansion cohort frame consists of four groups of cases:


Permanently ineligible cases (Group 2): These are cases that had been selected into the SDR
sample in a previous cycle but were later dropped from the sample due to ineligibility discovered

Survey of Doctorate Recpients

FINAL REPORT | 7

Page B-11

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

during subsequent SDR cycles. These include the deceased, those with no eligible degree, and nonU.S. citizens located abroad for two consecutive SDR cycles. NSF decided to include these known
ineligible cases on the frame to simplify database maintenance as these cases, if dropped from the
frame, will need to be brought back during post-survey data processing.


Maintenance cut cases (Group 3): These are the cases that have been dropped from the SDR
old cohort sample during 1995-2013 through random subsampling to maintain a stable sample size.
Without such maintenance cut, the SDR sample size would have increased over time due to the
addition of a new cohort sample at each cycle.



Non-selected cases (Group 4): These are doctorates that had been eligible for sample selection
but were never selected into the SDR sample in the previous cycles.



Previously ineligible cases (Group 5): These are doctorates that were not eligible for the SDR
based on previous target population definitions, i.e., new graduates with plans to leave the U.S. after
degree award. These cases are eligible for the 2015 SDR.

The 2015 new cohort frame includes 38,140 cases from the 2012 SED and 39,383 from the 2013 SED.
To ensure that all frame cases in these groups are defined consistently, only data available in the 2013
DRF are used as inputs, with the only exception being that data collected in previous SDR cycles are used
to determine age eligibility.5 The protocols for building the 2015 new cohort frame variables are applied
to all eligible cases in the 2015 SDR sampling frame and are described in Section 3.2.
For each frame component, Table 3.2 shows the frequency of eligible and ineligible cases for all records
in the 2013 DRF. This table accounts for eligible cases as well as cases determined to be ineligible for
inclusion on the 2015 SDR frame. The final 2015 SDR frame contains 1,102,985 cases.

Six cases—5 in the expansion cohort and 1 in the panel—are coded as “age eligible” and included in the 2015
frame based on SDR data although the DRF indicates that they are ineligible.
5

Survey of Doctorate Recpients

FINAL REPORT | 8

Page B-12

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

TABLE 3.2 2015 SDR frame eligibility status for all cases in the 2013 DRF
Sample Frame Component
Expansion
New
2013 Panel*
Cohort
Cohort

2015 SDR Frame Status

Overall

Eligible
00

Frame Eligible

45,936

979,526

77,523

1,102,985

Ineligible
01

Age ineligible

887

222,251

8

223,416

03

Deceased, according to the DRF

0

780

14

794

11

Non-SEH doctoral degree field

0

642,287

26,143

668,430

13b

Double Doc; first SEH doctorate earned before
SED 2012/2013

6

261

49

316

Overall
46,829
1,845,105
103,737
1,995,971
* The 2013 SDR sample included 249 cases determined to be ineligible for the 2015 SDR based on information obtained
during the 2013 survey. Most are known deceased cases. These cases are included in the 2015 SDR eligible frame shown
here in the Expansion cohort case count and are given a chance of selection. If selected, these will be immediately
finalized with an ineligible outcome for the 2015 cycle.

3.2 Construction of frame variables
Frame variables are used to support the sample design, including stratification variables, sorting variables,
and sample selection variables. All frame variables are constructed from DRF data with age as the only
exception. The primary variables used to stratify, sort, or assess eligibility for the 2015 SDR frame are as
follows:


PHDFIELD – doctorate degree field reported in the SED



SDRFLD15 – based on PHDFIELD, aggregated recoding of the doctorate degree field



PHDFY – year of doctorate degree award reported in the DRF



CENTURY – based on PHDFY, the century of doctorate degree award



RACETH15 and RACE15– these are racial group assignment derived from ethnicity and race data
reported in the DRF; the component variables from the DRF are renamed ASIAN15, BLACK15,
HISPANIC15, NATIVE15, PACIFIC15, and WHITE15 in the frame



URM15 – based on RACETH15, underrepresented minority status



BIRCIT15 – citizenship at birth based on data reported in the DRF

Survey of Doctorate Recpients

FINAL REPORT | 9

Page B-13

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report



SEX15 – gender reported in the DRF



LOCSTAT15– predicted residency location based on information provided in the SED at the time of
degree award



AGE15 – age of each frame member relative to the 2015 SDR reference date based on age data
reported in the DRF

When frame variables have missing data in the DRF, they are systematically imputed using a set of
imputation rules. Therefore, constructing the frame variables amounts to imputing missing data on these
variables. Missing data in the following frame variables are imputed: RACETH15, URM15, SEX15,
LOCSTAT15, BIRCIT15, and AGE15. PHDFY and PHDFIELD are key design variables that do not
contain any missing data on the DRF.
The details associated with each of these critical frame variables is described below including the
imputation rules and the amount of missing data for each of the frame variables where applicable.

PHDFIELD and SDRFLD15. PHDFIELD is used to define the 2015 SDR sampling strata, and
SDRFLD15 is used to support sample allocation as well as sample selection. PHDFIELD is never missing
in the DRF, so no imputation is required for this variable. Since SDRFLD15 is derived from PHDFIELD,
no imputation is required for SDRFLD15 either. The NSF required that all PHDFIELDs that represented
fields of degree obtained in the 21st century (i.e. from academic year 2001 and later) be included in the
frame and be used to form sampling strata. There are 36 eligible SEH fine fields of degree that are
recorded in the DRF but were discontinued before academic year 2001. Under the 2015 design, each fine
field of degree is its own sampling stratum, but these “20th century-only” fine fields are grouped together
to form a single composite stratum. These discontinued fields contain a total of 26,825 cases. Table 3.3
details these discontinued fields of degree, displaying their codes, descriptions and period in which
sample-eligible frame members earned degrees in those fields.

Survey of Doctorate Recpients

FINAL REPORT | 10

Page B-14

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

TABLE 3.3 20th Century discontinued fields of studya in 2015 SDR Frame
PHDFIELD
Code
007
032
040
042
045
054
060
065
140
156
171
186
205
219
224
225
235
322
323
354
506
521
545
547
549
554
555
562
563
567
573
575
619
679

Fine field of degree
Years in DRF
Animal Husbandry
1962-1982
Plant Protection/Pest Management
1988-1991
Food Sciences
1969-1989
Food Distribution
1994-1994
Soil Sciences
1968-1988
Fish and Wildlife Science
1964-1982
Wildlife
1983-1988
Forestry Science
1964-1988
Hydrobiology
1964-1979
Microbiology/Bacteriology
1961-1982
Genetics
1961-1982
Animal/Plant Physiology
1960-1960
Dentistry
1968-1968
Public Health/Epidemiology
1966-1982
Hospital Administration
1967-1977
Medical/Surgery
1964-1976
Optometry/Ophthalmology
1966-1966
Electrical Engineering
1961-1985
Electronics Engineering
1961-1983
Naval Architecture/Marine Engineering
1983-1991
Astronomy/Astrophysics
1962-1969
Agricultural/Food
1965-1979
Geophysics, Solid Earth
1962-1976
Fuel Technology/Petroleum Engineering
1967-1979
Mineralogy/Petrology/Geological Chemistry
1963-1969
Applied geology
1969-1991
Applied Geology/Geological Engineering
1965-1968
Electron Physics
1984-1991
Electromagnetism
1961-1979
Mechanics
1961-1976
Thermal Physics
1961-1981
Theoretical Physics
1961-1962
Human Engineering
1966-1966
Political Science/Public Administration
1960-1976
Total of 20th century discontinued fields
Total of 21st century fields
Overall

Number of Cases
565
13
1,720
1
1,284
679
142
1,309
132
4,651
1,918
1
1
973
37
25
1
7,157
1,081
64
147
221
428
70
95
279
18
23
135
50
161
2
1
3,441
26,825
1,076,160
1,102,985

aTwo

additional PHDFIELDS, Textile Engineering (375) and Experimental/Comparative & Physiological Psychology (616) were also
identified as discontinued fields. However, no frame case with these PHDFIELDS was age eligible for selection into the 2015 SDR.

Table 3.4 presents the full distribution of SDRFLD15. Please see the field of study coding taxonomies
crosswalk in Appendix E for the collapse of PHDFIELD into SDRFLD15.

Survey of Doctorate Recpients

FINAL REPORT | 11

Page B-15

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

TABLE 3.4 Frame Distribution of SDRFLD15
SDR Field
Chemistry
Physics/Astronomy
Earth/Ocean/Atmospheric Sciences
Mathematics
Computer/Information Sciences
Agricultural Sciences
Medical Sciences
NIH Biological Sciences
Other Biological Sciences
Psychology
Economics
Anthropology/Archaeology/Sociology
Other Social Sciences
Electrical/Electronics/Communications Engineering
Other Engineering
Overall

Total Cases
91,847
65,215
29,405
55,147
31,530
43,898
49,856
112,094
111,123
148,409
45,983
42,754
64,454
59,871
151,399

2013 Panel
3,726
2,587
1,312
2,234
1,434
1,782
2,357
4,651
4,681
6,030
1,916
2,043
2,710
2,545
5,928

Expansion Cohort
83,212
58,284
26,376
48,971
26,358
39,781
42,648
99,048
97,940
134,603
41,396
38,034
57,251
52,410
133,214

New Cohort
4,909
4,344
1,717
3,942
3,738
2,335
4,851
8,395
8,502
7,776
2,671
2,677
4,493
4,916
12,257

1,102,985

45,936

979,526

77,523

PHDFY and CENTURY. PHDFY represents the academic year (called ‘fiscal year’ in the SED) of
doctoral receipt. This variable is used to define the new cohort. It is also used to construct the CENTURY
indicator as one of the sorting variables to support systematic sample selection. For cases earning a degree
in the 20th century (PHDFY<2001), CENTURY is set to “20”; and those earning their degree in the 21st
century (PHDFY≥2001) are set to “21.” Since 1958, when the SED began to field its annual survey,
PHDFY is never missing. Therefore, PHDFY and CENTURY contain no imputed data. Tables 3.5 and
3.6 illustrate the distribution of these variables in the final frame.

TABLE 3.5 PHDFY Distribution by Cohort
PHDFY
1959-1969
1970-1979
1980-1989
1990-1999
2000-2009
2010-2011
2012-2013
Overall

Survey of Doctorate Recpients

2013
Panel
812
5,322
7,174
10,704
17,690
4,234
0

Expansion
Cohort
24,702
153,116
188,901
259,931
285,296
67,580
0

45,936

979,526

New
Cohort
0
0
0
0
0
0
77,523

Total
Cases
25,514
158,438
196,075
270,635
302,986
71,814
77,523

77,523

1,102,985

FINAL REPORT | 12

Page B-16

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

TABLE 3.6 CENTURY Distribution by Cohort
PHDFY
20th century
21st century
Overall

2013
Panel
25,218
20,178

Expansion
Cohort
653,166
326,360

45,936

979,526

New
Cohort
0
77,523

Total
Cases
678,384
424,601

77,523

1,102,985

RACETH15, RACE15 and URM15. RACETH15 represents race and ethnicity, and URM15
represents underrepresented minorities. They are constructed from the separate race/ethnicity variables
ASIAN15, BLACK15, HISPANIC15, NATIVE15, PACIFIC15, and WHITE15 after they are imputed.
RACE15 represents racial group independent of ethnicity and collapses individuals selecting more than
one race as multiracial.
There is a considerable amount of imputation in the 2015 SDR racial variables. Data on race and
ethnicity are entirely missing before 1973 since the SED only started collecting this data with the 1973
cycle. In addition, the racial category of “Native Hawaiian/Pacific Islander” did not exist in the SED until
the 2001 survey. The NSF-approved rules for assigning race and ethnicity are as follows:
1. Use reported data from the SED;
2. When ethnicity is missing, use the U.S. Census Bureau Hispanic surname list and impute

any matches as Hispanic ethnicity (if race is also missing and the surname is Hispanic,
impute the race to white);6
3. When race is missing, and ethnicity is either missing or non-Hispanic, use the GENESYS

Asian surname list7, and logically impute any matches as NH Asian;
4. When ethnicity is still missing, but race is reported, use place of birth to logically impute

ethnicity;
5. When race and ethnicity are both still missing, use place of birth to logically impute race

and ethnicity;

6

The 2015 new cohort cases and 2015 panel cases that joined the panel in the 2013 survey round were updated using
the Hispanic surname list based on the 2000 U.S. Census available as of 2011 located at
http://www.census.gov/genealogy/www/data/2000surnames/index.html. The 2015 panel cases that joined the panel
prior to the 2013 survey round were updated using the Hispanic surname list based on the 1990 U.S. Census.
7
Market Systems Group provides the GENESYS Sampling Systems suite of sampling tools, which includes this
algorithm that matches surnames to an Asian surname list for a nominal fee (http://www.m-sg.com/Web/genesys/index.aspx).

Survey of Doctorate Recpients

FINAL REPORT | 13

Page B-17

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

6. When race and ethnicity are both still missing and place of birth is missing, impute to NH

white.
The crosswalk of birth places to race and ethnicity imputation assignments is located in Appendix F,
Table F2. The sources for race and ethnicity data in the 2015 SDR frame are detailed in Tables 3.7 and
3.9.

TABLE 3.7 Race Data Sources: 2015 SDR Frame
Race Data Source
Self-reported
Surname imputation (Asian)
Birthplace imputation
Hot deck imputation
Default imputation (white)
Overall

Total
Cases
947,363
15,588
95,403
0
44,631

2013
Panel
39,825
549
3,854
0
1,708

Expansion
Cohort
836,286
13,601
90,790
0
38,849

New
Cohort
71,252
1,438
759
0
4,074

1,102,985

45,936

979,526

77,523

After all missing data on race are imputed, the variable RACE15 is created to tabulate race classifications
independent of ethnicity. In cases where one race is identified, the value of RACE15 is assigned to that
race. Otherwise, in cases where self-report indicates more than one race, RACE15 is assigned to ‘6’ for
“more than one race”. The frequencies of RACE15 are shown in Table 3.8 below.

TABLE 3.8 Frame Distribution of RACE15
Ethnicity Data Source
Asian
Black
Native
Pacific
White
More than one race
Overall

Survey of Doctorate Recpients

2013
Panel
11,095
2,605
199
84
31,414
539
45,936

Expansion
Cohort
228,920
29,599
2,646
696
712,301
5,364
979,526

New
Cohort
24,727
3,398
287
112
47,263
1,736
77,523

Total
Cases
264,742
35,602
3,132
892
790,978
7,639
,102,985

FINAL REPORT | 14

Page B-18

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

TABLE 3.9 Ethnicity Data Sources: 2015 SDR Frame
Ethnicity Data Source
Self-reported
Surname imputation (Hispanic)
Birthplace imputation
Hot deck imputation
Default imputation (non-Hispanic)
Overall

Total
Cases
962,550
3,552
87,032
0
49,851

2013
Panel
40,837
199
3,221
0
1,679

Expansion
Cohort
850,247
3,068
83,279
0
42,932

New
Cohort
71,466
285
532
0
5,240

1,102,985

45,936

979,526

77,523

RACETH15 is defined in the following hierarchical manner:


If a case is Hispanic or Latino, assign the case to the Hispanic value regardless of race;



If a case is not Hispanic (NH) and is black, assign the case to the NH black regardless of other race
selections;



If a case is not Hispanic or black, and is Asian, assign the case to the NH Asian regardless of other
race selections;



If a case is not Hispanic, black, or Asian, and is American Indian or Alaskan Native, assign the case
to the NH American Indian regardless of other race selections;



If a case is not Hispanic, black, Asian, or American Indian, and is Native Hawaiian or other Pacific
Islander, assign the case to the NH Pacific Islander regardless of other race selections; and



Otherwise, assign the case to NH white.

The distribution of the resulting race/ethnicity group assignments is shown in Table 3.10.

TABLE 3.10 Race/Ethnicity Assignment: 2015 SDR Frame
Race/ethnicity Group
Hispanic
NH-American Indian
NH-Asian
NH-Black
NH-Pacific Islander
NH-White
Overall

Survey of Doctorate Recpients

Total
Cases
46,114
4,157
266,747
36,046
984
748,937

2013 Panel
3,406
246
11,267
2,628
102
28,287

Expansion
Cohort
37,976
3,494
230,139
29,854
748
677,315

New
Cohort
4,732
417
25,341
3,564
134
43,335

1,102,985

45,936

979,526

77,523

FINAL REPORT | 15

Page B-19

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

URM15 is defined in the following manner:


If a case is Hispanic or Latino, assign the case to URM regardless of race;



If a case is not Hispanic (NH) and is American Indian, black, or Pacific Islander, or reports more than
one race, assign the case to URM; and



If a case is not Hispanic or not American Indian, black, or Pacific Islander, and is Asian or White,
assign the case to non-URM.

The distribution of the resulting URM15 variable is shown in Table 3.11.

TABLE 3.11 Frame Distribution of Underrepresented Minority (URM15)
URM
Yes
No

Total
Cases
90,355
1,012,630

2013 Panel
6,616
39,320

Expansion
Cohort
74,171
905,355

New Cohort
9,568
67,955

Overall

1,102,985

45,936

979,526

77,523

BIRCIT15. The BIRCIT15 variable indicates the sample member’s citizenship at birth, defined as
either U.S. or non-U.S. For all cases in the frame, this information is obtained from the SED. Cases that
do not have valid information on birth citizenship are imputed to be non-U.S. The sources for birth
citizenship data in the 2015 SDR frame files are detailed in Table 3.12. The distribution of the resulting
birth citizenship variable is shown in Table 3.13.

TABLE 3.12 Citizenship at Birth Sources: 2015 SDR Frame
Citizenship at Birth Data Source
Self-reported in SED
Citizenship imputed from DRF with
BIRTHPL and PDLOC
Default imputation (non-U.S. born)
Overall

Survey of Doctorate Recpients

Total Cases
1,055,996

2013 Panel
44,329

Expansion
Cohort
938,676

New Cohort
72,991

1,117
45,872

32
1,575

1,048
39,802

37
4,495

1,102,985

45,936

979,526

77,523

FINAL REPORT | 16

Page B-20

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

TABLE 3.13 Frame Distribution of Citizenship at Birth
Citizenship at Birth Assignment
U.S. born
Not U.S. born
Overall

Total
Cases
656,847
446,138

2013 Panel
27,839
18,097

Expansion
Cohort
589,281
390,245

New
Cohort
39,727
37,796

1,102,985

45,936

979,526

77,523

SEX15. SEX15 is defined from data in the SED. Cases with missing data on sex are imputed to be
female, giving these cases a higher probability of selection. The data sources for the sex variable in the
2015 frame are shown in Table 3.14. The distribution of the resulting sex variable is shown in Table
3.15.

TABLE 3.14 Data Sources for SEX15
Sex Data Source
Self-reported
Default imputation (female)

Total
Cases
1,101,208
1,777

2013 Panel
45,880
56

Expansion
Cohort
977,859
1,667

New
Cohort
77,469
54

Overall

1,102,985

45,936

979,526

77,523

Total
Cases
749,244
353,741

2013 Panel
29,321
16,615

Expansion
Cohort
675,528
303,998

New
Cohort
44,395
33,128

1,102,985

45,936

979,526

77,523

TABLE 3.15 Frame Distribution of SEX15
Sex Assignment
Male
Female
Overall

LOCSTAT15. The LOCSTAT15 variable indicates the last known residence location of the sample
member prior to the 2015 SDR; it is either in the U.S. or out of the U.S. The 2010 SDR was the first
cycle to use this variable.8 In the past two cycles, this variable was used to distinguish between NSDR and
ISDR cases. Under the 2015 SDR design, it is used as one of the sorting variables. For all cases in the

8

For more details about the LOCSTAT variable development for the 2010 SDR and continued for the 2013 and
2015 SDR, see the memoranda “2010 SDR Sample Frame Development Memo #3 – Sample Member Location
Variable” sent to Daniel Foley and Steve Cohen, NSF, on April 23, 2010 from Karen Grigorian, NORC, and Brenda
Cox, SRA, and “2015 SDR Frame Decisions – Frame File Layout” sent to Steve Proudfoot, NSF, on March 28,
2014 from Karen Grigorian and Lance Selfa, NORC.

Survey of Doctorate Recpients

FINAL REPORT | 17

Page B-21

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

2015 frame, LOCSTAT15 is derived from responses to the SED question about planned post-graduation
location. Any cases with no residency data from the SED are imputed to be in the U.S.
The sources for the location data in the 2015 SDR frame files are detailed in Table 3.16. The distribution
of the resulting location variable is shown in Table 3.17.

TABLE 3.16 Location Data Sources: 2015 SDR Frame
Location Data Source
SED
Default imputation (in the U.S.)
Overall

Total Cases
1,064,194
38,791

2013 Panel
44,297
1,639

Expansion
Cohort
948,386
31,140

New Cohort
71,511
6,012

1,102,985

45,936

979,526

77,523

TABLE 3.17 Frame Distribution of LOCSTAT15
Location Assignment
In the U.S.
Out of the U.S.
Overall

Total Cases
987,174
115,811

2013 Panel
40,832
5,104

Expansion
Cohort
877,149
102,377

New Cohort
69,193
8,330

1,102,985

45,936

979,526

77,523

BIRCIT15. The BIRCIT15 variable indicates the sample member’s citizenship at birth, defined as
either U.S. or non-U.S. For all cases in the frame, this information is obtained from the SED. Cases that
do not have valid information on birth citizenship are imputed to be non-U.S. The sources for birth
citizenship data in the 2015 SDR frame files are detailed in Table 3.18. The distribution of the resulting
birth citizenship variable is shown in Table 3.19.

TABLE 3.18 Citizenship at Birth Sources: 2015 SDR Frame
Citizenship at Birth Data Source
Self-reported in SED
Citizenship imputed from DRF with
BIRTHPL and PDLOC
Default imputation (non-U.S. born)
Overall

Survey of Doctorate Recpients

Total Cases

2013 Panel

Expansion
Cohort

New Cohort

1,055,996

44,329

938,676

72,991

1,117
45,872

32
1,575

1,048
39,802

37
4,495

1,102,985

45,936

979,526

77,523

FINAL REPORT | 18

Page B-22

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

TABLE 3.19 Frame Distribution of Citizenship at Birth
Citizenship at Birth Assignment
U.S. born
Not U.S. born

Total Cases
656,847
446,138

2013 Panel
27,839
18,097

Expansion
Cohort
589,281
390,245

New Cohort
39,727
37,796

1,102,985

45,936

979,526

77,523

Overall

AGE15. The AGEYR15variable indicates the sample member’s year of birth and is used to create
AGE15 and IAGE15. The primary sources of AGEYR15 are birth year data reported on the SED,
supplemented with birth year information collected from the SDR. Any missing data on AGEYR15 are
imputed from sample members’ bachelor’s degree year, if known, or from their doctorate degree year,
which is known for all sample members. The birth year imputation rules assume that sample members are
18 when they earned their bachelor’s degree, 21 when they earned their doctoral degree. These age
assumptions may not be realistic; they are intended to minimize frame undercoverage which could arise if
we eliminate those doctorates who are missing birth year but have earned a doctoral degree at a young
age. The sources for age in the 2015 SDR frame files are detailed in Table 3.20. The distribution of the
resulting age variable is shown in Table 3.21.

TABLE 3.20 Age Source: 2015 SDR Frame
Age Data Source
Self-reported in SED
BA Year Imputation
PhD Year Imputation
Overall

Survey of Doctorate Recpients

Total Cases
1,051,578
15,875
35,532

2013 Panel
43,833
652
1,451

Expansion
Cohort
935,485
14,022
30,019

New Cohort
72,260
1,201
4,062

1,102,985

45,936

979,526

77,523

FINAL REPORT | 19

Page B-23

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

TABLE 3.21 Frame Distribution of Age
Age Assignment
Under 35
35-39
40-44
45-49
50-54
55-59
60-64
65-75

Total Cases
111,950
129,168
123,227
123,076
127,462
121,617
123,245
243,240

2013 Panel
3,445
6,579
6,872
6,063
5,357
4,887
4,545
8,188

Expansion
Cohort
57,776
106,668
110,924
114,770
120,737
115,755
118,140
234,756

New Cohort
50,729
15,921
5,431
2,243
1,368
975
560
296

1,102,985

45,936

979,526

77,523

Overall

HCAPIN15. The disability status variable, HCAPIN15, is not used in the sampling process, but has
been included on the frame and in this reporting section as disability status is important to future reporting
and analysis. The HCAPIN15 variable indicates the sample member’s disability status – either disabled
or not disabled. For all cases in the 2015 sample, the disability information is obtained from the SED,
which has gathered data on disability since 1985. The historical data on disability in the DRF are
recorded in the variable DISABILITY1. Starting with the 2012 cycle, the SED is using the identical
disability question and code frame (summarized in DISABILITY2) as does the SDR. Therefore, defining
disability status for frame cases requires using both DISABILITY1 and DISABILITY2. Cases that never
reported disability status, including those who completed the SED before the disability questions were
introduced to the SED, are imputed to be non-disabled. This imputation means that the proportion of
disabled doctorates in the population should be much higher than known on the frame because the cases
with unknown disability status have been imputed to be not disabled. The sources for disability status in
the 2015 SDR frame files are presented in Table 3.22. The distribution of the resulting disability status
variable is shown in Table 3.23.

Survey of Doctorate Recpients

FINAL REPORT | 20

Page B-24

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

TABLE 3.22 Disability Status Source: 2015 SDR Frame
Disability Status Data Source
Self-reported in SED
Default imputation (not disabled)
Overall

Total Cases
679,405
423,580

2013 Panel
30,631
15,305

Expansion
Cohort
579,270
400,256

New Cohort
69,504
8,019

1,102,985

45,936

979,526

77,523

TABLE 3.23 Frame Distribution of Disability Status
Disability Status Assignment
Disabled
Not disabled
Overall

Survey of Doctorate Recpients

Total Cases
13,866
1,089,119

2013 Panel
559
45,377

Expansion
Cohort
9,220
970,306

New Cohort
4,087
73,436

1,102,985

45,936

979,526

77,523

FINAL REPORT | 21

Page B-25

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

SUMMARY OF FRAME VARIABLES DATA SOURCES. Table 3.24 summarizes the data
sources for the key frame variables subject to imputation. These results are shown by variable and by the
three main sample frame components.

TABLE 3.24 Data Sources for Sample Frame Variables Subject to Imputation and/or
Derivation: 2015 SDR Frame
2015 SDR Sample Frame Cases
Sample Frame
Component
2013 Panel

Sample Frame Variable
Race (RACE15)
Ethnicity (HISPANIC15)
Sex (SEX15)
Location (LOCSTAT15)
Citizenship at birth (BIRCIT15)
Disability status (HCAPIN15)
Birth year (AGEYR15)

Reported Values in
the Final Frame
39,825
40,837
45,880
44,297
44,329
30,631
43,828

Imputed from a
Non-default Rule
4,403
3,420
n/a
n/a
32
n/a
2,108

Assigned Default
Imputation
1,708
1,679
56
1,639
1,575
15,305
n/a

Expansion Cohort

Race (RACE15)
Ethnicity (HISPANIC15)
Sex (SEX15)
Location (LOCSTAT15)
Citizenship at birth (BIRCIT15)
Disability status (HCAPIN15)
Birth year (AGEYR15)

836,286
850,247
977,859
948,386
938,676
579,270
935,484

104,391
86,347
n/a
n/a
1,048
n/a
44,042

38,849
42,932
1,667
31,140
39,802
400,256
n/a

New Cohort

Race (RACE15)
Ethnicity (HISPANIC15)
Sex (SEX15)
Location (LOCSTAT15)
Citizenship at birth (BIRCIT15)
Disability status (HCAPIN15)
Birth year (AGEYR15)

71,252
71,466
77,469
71,511
72,991
69,504
72,260

2,197
817
n/a
n/a
37
n/a
5,263

4,074
5,240
54
6,012
4,495
8,019
n/a

Overall

Race (RACE15)
Ethnicity (HISPANIC15)
Sex (SEX15)
Location (LOCSTAT15)
Citizenship at birth (BIRCIT15)
Disability status (HCAPIN15)
Birth year (AGEYR15)

947,363
962,550
1,101,208
1,064,194
1,055,996
679,405
1,051,572

110,991
90,584
n/a
n/a
1,117
n/a
51,413

44,631
49,851
1,777
38,791
45,872
423,580
n/a

Survey of Doctorate Recpients

FINAL REPORT | 22

Page B-26

SDR OMB Package for the 2017 Cycle

NORC

4

| SDR 2015 Sample Design and Implementation Report

Sample Design

4.1 Precision Requirements
The SDR sample design has undergone several major changes since its inception in 1973, reflecting
changing estimation objectives and budgetary situations. For the past few cycles, the SDR was designed
to produce estimates by various analytical domains defined by aggregated field of degree, gender, race
and ethnicity, citizenship at birth, and disability status. The existing SDR sample stratification and
allocation system reflects these estimation goals. A new significant change to the 2015 SDR design is a
major sample size expansion to support employment outcome estimates by fine field of degree (FFOD).
The sample size of the 2013 SDR is 47,078 cases, while the current expansion calls for a sample size
increase to 120,000 cases for the 2015 SDR.
With the marked increase in the overall sample size, the estimation capability of the 2015 SDR is
expected to increase substantially. To guide the SDR sample redesign, the NSF specified general
requirements regarding the analytical objectives of the 2015 sample. The following comes from the
document “Requirements of Sample Expansion and Sample Redesign of the Survey of Doctorate
Recipients” and its addendum (full text in Appendix A):


Specified precision at FFOD level: Producing employment outcome estimates at the SED fine
field of degree (FFOD) level for the entire SDR eligible population regardless of their residential
location and time of receiving doctorates. The precision is required to be within 5% margin of errors
at the 95% confidence level for important outcomes.9



Maintain precision of key 2013 domains: The overall expanded sample should maintain the
existing 2013 estimation capability at the aggregated degree levels and for domains defined by
various demographic characteristics currently used in NCSES publications. This set of requirements
will be examined by comparing the estimates’ precision levels derived under the proposed designs
and the actual 2013 SDR sample results.

The expansion of the SDR, along with its new estimation objectives, calls for a significant redesign of the
SDR sample. The objectives of the SDR redesign include meeting the newly defined estimation
objectives, resolving any longstanding sampling issues to improve efficiency, creating a unified sample
design for NSDR (National SDR) and ISDR (International SDR), and constructing a flexible and

9

This requirement was later relaxed to a 5% margin of error at the 90% confidence level.

Survey of Doctorate Recpients

FINAL REPORT | 23

Page B-27

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

sustainable design for the growing demands of SDR data. The rest of this section describes design
considerations and the final design chosen for the 2015 SDR sample redesign.
4.2 Design Approaches Considered
This subsection briefly discusses the two broad design approached considered for the 2015 SDR design.
More detailed descriptions of these approaches and design options are presented in Appendix B.
Broadly speaking, the 2015 SDR frame may be thought of as consisting of two overlapping frames.
Frame A, which contains the first four frame groups in Table 3.1, covers the portion of the SDR
population that is represented by the existing 2013 SDR sample, or the panel sample. Frame B
encompasses Frame A as well as the population that is not represented by the panel sample, namely,
frame groups 5 and 6 in Table 3.1, including new cohort doctorates awarded in 2012 and 2013 as well as
those that were excluded from the SDR frame prior to 2000. Therefore, Frame A is completely nested
within frame B which includes all six frame groups in Table 3.1
Under the guidance of NCSES, NORC considered two major design approaches for the 2015 SDR
redesign: dual frame design, and single frame design with a fresh new sample, as discussed below.
Dual Frame Design
The main motivation of the dual frame design is to preserve the existing SDR panel, both to reduce data
collection cost and improve data utility. Under this approach, the 2015 SDR sample would include two
independent and partially overlapping samples: the panel sample (i.e., the existing 2013 SDR sample)
from Frame A and an independent expansion sample from Frame B. For estimation, these two samples
would be combined through a dual frame method to derive the overall 2015 estimates.
Let’s designate the existing panel sample from Frame A as sample

a

and the new expansion sample to

be selected from Frame B sample b . The two samples would be first be weighted separately according to
their respective sample design, generating two sets of sampling weights

w(aj )

and

w(bj )

. Then a single set

of weights would be created for the combined sample through a combining factor. For a sample member j
selected into either sample, its sampling weight will be calculated as

Survey of Doctorate Recpients

FINAL REPORT | 24

Page B-28

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

w(jb ) , if j  sample b and not frame A
 (a)
w j , if j  sample a and frame A
wj  
(b )
1   w j , if j  sample b and frame A
 (a)
(b )
w j  1   w j , if j  both sample a and sample b and frame A

The first category includes cases that are in Frame B but not in Frame A, representing the nonoverlapping portion of the frame. For this portion of the population, the estimate will be based only on a
subset of sample b . The other three categories include samples from the overlapping portion of the
population. For this portion of the population, there are two estimates, one based on sample

a and the

other based on a subset of sample b .
The combining factor

 is defined as


In this expression,
b
neff

a
neff

(a)
neff
(a)
(b )
neff
 neff

is the effective sample size associated with sample

a selected from Frame A; and

is the effective sample size associated with sample b selected from Frame A. The effective sample

size is the expected number of complete surveys divided by the design effect due to unequal weighting.
With a single set of weights defined, the usual Horvitz-Thompson estimator can be used to derive point
estimates after proper adjustments for eligibility, nonresponse, and frame coverage.

Under this dual frame estimation approach, the effective sample size from sample a
each fine field. Therefore, sample size determination is to estimate

(b )
neff

such that

(a )
neff

, is known for

1.96 * V ( pˆ )  .05

(b )

n
when estimating a population proportion P for a fine field. The quantity eff can then be converted to a
nominal sample size based on the design effect and expected completion rate.

Survey of Doctorate Recpients

FINAL REPORT | 25

Page B-29

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

Single Frame with a Fresh New Sample

Under this design approach, a fresh new sample would be selected from Frame B, and the
existence of the panel sample has no bearing on the 2015 SDR design. The sample will be
stratified by FFOD only. Sample allocation to the strata is determined to balance the competing
estimation goals discussed earlier. For a sample member j , its sampling weight will be

wj 

where

pj

1
pj

is the inclusion probability under the sample design. The Horvitz-Thompson estimator can be

used to derive point estimates after proper adjustments for eligibility, nonresponse, and frame coverage.
4.3 Sample Allocation
We now discuss the Fresh New Sample approach that the NSF decided to adopt. Under this approach, the
SDR sample is stratified by FFOD to 216 sampling strata, including the discontinued 20th century fields
strata. As discussed below, the sample of 120,000 cases is allocated to the strata in two steps. The twostep allocation is implemented to achieve a balance between satisfying the fine field level estimation
requirement and maintaining the existing estimation capabilities of the SDR with regard to the key
analytical domains under the prior design. In particular, the second step is intended to improve the
representation of the population by the 15-category aggregate degree fields.10 This measure is
implemented because, after step one, aggregate fields with a large population but consisting of a small
number of fine fields (e.g., Computer/Information Sciences) are underrepresented, while aggregate fields
with a small population but consisting of a large number of fine fields (e.g., Agricultural Sciences) are
overrepresented. The objective of the two step allocation is to make the representation of aggregate
degree fields more proportional to the population.
NSF decided to allocate 1,000 cases to the stratum that represents the discontinued fields. The remaining
119,000 cases are allocated to the other 215 strata as described in Steps 1 and 2 below.

10

The 15 categories are: Chemistry, Physics/Astronomy, Earth/Ocean/Atmospheric Sciences, Mathematics,
Computer/Information Science, Agricultural Sciences, Medical Sciences, NIH Biological Sciences, Other Biological Sciences,
Psychology, Economics, Anthropology/Archeology/Sociology, Other Social Sciences, Electrical/Electronics/Communications,
and Other engineering.

Survey of Doctorate Recpients

FINAL REPORT | 26

Page B-30

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

Step 1
The first step features an equal allocation to each stratum. For a population proportion centered at 50
percent, the first step allocation is designed to achieve a 5 percent margin of error (MOE) with 90%
confidence. The following summarizes how the first step allocation is derived.


Assume that the population proportion is 𝑃̂ = 0.50 to derive the most conservative sample size
estimate;



2

𝑍90
The number of complete surveys needed per stratum is 𝑛𝑐𝑜𝑚𝑝𝑙𝑒𝑡𝑒𝑑 = (𝑃̂ ∗ (1 − 𝑃̂)) ∗ (𝑀𝑂𝐸
) , where

Z90 is the critical value of the standard normal distribution for a 2-tailed test at a 90% confidence level
(approximately 1.645), and MOE = 5%


Assume that the completion rate is 70%, the number of cases to sample per stratum is 𝑛𝑠𝑎𝑚𝑝𝑙𝑒𝑑 =
𝑛𝑐𝑜𝑚𝑝𝑙𝑒𝑡𝑒𝑑
0.70



In strata where the allocated sample size exceeds the number of cases on the frame, the stratum
sample size is set to equal to the frame size, 𝑛1𝐹𝐹𝑂𝐷 = min(𝑁𝐹𝐹𝑂𝐷 , 𝑛𝑠𝑎𝑚𝑝𝑙𝑒𝑑 ), where 𝑁𝐹𝐹𝑂𝐷 = Total
frame size for the FFOD



The total sample allocated in step one is 𝑇𝑜𝑡𝑎𝑙_𝑆𝑡𝑒𝑝1 = ∑𝐹𝐹𝑂𝐷 𝑛1𝐹𝐹𝑂𝐷

The nominal sample size allocated to each stratum is 387, which will produce 271 complete surveys with
an expected completion rate of 70%, enough to satisfy the stratum level precision requirements. Note that
the finite population correction factor (FPC) is not incorporated in the sample size estimation described
above. When sampling from a finite population, the variance of the mean is reduced by a factor
( N  n) / N , called FPC, where N is the population size and

n is the sample size. For sample size

estimation, the application of the FPC helps to reduce the sample size necessary to meet the specified
precision requirement. To be conservative, NCSES and NORC decided not to apply the FPC when
estimating the sample size per stratum. The effect of ignoring the FPC is to overestimate the standard
error of the mean; but it offers additional insurance in case the completion rate is lower than expected.
For fine fields with less than 387 cases in the frame, all frame cases are included in the sample. A total of
77,965 cases, or 65 percent of the total sample, are allocated in the first step. The first step allocation
represents the minimum allocation to each stratum and assures that the final sample will achieve the
required level of precision at the fine field level.

Survey of Doctorate Recpients

FINAL REPORT | 27

Page B-31

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

Step 2
The second step allocates the remaining 35 percent of the sample proportionately to the 15 SDR broad
field categories, as represented by the design variable SDRFLD15. This second step allocation is designed
to allocate the balance of the sample in such a way as to minimize the variation in sampling weights for
the full sample, given the first step allocation. The second step allocation is carried out as follows:


First, calculate the expected proportional allocation to the 15 broad field categories based on the
overall frame distribution across the 15 broad fields. The fine fields within the discontinued 20th
century fields stratum also participate in this calculation. The expected proportional allocation to each
broad field of degree (BFOD) is: 𝐸𝑥𝑝𝑒𝑐𝑡𝑒𝑑_𝐴𝑙𝑙𝑜𝑐𝑎𝑡𝑖𝑜𝑛𝐵𝐹𝑂𝐷 = 𝑇𝑜𝑡𝑎𝑙_𝑆𝑎𝑚𝑝𝑙𝑒 ∗

𝑁𝐵𝐹𝑂𝐷
,
𝑁

where

𝑁𝐵𝐹𝑂𝐷 is the total number of frame cases per broad field.


Second, subtract the total first step allocation for each broad field from the expected allocation to get
the second step allocation per broad field category. For those broad field categories (Agricultural
Sciences and Earth/Ocean/Atmospheric Sciences) that have already exceeded the expected allocation
after the first step allocation, allocate 0 cases in the second step. If the step one allocation per broad
field is 𝑆𝑡𝑒𝑝1𝐵𝐹𝑂𝐷 = ∑𝐹𝐹𝑂𝐷∈𝐵𝐹𝑂𝐷 𝑛1𝐹𝐹𝑂𝐷 , then the step two allocation to each broad field is



𝑆𝑡𝑒𝑝2_𝐴𝑙𝑙𝑜𝑐𝑎𝑡𝑖𝑜𝑛_𝑎𝐵𝐹𝑂𝐷 = max(0, 𝐸𝑥𝑝𝑒𝑐𝑡𝑒𝑑_𝐴𝑙𝑙𝑜𝑐𝑎𝑡𝑖𝑜𝑛𝐵𝐹𝑂𝐷 − 𝑆𝑡𝑒𝑝1𝐵𝐹𝑂𝐷 ) . The step two
allocation to the broad field is then adjusted to reflect the fact that two broad fields would not receive
additional allocation in step two. The adjusted step two allocation to the broad field
is: 𝑆𝑡𝑒𝑝2_𝐴𝑙𝑙𝑜𝑐𝑎𝑡𝑖𝑜𝑛_𝑏𝐵𝐹𝑂𝐷 = 𝑆𝑡𝑒𝑝2_𝐴𝑙𝑙𝑜𝑐𝑎𝑡𝑖𝑜𝑛_𝑎𝐵𝐹𝑂𝐷 ∗

𝑇𝑜𝑡𝑎𝑙_𝑆𝑡𝑒𝑝2
,
𝑇𝑜𝑡𝑎𝑙_𝑆𝑡𝑒𝑝2_𝐴𝑙𝑙𝑜𝑐𝑎𝑡𝑖𝑜𝑛_𝑎

where

𝑇𝑜𝑡𝑎𝑙_𝑆𝑡𝑒𝑝2 = 119,000 − 𝑇𝑜𝑡𝑎𝑙_𝑆𝑡𝑒𝑝1, and 𝑇𝑜𝑡𝑎𝑙_𝑆𝑡𝑒𝑝2_𝐴𝑙𝑙𝑜𝑐𝑎𝑡𝑖𝑜𝑛_𝑎 =
∑ 𝑆𝑡𝑒𝑝2_𝐴𝑙𝑙𝑜𝑐𝑎𝑡𝑖𝑜𝑛_𝑎𝐵𝐹𝑂𝐷


Third, within each broad field category, proportionately allocate the second step allocation to each
fine field stratum based on the frame count per fine field stratum11. This allocation is calculated as
𝑛2𝐹𝐹𝑂𝐷 = 𝑆𝑡𝑒𝑝2_𝐴𝑙𝑙𝑜𝑐𝑎𝑡𝑖𝑜𝑛_𝑏𝐵𝐹𝑂𝐷 ∗ (∑

𝑁𝐹𝐹𝑂𝐷

𝐹𝐹𝑂𝐷∈𝐵𝐹𝑂𝐷 𝑁𝐹𝐹𝑂𝐷



)

The final allocation to each fine field stratum is the sum of step one and step two allocations or the
frame size if the sum exceeds the frame size, i.e., 𝐹𝑖𝑛𝑎𝑙_𝐴𝑙𝑙𝑜𝑐𝑎𝑡𝑖𝑜𝑛𝐹𝐹𝑂𝐷 = min(𝑁𝐹𝐹𝑂𝐷 , 𝑛1𝐹𝐹𝑂𝐷 +
𝑛2𝐹𝐹𝑂𝐷 )

11

The fine fields that make up the discontinued fields stratum do not receive any allocation.

Survey of Doctorate Recpients

FINAL REPORT | 28

Page B-32

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

When a fine field does not have enough cases to support the final allocation, the total allocation is equal
to the frame total. In that situation, the shortage is allocated to the discontinued field stratum. For this
reason, the final allocation to the discontinued stratum is slightly over 1,000. In the final allocation, a total
of 118,916 cases are allocated to the 215 fine field strata, with the remaining 1,084 cases allocated to the
216th stratum representing the 20th century discontinued fields. Appendix Table G.1 shows the step one,
step two, and total allocation by 2015 sampling stratum. For comparison, Appendix Table G.2 shows the
total sample allocation by 2013 sampling stratum.

Survey of Doctorate Recpients

FINAL REPORT | 29

Page B-33

SDR OMB Package for the 2017 Cycle

NORC

5.

| SDR 2015 Sample Design and Implementation Report

Sample Selection

Within each of the 216 strata, a random sample is selected systematically with probability proportionate
to size (PPS). PPS sampling is adopted as a vehicle to oversample underrepresented racial and ethnic
minorities (URM), women, and the 2013 panel cases. Oversampling of URM and women allows the
sample to sustain the estimation capabilities under the prior SDR design. The addition of a panel
oversample maintains the simplicity of a fresh new sample, but allows for limited longitudinal analysis
using earlier waves of the SDR. The oversampling is achieved by assigning a measure of size to each
frame member and then selecting the sample systematically with PPS within each stratum. Each frame
case is assigned a measure of size (MOS), as follows:


Male URM: 2.0



Female URM: 2.5



Female non-URM: 1.5



Panel cases, regardless of gender or URM status: 5.0



All other cases: 1.0

p  nh * MOSi /
Under PPS sampling, the selection probability for a case i in stratum h is hi

 MOS ,
i

where nh is the stratum sample size, MOSi is the measure of size for case i , and the summation is over
all frame cases within a stratum. For cases with large MOS, the selection probability may be equal to or
greater than 1. Such cases are identified first because they would be selected with certainty into the
sample. For each stratum, the certainty cases are identified as follows:
1. Sort the frame cases in descending order by MOS;
2. Sum MOS across all frame cases to get the total MOS Total_MOS;
3. Denote the total allocated sample size as Total_Allocated ;
4. Carry out the following procedures, starting at the top of the sorted frame

a. If MOS ≥ (Total_MOS/Total_Allocated) then this case is a certainty selection:
i. Set sampling weight = 1;
ii. Move the case to a separate file that contains all certainty cases;

Survey of Doctorate Recpients

FINAL REPORT | 30

Page B-34

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

iii. Recompute Total_MOS = Total_MOS – MOS;
iv. Recompute Total_Allocated = Total_Allocated – 1;
v. Return to 4.a to evaluate the next case on the sorted frame.
b. If MOS < (Total_MOS/Total_Allocated) then this case and all cases following this
case are not certainty cases
i. All non-certainty cases constitute the frame for systematic PPS sampling;
ii. The final sample consists of the certainty cases and those selected from the rest
of the frame through systematic PPS sampling.
Before systematic sampling, the frame is sorted by the following variables to impose an implicit
stratification within each stratum. The sorting variables are:


CENTURY



RACETH15



BIRCIT15



SEX15



LOCSTAT15



SDRFLD15



PHDFY

The purpose of implicit stratification is to improve the representation of the sample with respect to the
sorting variables. Note that sorting by SDRFLD15 (the 15 broad fields) is only effective in the
discontinued stratum because all the other strata represent a single field. The purpose of sorting by
SDRFLD15 within the discontinued fields is to ensure a proportional representation of the broad fields
within the discontinued stratum.
With the certainty cases set aside, the SAS procedure PROC SURVEYSELECT is used to carry out the
systematic sampling within each stratum. Systematic sampling selects cases at a fixed interval throughout
the stratum after a random start. PROC SURVEYSELECT uses a fractional sampling interval to provide
exactly the specified sample size. The interval within a stratum is
(selections) for a case is

𝑛∗ 𝑀𝑂𝑆𝑖
.
∑ 𝑀𝑂𝑆𝑖

∑ 𝑀𝑂𝑆𝑖
𝑛

. The expected number of hits

The sampling weight is the inverse of the expected number of hits. The

final sample includes all the certainty cases and those selected through PROC SURVEYSELECT.

Survey of Doctorate Recpients

FINAL REPORT | 31

Page B-35

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

Subsequent to selection, the selected sample along with the quality assurance procedures were sent to the
NSF for review and approval.12

12

See the memo sent to Emilda Rivers and Steve Proudfoot at the NSF from Michael Yang, Lance Selfa, and Karen
Grigorian at NORC entitled “2015 SDR – Sample Selection Quality Control Procedures and Results” issued on 16
February 2015 as well as the sample review tables in the attachment file named ” 2015 SDR Expansion Sample
Allocation and Review Tables.zip.”

Survey of Doctorate Recpients

FINAL REPORT | 32

Page B-36

SDR OMB Package for the 2017 Cycle

NORC

6.

| SDR 2015 Sample Design and Implementation Report

SDR 2017 and Beyond

The NSF adopted the fresh new sample approach to meet new estimation goals, to simplify the sample
design and estimation procedures, and to resolve longstanding frame and sampling issues accumulated
over time. With a fresh new sample, the 2015 SDR represents a significant turning point in the hitherto
longitudinal sample design dating back to 1973. Although the SDR sample has undergone several major
redesigns, for the first time since the 1975 SDR, the new sample does not include a substantial panel
component. Instead, the 2015 SDR is expected to be the starting point of a new longitudinal data series.
The NSF has not yet provided any guidelines for the 2017 SDR sample design, but it is most likely that
the 2015 SDR sample will form the sampling frame for the old cohort sample for 2017 while a new
cohort sample will be selected from the new cohort frame consisting of SEH doctorates awarded in 2014
and 2015 academic years. Unless the analytical objectives change, we expect the 2017 SDR to follow the
same stratification scheme and sample collection procedures.
Assuming that the 2017 SDR sample size will be kept at the current level, a maintenance cut to the old
cohort sample will be necessary while adding a new cohort sample. To preserve the oversampling of
URM, women, and the 2013 panel cases, we expect the 2017 SDR old cohort sample to be a
straightforward equal probability random sample within each stratum. Like the 2015 SDR, the 2017 new
cohort sample will be stratified by FFOD, and the sample allocation will be guided by the analytical
objectives specified by the NSF. If the analytical objectives stay the same, for example, the 2015
allocation procedures may be adapted to allocate the 2017 new cohort sample.

Survey of Doctorate Recpients

FINAL REPORT | 33

Page B-37

SDR OMB Package for the 2017 Cycle

NORC

7.

| SDR 2015 Sample Design and Implementation Report

Concluding Remarks

The current SDR sample redesign may have significant implications for post-survey data processing
procedures such as weighting adjustment, missing data imputation, and variance estimation. We conclude
this report by discussing these likely implications. Additional research will be needed to modify these
procedures if they turn out to be necessary.
Starting from the 2010 cycle, the SDR has moved from the traditional weighting class method to the
model-based propensity score method for noncontact and nonresponse weighting adjustments. A
propensity score is predicted from a logistic regression model for both eligibility determination and
interview cooperation; these scores are then used, either directly or indirectly, to adjust the original
sampling weight to compensate for noncontact and nonresponse prior to a final poststratification
adjustment. Given the changes to the sample design, the 2015 noncontact and nonresponse models are
likely to be different from the models of the prior rounds. For example, additional predictor variables may
need to be included to capture the noncontact and nonresponse pattern associated with the expansion
cohort cases that appear in the SDR sample for the first time. Furthermore, with the newly defined
estimation goals, the poststratification procedures may also need to be revamped to match the poststrata
with analysis domains. For example, it may be necessary to define the poststrata by fine field of degree,
among other key factors.
The SDR conducts extensive missing data imputation, using basically the same set of imputation
procedures in the past few cycles. With the sample redesign, these imputation procedures may need to be
adapted to the 2015 SDR. For example, many variables are imputed through multivariate regression
models; these models may need to be updated to reflect the new sampling frame and relevant features of
the sample design. The sorting variables under hot deck imputation may also need to be updated by new
variables and new models.
The SDR has used a replication method for variance estimation to account for its complex design features
that cannot be adequately captured with the Taylor Series linearization method. The current successive
difference replication method (SDRM) is designed for systematic samples where implicit stratification
puts similar cases close to each other on the sampling frame. Under the existing strategy initially
developed by the Census Bureau, the replicates are formed as if a systematic sample is selected from a
single large stratum. While this method may effectively account for the reduction in variance resulting

Survey of Doctorate Recpients

FINAL REPORT | 34

Page B-38

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

from implicit stratification within the explicit strata, it may not account for the impact of explicit
stratification on the sampling variance. NORC has proposed to the NSF to explore alternative variance
estimation approaches to improve potentially both statistical and cost efficiency. With the 2015 sample
redesign, it may be a good time to revisit the current procedures. For example, we would like to compare
the SDRM with a simpler and more efficient procedure based on a Taylor Series or Jackknife method. In
case the SDRM does not lead to noticeable reduction in the variances, the Taylor series or Jackknife
methods would make more efficient alternatives.
In addition to SDRM replicate weights, the SDR also provides estimated Generalized Variance Functions
(GVFs) for a set of key NSDR and ISDR domains. The GVFs are valuable because they provide a
mechanism for data users to compute the variance of estimates not directly provided by the SDR. With
the sample redesign, it may be necessary to redefine the GVF definitions so that they match with key
analysis domains.

Survey of Doctorate Recpients

FINAL REPORT | 35

Page B-39

SDR OMB Package for the 2017 Cycle

NORC

| SDR 2015 Sample Design and Implementation Report

Appendices removed

Survey of Doctorate Recpients

Page B-40

SDR OMB Package for the 2017 Cycle

Attachment C
First Federal Register Announcement

Survey of Doctorate Recipients

Page C-1

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page C-2

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page C-3

SDR OMB Package for the 2017 Cycle

Attachment D
Draft 2017 SDR Questionnaire

Survey of Doctorate Recipients

Page D-1

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-2

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-3

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-4

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-5

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-6

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-7

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-8

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-9

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-10

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-11

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-12

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-13

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-14

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-15

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-16

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-17

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-18

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-19

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-20

SDR OMB Package for the 2017 Cycle

Survey of Doctorate Recipients

Page D-21

SDR OMB Package for the 2017 Cycle

Attachment E
Draft 2017 SDR Survey Mailing Materials
The 2017 Survey of Doctorate Recipients (SDR) contact materials are consistent with materials utilized in past
cycles. The various materials used in outreach to the SDR sample members are listed below. All letters will be
printed on the SDR 2017 letterhead found on pages E-2 and E-3.
































Mail Start Mode Prenotice Letter for 2015 Cooperative (M_PN1)
Mail Start Mode Prenotice Letter for 2015 Nonresponse Retirees/2015 NIR/Partial Responders (M_PN2)
Mail Start Mode Prenotice Letter for 2017 New Cohort (M_PN3)
CATI Start Mode Prenotice Letter for 2015 Cooperative (C_PN1)
CATI Start Mode Prenotice Letter for 2015 Nonresponse Cases/2017 New Cohort (C_PN2)
Web Start Mode Initial Notice sent via USPS to 2015 Cooperative cases (W_IC1_u)
Web Start Mode Initial Notice sent via USPS to 2015 Cooperative Retiree cases (W_IC2_u)
Web Start Mode Initial Notice sent via USPS to 2015 Nonresponse Cases (W_IC3_u)
Web Start Mode Initial Notice sent via USPS to 2015 Cooperative Green Appeal (W_IC5_u)
Web Start Mode Initial Notice sent via USPS to 2017 New Cohort and 2015 Panel Locating Problems
(W_IC7_u)
Web Start Mode Enclosed InfoCard - USPS: 2017 New Cohort and 2015 Panel Locating Problems (InfoCard)
Mail Start Mode First Questionnaire Cover Letter for 2015 Cooperative cases (M_Quex1_1)
Mail Start Mode First Questionnaire Cover Letter for 2015 Cooperative Retiree cases (M_Quex1_2)
Mail Start Mode First Questionnaire Cover Letter for 2015 Nonresponse Retiree cases (M_Quex1_3)
Mail Start Mode First Questionnaire Cover Letter for Refusals/NIR/Partial Responders (M_Quex1_4)
Mail Start Mode First Questionnaire Cover Letter for 2017 New Cohort cases (M_Quex1_5)
Mail Start Mode First Questionnaire Cover Letter for Money Cooperative cases (incentive) (M_Quex1_10)
Web Start Mode Follow-up Notice sent via USPS to 2015 Cooperative cases (W_FN1_u)
Web Start Mode Follow-up Notice sent via USPS to 2015 Locating/2015 Nonresponse Cases (W_FN2_u)
Web Start Mode Follow-up Notice sent via USPS to 2015 Cooperative Green Appeal (W_FN3_u)
Web Start Mode Follow-up Notice sent via USPS to New Cohort (incentive) (W_FN10_u)
CATI Start Mode Prompting Message sent via USPS to Panel (C_PM_u)
CATI Start Mode Prompting Message sent via USPS to New Cohort (incentive) (C_PM2_u)
CATI and Web Start Modes First Questionnaire Cover Letter sent to 2015 Non-retired cases (CW_Quex1_6)
CATI and Web Start Modes First Questionnaire Cover Letter sent to 2015 Cooperative Retiree cases
(CW_Quex1_7)
CATI and Web Start Modes First Questionnaire Cover Letter sent to New Cohort (CW_Quex1_10)
Mail Start Second Questionnaire Cover Letter to all sample member types (Quex2)
Mail Start Mode Second Questionnaire Cover Letter to New Cohort (incentive) (Quex2_10)
CATI and Web Start Survey Request Letter: Panel except Locating Problems (CW_InfoCard_letter_for_SMs)
CATI and Web Start Survey Request Letter: New Cohort and Panel Locating Problems
(CW_Letter_with_Graphic_not_InfoCard)
Postcard Text sent to all sample member types

Survey of Doctorate Recipients

Page E-1

SDR OMB Package for the 2017 Cycle

SDR Letterhead text (front)

Survey of Doctorate Recipients

Page E-2

SDR OMB Package for the 2017 Cycle

SDR Letterhead text (back)

Survey of Doctorate Recipients

Page E-3

SDR OMB Package for the 2017 Cycle

Mail Start Mode Prenotice Letter: 2015 Cooperative (M_PN1.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
In a few days, you should receive a questionnaire in the mail for the 2017 Survey of Doctorate Recipients
(SDR). Thank you for your past contribution to this unique study of doctorate holders, sponsored by
the National Science Foundation (NSF) and the National Institutes of Health. The participation of
doctorate holders like you has provided governmental organizations, businesses, and academic institutions
with crucial information concerning the availability of highly-educated personnel in a variety of fields.
The SDR has been conducted biennially since 1973 and is the only source of data on the careers of science,
engineering, and health doctorate holders from U.S. academic institutions. The value of the data obtained
over the years with the help of its participants is immeasurable, and your continued participation ensures
the validity and accuracy of the survey data. Aggregated results from earlier rounds of the study are
publicly available at the NSF website listed below.
All information you provide will be kept strictly confidential and safeguarded in accordance with the Privacy
Act of 1974 and the Confidential Information Protection and Statistical Efficiency Act of 2002.
NORC at the University of Chicago is the survey contractor conducting this survey on our behalf. If you do
not receive a questionnaire within two weeks or have any questions regarding this study, please contact
NORC via the toll-free number or email address listed below. A secure, online version of the survey is also
available; please contact NORC for your unique survey PIN and password if you prefer to complete the
2017 SDR online. Staff members are available from 9 a.m. to 9 p.m. (Central Time) to assist you.
I would greatly appreciate your continued participation in this significant effort.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-4

SDR OMB Package for the 2017 Cycle

Mail Start Mode Prenotice Letter: 2015 Nonresponse Retirees/2015 NIR/Partial Responders
(M_PN2.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
In a few days, you should receive a questionnaire in the mail for the 2017 Survey of Doctorate Recipients
(SDR), which is sponsored by the National Science Foundation (NSF) and the National Institutes of Health.
The SDR has been conducted biennially since 1973 and is the only source of data on the careers of
science, engineering, and health doctorate holders from U.S. academic institutions. Analysis and reports
prepared using this survey data have provided governmental organizations, businesses, and academic
institutions with crucial information concerning the availability of highly-educated personnel in a variety of
fields.
Your response is needed, regardless of your current employment status or occupation. Because you were
scientifically selected to represent a crucial segment of the population, we cannot substitute any other
person for you. Your involvement in this ongoing effort will help to ensure the validity and accuracy of the
survey data. Aggregated results from earlier rounds of the study are publicly available at the NSF website
listed below.
All information you provide will be kept strictly confidential and safeguarded in accordance with the Privacy
Act of 1974 and the Confidential Information Protection and Statistical Efficiency Act of 2002.
NORC at the University of Chicago is the survey contractor conducting this survey on our behalf. If you do
not receive a questionnaire within two weeks or have any questions regarding this study, please contact them
via the toll-free number or email address listed below. A secure, online version of the survey is also
available; please contact NORC for your unique PIN and password if you prefer to complete the 2017
SDR online. Staff members are available from 9 a.m. to 9 p.m. (Central Time) to assist you.
I would greatly appreciate your cooperation in this significant effort.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-5

SDR OMB Package for the 2017 Cycle

Mail Start Mode Prenotice Letter: New Cohort (M_PN3.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
In a few days, you should receive a questionnaire in the mail for the 2017 Survey of Doctorate Recipients
(SDR), which is sponsored by the National Science Foundation (NSF) and the National Institutes of Health.
The SDR has been conducted biennially since 1973 and is the only source of data on the careers of
science, engineering, and health doctorate holders from U.S. academic institutions. Analysis and reports
prepared using this survey data have provided governmental organizations, businesses, and academic
institutions with crucial information concerning the availability of highly-educated personnel in a variety of
fields.
You were scientifically selected to represent a crucial segment of the population of individuals earning a
research doctorate in the U.S. As such, we cannot substitute any other person for you. Your involvement
in this ongoing effort will help ensure the validity and accuracy of the survey data. Aggregated results from
earlier rounds of the study are publicly available at the NSF website listed below.
All information you provide will be kept strictly confidential and safeguarded in accordance with the Privacy
Act of 1974 and the Confidential Information Protection and Statistical Efficiency Act of 2002.
NORC at the University of Chicago is the survey contractor conducting this survey on our behalf. If you do
not receive a questionnaire within two weeks or have any questions regarding this study, please contact them
via the toll-free number or email address listed below. A secure, online version of the survey is also
available; please contact NORC for your unique survey PIN and password if you prefer to complete
the 2017 SDR online. Staff members are available from 9 a.m. to 9 p.m. (Central Time) to assist you.
I would greatly appreciate your cooperation in this significant effort.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-6

SDR OMB Package for the 2017 Cycle

CATI Start Mode Prenotice Letter: 2015 Cooperative (C_PN1.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We will be calling you soon to request your participation in the 2017 Survey of Doctorate Recipients (SDR).
Thank you for your past contribution to this unique study of doctorate holders, sponsored by the
National Science Foundation (NSF) and the National Institutes of Health. The participation of doctorate
holders like you has provided governmental organizations, businesses, and academic institutions with crucial
information concerning the availability of highly-educated personnel in a variety of fields.
The SDR has been conducted biennially since 1973 and is the only source of data on the careers of science,
engineering, and health doctorate holders from U.S. academic institutions. The value of the information
obtained over the years with the help of its participants is immeasurable, and your continued participation
ensures the validity and accuracy of the survey data. Aggregated results from earlier rounds of the study
are publicly available at the NSF website listed below.
All information you provide will be kept strictly confidential and safeguarded in accordance with the Privacy
Act of 1974 and the Confidential Information Protection and Statistical Efficiency Act of 2002.
NORC at the University of Chicago is the survey contractor conducting this survey on our behalf. The
interview should take about 25 minutes. If you do not receive a call within the next two weeks or have any
questions regarding this study, please contact NORC via the toll-free number or email address listed below.
A secure, online version of the survey is also available; please contact NORC for your unique survey PIN
and password if you prefer to complete the 2017 SDR online. Staff members are available from 9 a.m. to
9 p.m. (Central Time) to assist you.
I would greatly appreciate your continued participation in this significant effort.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-7

SDR OMB Package for the 2017 Cycle

CATI Start Mode Prenotice Letter: 2015 Nonresponse /2017 New Cohort (C_PN2.docx)
This letter will be sent to sample members printed on the
SDR 2017 letterhead as found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We will be calling you soon to request your participation in the 2017 Survey of Doctorate Recipients (SDR),
sponsored by the National Science Foundation (NSF) and the National Institutes of Health. The SDR has
been conducted biennially since 1973 and is the only source of data on the careers of science,
engineering, and health doctorate holders from U.S. academic institutions. Analysis and reports
prepared using this survey data have provided governmental organizations, businesses, and academic
institutions with crucial information concerning the availability of highly-educated personnel in a variety of
fields.
You were scientifically selected to represent a crucial segment of the population of individuals earning a
research doctorate in the U.S. As such, we cannot substitute any other person for you. Your involvement
in this ongoing effort will help ensure the validity and accuracy of the survey data. Aggregated results from
earlier rounds of the study are publicly available at the NSF website listed below.
All information you provide will be kept strictly confidential and safeguarded in accordance with the Privacy
Act of 1974 and the Confidential Information Protection and Statistical Efficiency Act of 2002.
NORC at the University of Chicago is the survey contractor conducting this survey on our behalf. The
interview should take about 25 minutes. If you do not receive a call within the next two weeks or have any
questions regarding this study, please contact NORC via the toll-free number or email address listed below.
A secure, online version of the survey is also available; please contact NORC for your unique survey PIN
and password if you prefer to complete the 2017 SDR online. Staff members are available from 9 a.m. to 9
p.m. (Central Time) to assist you.
We would greatly appreciate your cooperation in this important effort.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-8

SDR OMB Package for the 2017 Cycle

Web Start Mode Initial Notice - USPS: 2015 Cooperative (W_IC1_u.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
Thank you for your past participation in the Survey of Doctorate Recipients (SDR), sponsored by the
National Science Foundation (NSF) and the National Institutes of Health. Your responses, aggregated with
other participants’ responses, help academic and governmental organizations with decision-making, for
instance, in the anticipation of personnel shortages and funding needs for research and development. The
information you provided has also proven valuable for students who want to learn about the occupational
potential of their graduate education.
Please continue to participate by completing the 2017 SDR online. Because the survey is secure and
confidential, you’ll need to enter your unique and case-sensitive PIN and password to access the survey after
going to the survey website.
https://websurvey.norc.org/2017sdr
PIN: [WEBPIN]

Password: [WEBPWD]

To ensure receipt, we are also sending you access to the survey to the email address we have on record
for you. The information you provide will be collected by NORC at the University of Chicago, the survey
contractor conducting SDR on our behalf. Your data will be kept strictly confidential and safeguarded in
accordance with the Privacy Act of 1974 and the Confidential Information Protection and Statistical
Efficiency Act of 2002.
If you have any questions regarding the survey or would like to request a paper version of the survey, please
contact NORC via the toll-free number or email address listed below. Staff members are available from 9
a.m. to 9 p.m. (Central Time) to assist you.
Thank you in advance for contributing to the SDR. We look forward to receiving your online survey.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-9

SDR OMB Package for the 2017 Cycle

Web Start Mode Initial Notice - USPS: 2015 Cooperative Retiree (W_IC2_u.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
Thank you for your past participation in the Survey of Doctorate Recipients (SDR), sponsored by the
National Science Foundation (NSF) and the National Institutes of Health. Your responses, aggregated with
other participants’ responses, help academic and governmental organizations with decision-making, for
instance, in the anticipation of personnel shortages and funding needs for research and development. The
information you provided has also proven valuable for students who want to learn about the occupational
potential of their graduate education.
Please continue to participate by completing the 2017 SDR online. Because the survey is secure and
confidential, you’ll need to enter your unique and case-sensitive PIN and password to access the survey after
going to the survey website.
https://websurvey.norc.org/2017sdr
PIN: [WEBPIN]

Password: [WEBPWD]

To ensure receipt, we are also sending you access to the survey to the email address we have on record
for you. The information you provide will be collected by NORC at the University of Chicago, the survey
contractor conducting SDR on our behalf. Your data will be kept strictly confidential and safeguarded in
accordance with the Privacy Act of 1974 and the Confidential Information Protection and Statistical
Efficiency Act of 2002.
Our records show that you were retired when you last participated in the SDR. Your response is vital,
regardless of your employment situation. We want to hear from you whether you are retired and not working,
retired but working part time, or back to working full time.
If you have any questions regarding the survey, please contact NORC via the toll-free number or email
address listed below. Staff members are available from 9 a.m. to 9 p.m. (Central Time) to assist you.
Thank you in advance for contributing to the SDR. We look forward to receiving your online survey.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation
Survey of Doctorate Recipients

ID: [SDR ID]

Page E-10

SDR OMB Package for the 2017 Cycle

Web Start Mode Initial Notice - USPS: 2015 Nonresponse (W_IC3_u.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We are requesting your participation in the 2017 Survey of Doctorate Recipients (SDR). The study is
sponsored by the National Science Foundation (NSF) and the National Institutes of Health and is the only
source of data on the careers of science, engineering, and health doctorate holders from U.S. academic
institutions.
Your response is needed whatever your current employment status or occupation. Because you were
scientifically selected to represent a crucial segment of the population, we cannot substitute any other
person for you. Your involvement in this effort will help ensure the validity and accuracy of the survey
results.
Please contribute to this research by completing the SDR online. Because the survey is secure and
confidential, you’ll need to enter your unique and case-sensitive PIN and password to access the survey after
going to the survey website.
https://websurvey.norc.org/2017sdr
PIN: [WEBPIN]

Password: [WEBPWD]

To ensure receipt, we are also sending you access to the survey to the email address we have on record
for you. The information you provide will be collected by NORC at the University of Chicago, the survey
contractor conducting SDR on our behalf. Your data will be kept strictly confidential in accordance with the
Privacy Act of 1974 and the Confidential Information Protection and Statistical Efficiency Act of 2002.
If you have any questions regarding the survey or would like to request a paper version of the survey, please
contact the survey contractor, NORC at the University of Chicago, via the toll-free number or email address
listed below. Staff members are available from 9 a.m. to 9 p.m. (Central Time) to assist you.
Thank you in advance for contributing to the SDR. We look forward to receiving your online survey.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-11

SDR OMB Package for the 2017 Cycle

Web Start Mode Initial Notice - USPS: 2015 Cooperative Green Appeal (W_IC5_u.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
Thank you for your past participation in the Survey of Doctorate Recipients (SDR), sponsored by the
National Science Foundation (NSF) and the National Institutes of Health. Your responses, aggregated with
other participants’ responses, help academic and governmental organizations with decision-making, for
instance, in the anticipation of personnel shortages and funding needs for research and development. The
information you provided has also proven valuable for students who want to learn about the occupational
potential of their graduate education.
We are asking you to complete the 2017 SDR online rather than a paper questionnaire sent in the mail to
promote a more efficient and eco-friendly way to participate in the 2017 SDR. We hope you will support the
NSF's efforts to conserve resources.
Please continue to participate by completing the survey online. Because the survey is secure and
confidential, you’ll need to enter your unique and case-sensitive PIN and password to access the survey after
going to the survey website.
https://websurvey.norc.org/2017sdr
PIN: [WEBPIN]

Password: [WEBPWD]

To ensure receipt, we are also sending you access to the survey to the email address we have on record
for you. The information you provide will be collected by NORC at the University of Chicago, the survey
contractor conducting SDR on our behalf. Your data will be kept strictly confidential and safeguarded in
accordance with the Privacy Act of 1974 and the Confidential Information Protection and Statistical
Efficiency Act of 2002.
If you have any questions regarding the survey or would rather complete the paper version of the survey,
please contact NORC via the toll-free number or email address listed below. Staff members are available
from 9 a.m. to 9 p.m. (Central Time) to assist you.
Thank you in advance for contributing to the SDR. We look forward to receiving your online survey.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation
Survey of Doctorate Recipients

ID: [SDR ID]

Page E-12

SDR OMB Package for the 2017 Cycle

Web Start Mode Initial Notice - USPS: 2017 New Cohort and 2015 Panel Locating Problems
(W_IC7_u.docx)
Please see pages D-14 and D-15 for the referenced enclosed InfoCard
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.

[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We are requesting your participation in the 2017 Survey of Doctorate Recipients (SDR), sponsored by
the National Science Foundation (NSF) and the National Institutes of Health. The SDR collects career
outcome information from a highly educated and important population – individuals who have earned a
science, engineering, or health doctorate degree from a U.S. academic institution.
Please complete this important survey online using the access information below. Your PIN and
password are unique, and, for your convenience, we are sending this same information to the email address
we have on record for you.
https://websurvey.norc.org/2017sdr
PIN:

[WEBPIN]

Password:

[WEBPWD]

Your response will help us better understand the career and employment patterns of the U.S. trained
doctorate population. Your response is needed regardless of your present employment situation, whether
you are working, retired, or looking for a job. We also understand that highly-trained individuals like
yourself are mobile, and we want to hear from you, wherever you currently reside.
To thank you for your consideration, you’ll find a card enclosed that provides more information about the
SDR with an attached bookmark. If you have any questions regarding the survey, please contact NORC via
the toll-free number or email address listed below. Staff members are available from 9 a.m. to 9 p.m. (Central
Time) to assist you.
We look forward to receiving your online survey.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

SDR ID: [SDR ID]

Page E-13

SDR OMB Package for the 2017 Cycle

Web Start Mode Enclosed InfoCard - USPS: 2017 New Cohort and 2015 Panel Locating
Problems (InfoCard) OUTSIDE VIEW
Graphic and statistics, as well as color of the InfoCard will be updated utilizing the 2015 SDR
data.

Survey of Doctorate Recipients

Page E-14

SDR OMB Package for the 2017 Cycle

Web Start Mode Enclosed InfoCard - USPS: 2017 New Cohort and 2015 Panel Locating
Problems (InfoCard) INSIDE VIEW
Graphic and statistics, as well as color of the InfoCard will be updated utilizing the 2015 SDR
data.

Survey of Doctorate Recipients

Page E-15

SDR OMB Package for the 2017 Cycle

Mail Start Mode First Questionnaire Cover Letter: 2015 Cooperative (M_Quex1_1.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
Thank you for your past participation in the Survey of Doctorate Recipients (SDR). For over 40 years,
you and other SDR respondents have contributed to an irreplaceable collection of longitudinal data about
doctorate holders. Your responses, aggregated with other participants’ responses, help academic and
governmental organizations with decision-making, for instance, in the anticipation of personnel shortages
and funding needs for research and development. The information you provided has also proven valuable for
students who want to learn about the occupational potential of their graduate education.
At this time, we are asking for you to complete the enclosed 2017 SDR questionnaire. The study is
sponsored by the National Science Foundation (NSF) and the National Institutes of Health. Your response,
regardless of your employment situation, is vital to the creation of an accurate picture of the doctorate
population. We want to learn how many doctorate holders are working in or out of their field of study, are
seeking employment, are retired, or are in other employment-related situations. We can only learn about how
the career patterns of doctorate recipients change over time from you. Aggregated results from earlier rounds
of the study are publicly available at the NSF website listed below.
Please complete the enclosed questionnaire and return it in the postage-paid envelope to NORC.
If you have any questions about the survey, please contact the survey contractor, NORC at the University of
Chicago, via the toll-free number or email address listed below. Staff members are available from 9 a.m. to 9
p.m. (Central Time) to assist you.
Thank you for your continued participation. We look forward to receiving your completed survey.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-16

SDR OMB Package for the 2017 Cycle

Mail Start Mode First Questionnaire Cover Letter: 2015 Cooperative Retiree (M_Quex1_2.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.

[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
Thank you for your past participation in the Survey of Doctorate Recipients (SDR). For over 40 years,
you and other SDR respondents have contributed to an irreplaceable collection of longitudinal data about
doctorate holders. Your responses, aggregated with other participants’ responses, help academic and
governmental organizations with decision-making, for instance, in the anticipation of personnel shortages
and funding needs for research and development. The information you provided has also proven valuable for
students who want to learn about the occupational potential of their graduate education.
At this time, we are asking for you to complete the enclosed 2017 SDR questionnaire. The study is
sponsored by the National Science Foundation (NSF) and the National Institutes of Health. Our records show
that you were retired when you last participated in the SDR. Your response, regardless of your employment
situation, is vital to the creation of an accurate picture of the doctorate population. We can only learn about
how the career patterns of doctorate recipients change over time from you. Aggregated results from earlier
rounds of the study are publicly available at the NSF website listed below.
Please complete the enclosed questionnaire and return it in the postage-paid envelope to NORC.
If you have any questions about the survey, please contact the survey contractor, NORC at the University of
Chicago, via the toll-free number or email address listed below. Staff members are available from 9 a.m. to 9
p.m. (Central Time) to assist you.
Thank you for your continued participation. We look forward to receiving your completed survey.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-17

SDR OMB Package for the 2017 Cycle

Mail Start Mode First Questionnaire Cover Letter: 2015 Nonresponse Retiree
(M_Quex1_3.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We are requesting your participation in the 2017 Survey of Doctorate Recipients (SDR). The study is
sponsored by the National Science Foundation (NSF) and the National Institutes of Health and an
irreplaceable collection of longitudinal data on the careers of science, engineering, and health doctorate
holders from U.S. academic institutions.
Our records show that you were retired during a previous round of the SDR. Your response, regardless of
your current employment situation, is vital to the creation of an accurate picture of the doctorate population.
We can only learn about the extent to which retirement decisions change or remain stable over time from
you. Aggregated results from earlier rounds of the study are publicly available at the NSF website listed
below.
Please complete the enclosed questionnaire and return it in the postage-paid envelope to NORC at the
University of Chicago.
If you have any questions about the survey, please contact the survey contractor, NORC at the University of
Chicago, via the toll-free number or email address listed below. Staff members are available from 9 a.m. to 9
p.m. (Central Time) to assist you.
Thank you in advance for your participation. We look forward to receiving your completed survey.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-18

SDR OMB Package for the 2017 Cycle

Mail Start Mode First Questionnaire Cover Letter: Refusals/NIR/Partial Responders
(M_Quex1_4.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We are requesting your participation in the 2017 Survey of Doctorate Recipients (SDR). The study is
sponsored by the National Science Foundation (NSF) and the National Institutes of Health and is an
irreplaceable collection of longitudinal data on the careers of science, engineering, and health doctorate
holders from U.S. academic institutions.
Your response, regardless of your employment situation, is vital to the creation of an accurate picture of the
doctorate population. We want to learn how many doctorate holders are working in or out of their field of
study, are seeking employment, are retired, or are in other employment-related situations. We can only learn
about how the career patterns of doctorate recipients change over time from you.
Your data will be kept strictly confidential and safeguarded in accordance with the Privacy Act of 1974
and the Confidential Information Protection and Statistical Efficiency Act of 2002. The data collected will be
aggregated and used to prepare scientific reports, articles, and statistical summaries, but any information
released publicly will maintain the confidentiality of all participants. Results from earlier studies are
available at the NSF website listed below.
Please complete the enclosed questionnaire and return it in the postage-paid envelope to NORC.
If you have any questions about the survey, please contact the survey contractor, NORC at the University of
Chicago, via the toll-free number or email address listed below. Staff members are available from 9 a.m. to 9
p.m. (Central Time) to assist you.
Thank you in advance for your participation. We look forward to receiving your completed survey.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-19

SDR OMB Package for the 2017 Cycle

Mail Start Mode First Questionnaire Cover Letter: 2017 New Cohort (M_Quex1_5.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We are requesting your participation in the 2017 Survey of Doctorate Recipients (SDR). The study is
sponsored by the National Science Foundation (NSF) and the National Institutes of Health.
As mentioned in our previous correspondence, you were scientifically selected to represent a crucial segment
of the population and as such, we cannot substitute any other person for you. Your response, regardless of
your employment situation, is vital to the creation of an accurate picture of the doctorate population.
Please complete the enclosed questionnaire and return it in the postage-paid envelope to NORC at the
University of Chicago—or complete an online survey by using the access information below. Because the
survey is secure and confidential, you’ll need to enter your unique and case-sensitive PIN and password to
access the survey after going to the survey website.
https://websurvey.norc.org/2017sdr
PIN: [WEBPIN]

Password: [WEBPWD]

If you have any questions about the survey, please contact the survey contractor, NORC at the University of
Chicago, via the toll-free number or email address listed below. Staff members are available from 9 a.m. to 9
p.m. (Central Time) to assist you.
We would greatly appreciate your cooperation in this significant effort. We look forward to receiving
your completed survey.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-20

SDR OMB Package for the 2017 Cycle

Mail Start Mode First Questionnaire Cover Letter: Money Cooperative (M_Quex1_10.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We are requesting your participation in the 2017 Survey of Doctorate Recipients (SDR). The study is
sponsored by the National Science Foundation (NSF) and the National Institutes of Health and is an
irreplaceable collection of longitudinal data on the careers of science, engineering, and health doctorate
holders from U.S. academic institutions.
Please find a $30 token of our appreciation enclosed to thank you for your past contribution and in
advance for your 2017 SDR participation.Your response, regardless of your employment situation, is vital
to the creation of an accurate picture of the doctorate population. We want to learn how many doctorate
holders are working in or out of their field of study, are seeking employment, are retired, or are in other
employment-related situations. We can only learn about how the career patterns of doctorate recipients
change over time from you.
Your data will be kept strictly confidential and safeguarded in accordance with the Privacy Act of 1974
and the Confidential Information Protection and Statistical Efficiency Act of 2002. The data collected will be
aggregated and used to prepare scientific reports, articles, and statistical summaries, but any information
released publicly will maintain the confidentiality of all participants. Results from earlier studies are
available at the NSF website listed below.
Please complete the enclosed questionnaire and return it in the postage-paid envelope to NORC.
If you have any questions about the survey, please contact the survey contractor, NORC at the University of
Chicago, via the toll-free number or email address listed below. Staff members are available from 9 a.m. to 9
p.m. (Central Time) to assist you.
Thank you in advance for your participation. We look forward to receiving your completed survey.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-21

SDR OMB Package for the 2017 Cycle

Web Start Mode Follow-up Notice - USPS: 2015 Cooperative (W_FN1_u.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We would like your help in completing the 2017 Survey of Doctorate Recipients (SDR). The SDR is
sponsored by the National Science Foundation (NSF) and the National Institutes of Health.
Your participation helps to ensure that SDR information and statistics are valid and complete. Because
you were scientifically selected for the SDR, we cannot substitute any other person for you. Your response,
regardless of your employment situation, is vital to the creation of an accurate picture of the doctorate
population. We want to learn how many doctorate holders are working in or out of their field of study, are
seeking employment, are retired, or are in other employment-related situations.
Please complete this important survey online using the access information below.
https://websurvey.norc.org/2017sdr
PIN: [WEBPIN]

Password: [WEBPWD]

If you would rather complete this survey on the telephone or via a paper questionnaire, please let us know by
contacting the study’s toll-free number or email address listed below. For your convenience, this information
has also been sent to the email address we have on record for you.
If you have any questions regarding the survey, please contact the survey contractor, NORC at the University
of Chicago, where staff members are available from 9 a.m. to 9 p.m. (Central Time) to assist you.
Thank you in advance for your cooperation in this important effort.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation
P.S. If you have already completed the 2017 survey, thank you so much for your time!
ID: [SDR ID]

Survey of Doctorate Recipients

Page E-22

SDR OMB Package for the 2017 Cycle

Web Start Mode Follow-up Notice - USPS: 2015 Locating/2015 Nonresponse (W_FN2_u.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We would like your help in completing the 2017 Survey of Doctorate Recipients (SDR). The SDR is
sponsored by the National Science Foundation (NSF) and the National Institutes of Health.
Your participation helps to ensure that SDR information and statistics are valid and complete. Science,
engineering, and health doctorate holders are highly trained and mobile, moving from one country to another
to take advantage of opportunities in their fields. We want to gather information on the productivity and
career paths of individuals like you, wherever you may live or work. Because you were scientifically selected
for the SDR, we cannot substitute any other person for you.
Please complete this important survey online using the access information below.
https://websurvey.norc.org/2017sdr
PIN: [WEBPIN]

Password: [WEBPWD]

If you would rather complete this survey on the telephone or via a paper questionnaire, please let us know by
contacting the study’s toll-free number or email address listed below. For your convenience, this information
has also been sent to the email address we have on record for you.
If you have any questions regarding the survey, please contact the survey contractor, NORC at the University
of Chicago, where staff members are available from 9 a.m. to 9 p.m. (Central Time) to assist you.
Thank you in advance for your cooperation in this important effort.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation
P.S. If you have already completed the 2017 survey, thank you so much for your time!
ID: [SDR ID]

Survey of Doctorate Recipients

Page E-23

SDR OMB Package for the 2017 Cycle

Web Start Mode Follow-up Notice - USPS: 2015 Cooperative Green Appeal (W_FN3_u.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We would like your help in completing the 2017 Survey of Doctorate Recipients (SDR). The SDR is
sponsored by the National Science Foundation (NSF) and the National Institutes of Health.
Your participation helps ensure that SDR information and statistics are valid and complete. Because
you were scientifically selected for the SDR, we cannot substitute any other person for you. Your response,
regardless of your employment situation or location, is vital to the creation of an accurate picture of the
doctorate population. We want to learn how many doctorate holders are working in or out of their field of
study, are seeking employment, are retired, or are in other employment-related situations.
Please complete this important survey online using the access information below.
https://websurvey.norc.org/2017sdr
PIN: [WEBPIN]

Password: [WEBPWD]

We are asking you to complete the survey online in an effort to make the SDR a more sustainable and
environmentally-friendly program. If you would rather complete this survey on the telephone or via a
paper questionnaire, please let us know by contacting the study’s toll-free number or email address listed
below. For your convenience, this information has also been sent to the email address we have on record for
you.
If you have any questions regarding the survey, please contact the survey contractor, NORC at the University
of Chicago, where staff members are available from 9 a.m. to 9 p.m. (Central Time) to assist you.
Thank you in advance for your cooperation in this important effort.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation
P.S. If you have already completed the 2017 survey, thank you so much for your time!
ID: [SDR ID]

Survey of Doctorate Recipients

Page E-24

SDR OMB Package for the 2017 Cycle

Web Start Mode Follow-up notice - USPS: New Cohort (W_FN4_u.docx)

This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
Please find a $30 token of our appreciation enclosed to thank you in advance for participating in the
2017 Survey of Doctorate Recipients (SDR).
Your participation helps to ensure that SDR information and statistics are valid and complete. Because
you were scientifically selected for the SDR, we cannot substitute any other person for you. Doctorate holders
in science, engineering, and health are highly trained and mobile, moving to take advantage of opportunities in
their fields. Understanding the productivity and career paths of individuals like you is important no matter
where they live and work.
Please complete this important survey online using the access information below.
https://websurvey.norc.org/2017sdr
PIN: [WEBPIN]

Password: [WEBPWD]

If you would rather complete this survey on the telephone or via a paper questionnaire, please let us know by
contacting the study’s toll-free number or email address listed below. For your convenience, this information
has also been sent to the email address we have on record for you.
The 2017 SDR is sponsored by the National Science Foundation (NSF) and the National Institutes of Health.
The information you provide will be collected by NORC at the University of Chicago, the survey contractor
conducting the SDR on our behalf. If you have any questions regarding the survey, please contact NORC,
where staff members are available from 9 a.m. to 9 p.m. (Central Time) to assist you.
Thank you in advance for your cooperation in this important effort.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation
P.S. If you have already completed the 2017 survey, thank you so much for your time!
SDR ID: [SDR ID]

Survey of Doctorate Recipients

Page E-25

SDR OMB Package for the 2017 Cycle

CATI Start Mode Prompting Message - USPS: Panel (C_PM_u.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We recently attempted to reach you by telephone to complete the 2017 Survey of Doctorate Recipients
(SDR). The SDR is sponsored by the National Science Foundation (NSF) and the National Institutes of
Health. We regret that we have been unable to reach you.
Your participation helps to ensure that SDR information and statistics are valid and complete. Because
you were scientifically selected for the SDR, we cannot substitute any other person for you. Your response,
regardless of your employment situation, is vital to the creation of an accurate picture of the doctorate
population. We want to learn how many doctorate holders are working in or out of their field of study, are
seeking employment, are retired, or are in other employment-related situations.
The data you provide will be kept strictly confidential and safeguarded in accordance with the Privacy Act of
1974 and the Confidential Information Protection and Statistical Efficiency Act of 2002.
We are offering three different ways by which you may participate.
1. Paper:
2. Telephone:
3. Online:

To request a paper version of the survey, please call 1-800-685-1663.
If you prefer a telephone interview, please call 1-800-685-1663.
If you prefer to complete the survey online, please go to the following URL and enter
your unique and case-sensitive PIN and password to securely access the survey:
https://survey.norc.org/2017sdr
PIN: [WEBPIN]

Password:

[WEBPWD]

For your convenience, we are also sending you access to the online survey to the email address we have on
record for you. If you have any questions regarding the survey, please contact the survey contractor, NORC
at the University of Chicago, via the toll-free number or email address listed below. Staff members are
available from 9 a.m. to 9 p.m. (Central Time) to assist you.
Thank you in advance for your cooperation in this important effort.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]-113

Page E-26

SDR OMB Package for the 2017 Cycle

CATI Start Mode Prompting Message - USPS: New Cohort (C_PM2_u.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
Please find a $30 token of our appreciation enclosed to thank you in advance for participating in the
2017 Survey of Doctorate Recipients. We recently attempted to reach you by telephone to complete the
survey and regret we have been unable to reach you. The SDR is sponsored by the National Science
Foundation (NSF) and the National Institutes of Health.
Your participation helps to ensure that SDR information and statistics are valid and complete. Because
you were scientifically selected for the SDR, we cannot substitute any other person for you. Your response,
regardless of your employment situation, is vital to the creation of an accurate picture of the doctorate
population. We want to learn how many doctorate holders are working in or out of their field of study, are
seeking employment, are retired, or are in other employment-related situations.
The data you provide will be kept strictly confidential and safeguarded in accordance with the Privacy Act of
1974 and the Confidential Information Protection and Statistical Efficiency Act of 2002.
We are offering three different ways by which you may participate.
1. Paper:
2. Telephone:
3. Online:

To request a paper version of the survey, please call 1-800-685-1663.
If you prefer a telephone interview, please call 1-800-685-1663.
If you prefer to complete the survey online, please go to the following URL and enter
your unique and case-sensitive PIN and password to securely access the survey:
https://survey.norc.org/2017sdr
PIN: [WEBPIN]

Password:

[WEBPWD]

For your convenience, we are also sending you access to the online survey to the email address we have on
record for you. If you have any questions regarding the survey, please contact the survey contractor, NORC
at the University of Chicago, via the toll-free number or email address listed below. Staff members are
available from 9 a.m. to 9 p.m. (Central Time) to assist you.
Thank you in advance for your cooperation in this important effort.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

ID: [SDR ID]

Page E-27

SDR OMB Package for the 2017 Cycle

CATI and Web Start Mode First Questionnaire Cover Letter: 2015 Non-retirees
(CW_Quex1_6.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],

We are requesting your participation in the 2017 Survey of Doctorate Recipients, sponsored by the
National Science Foundation (NSF) and the National Institutes of Health. Recently, we tried to contact you
by email and telephone but have not received your completed questionnaire.
Whether you are working in or out of your doctoral field of study, seeking employment, retired, or in
another employment-related situation, your response is vital. We can only learn about how the career
patterns of doctorate recipients change over time from you. This study is conducted every two years and
takes an average of 25 minutes of your time. Results from earlier studies are available at the NSF website
below.
For your convenience, we are offering three different ways by which you can participate.
1. Paper:
2. Telephone:
3. Online:

Please fill out the enclosed questionnaire and return it using the enclosed envelope.
If you prefer a telephone interview, please call 1-800-685-1663.
If you prefer to complete this survey online, please go to the following URL and enter your
unique and case-sensitive PIN and password to securely access the survey:
https://survey.norc.org/2017sdr
PIN: [WEBPIN]
Password: [WEBPWD]

If you have any questions regarding the survey, please contact the survey contractor, NORC at the University
of Chicago, via the toll-free number or email address listed below. Staff members are available from 9 a.m.
to 9 p.m. (Central Time) to assist you.
Thank you for your help with this important effort. We look forward to your response.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation
ID: [SU_ID]

Survey of Doctorate Recipients

Page E-28

SDR OMB Package for the 2017 Cycle

CATI and Web Start Mode First Questionnaire Cover Letter: 2015 Cooperative Retiree
(CW_Quex1_7.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],

We are requesting your participation in the 2017 Survey of Doctorate Recipients, sponsored by the
National Science Foundation (NSF) and the National Institutes of Health. Recently, we tried to contact you
by email and telephone but have not received your completed questionnaire.
Whether you are retired and not working, retired but working part time, or back to working full time,
your response is vital. We can only learn about the extent to which retirement decisions change or remain
stable over time from you. Aggregated results from earlier rounds of the study are publicly available at the
NSF website listed below.
For your convenience, we are offering three different ways by which you can participate.
1. Paper:
2. Telephone:
3. Online:

Please fill out the enclosed questionnaire and return it using the enclosed envelope.
If you prefer a telephone interview, please call 1-800-685-1663.
If you prefer to complete this survey online, please go to the following URL and enter your
unique and case-sensitive PIN and password to securely access the survey:
https://survey.norc.org/2017sdr
PIN: [WEBPIN]
Password: [WEBPWD]

If you have any questions regarding the survey, please contact the survey contractor, NORC at the University
of Chicago, via the toll-free number or email address listed below. Staff members are available from 9 a.m.
to 9 p.m. (Central Time) to assist you.
Thank you for your help with this important effort. We look forward to your response.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation
ID: [SU_ID]

Survey of Doctorate Recipients

Page E-29

SDR OMB Package for the 2017 Cycle

CATI and Web Start Mode First Questionnaire Cover Letter: New Cohort (CW_Quex1_10.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
I hope you received the $30 token of our appreciation recently mailed to you for the 2017 Survey of
Doctorate Recipients (SDR). We are requesting your participation in this important study, which is
sponsored by the National Science Foundation (NSF) and the National Institutes of Health. We have not yet
received your completed questionnaire and hope you will contribute to 2017 SDR.
Whether you are working in or out of your doctoral field of study, seeking employment, retired, or in
another employment-related situation, your response is vital. We can only learn about how the career
patterns of doctorate recipients change over time from you. This study is conducted every two years and
takes an average of 25 minutes of your time. Results from earlier studies are available at the NSF website
below.
For your convenience, we are offering three different ways by which you can participate.
1. Paper:
Please fill out the enclosed questionnaire and return it using the enclosed envelope.
2. Telephone: If you prefer a telephone interview, please call 1-800-685-1663.
3. Online:
If you prefer to complete this survey online, please go to the following URL and enter
your unique and case-sensitive PIN and password to securely access the survey:
https://survey.norc.org/2017sdr
PIN: [WEBPIN]
Password: [WEBPWD]
If you have any questions regarding the survey, please contact the survey contractor, NORC at the University
of Chicago, via the toll-free number or email address listed below. Staff members are available from 9 a.m.
to 9 p.m. (Central Time) to assist you.
Thank you for your help with this important effort. We look forward to your response.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation
ID: [SU_ID]

Survey of Doctorate Recipients

Page E-30

SDR OMB Package for the 2017 Cycle

Mail Start Second Questionnaire Cover Letter: All (Quex2.docx)
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We recently mailed you a 2017 Survey of Doctorate Recipients (SDR) questionnaire and have not yet
received your completed survey. Your participation greatly influences the accuracy, usefulness, and overall
success of the survey and its results and will take about 25 minutes of your time.
Your response is needed wherever you live and whatever your employment status is. Please complete
this important survey. For your convenience, we are offering three different participation options.
1. Paper:
2. Telephone:
3. Online:

Please fill out the enclosed questionnaire and return it using the enclosed envelope.
If you prefer a telephone interview, please call 1-800-685-1663.
If you prefer to complete the survey online, please go to the following URL and enter
your unique and case-sensitive PIN and password to securely access the survey:
https://survey.norc.org/2017sdr
PIN: MPMUK
Password: [WEBPWD]

If you have any questions regarding the survey, please contact the survey contractor, NORC at the University
of Chicago, via the toll-free number or email address listed below. Staff members are available from 9 a.m.
to 9 p.m. (Central Time) to assist you.
Thank you in advance for your cooperation in this important effort, which is sponsored by the
National Science Foundation (NSF) and the National Institutes of Health.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation
ID: [SU_ID]

Survey of Doctorate Recipients

Page E-31

SDR OMB Package for the 2017 Cycle

Mail Start Second Questionnaire Cover Letter: New Cohort (Quex2_10.docx)

This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
We recently mailed you a 2017 Survey of Doctorate Recipients (SDR) questionnaire and have not yet
received your completed survey. Your participation greatly influences the accuracy, usefulness, and overall
success of the survey and its results and will take about 25 minutes of your time.
Please find a $30 token of appreciation enclosed to thank you in advance for participating in the 2017
SDR. Your response is needed wherever you live and whatever your employment status is.
Please complete this important survey. For your convenience, we are offering three different participation
options.
1. Paper:
2. Telephone:
3. Online:

Please fill out the enclosed questionnaire and return it using the enclosed envelope.
If you prefer a telephone interview, please call 1-800-685-1663.
If you prefer to complete the survey online, please go to the following URL and enter
your unique and case-sensitive PIN and password to securely access the survey:
https://survey.norc.org/2017sdr
PIN: MPM2IR
Password: [WEBPWD]

If you have any questions regarding the survey, please contact the survey contractor, NORC at the University
of Chicago, via the toll-free number or email address listed below. Staff members are available from 9 a.m.
to 9 p.m. (Central Time) to assist you.
Thank you in advance for your cooperation in this important effort, which is sponsored by the
National Science Foundation (NSF) and the National Institutes of Health.
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation
ID: [SU_ID]

Survey of Doctorate Recipients

Page E-32

SDR OMB Package for the 2017 Cycle

CATI and Web Start Survey Request Letter: Panel except Locating Problems
(CW_InfoCard_letter_for_SMs.docx)
Please see pages D-14 and D-15 for the referenced enclosed InfoCard
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]

Dear Dr. [LAST NAME],
We have not yet received your completed 2017 Survey of Doctorate Recipients (SDR) and are
requesting your participation. Your participation will enhance the accuracy, usefulness, and overall
success of the survey and will help us better understand the career and employment patterns of doctorate
holders like you.
Your response is important wherever you live worldwide and whatever your employment status. Enclosed
you’ll find a card that provides additional information about the SDR and contains an attached bookmark.
To complete the 2017 SDR online, go to the following URL and enter your unique PIN and password:
https://survey.norc.org/2017sdr
PIN:

[WEBPIN]

Password:

[WEBPWD]

If you prefer a telephone interview or a mailed paper questionnaire, contact NORC via email or telephone
using the information listed below. Staff members can assist you from 9 a.m. to 9 p.m. (U.S. Central Time).
Sincerely,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation
SDR ID: [SDR ID]

Survey of Doctorate Recipients

Page E-33

SDR OMB Package for the 2017 Cycle

CATI and Web Start Survey Request Letter: New Cohort and Panel Locating Problems
(CW_Letter_with_Graphic_not_InfoCard.docx)
Graphic and statistics will be updated utilizing the 2015 SDR data.
This letter will be sent to sample members printed on the SDR 2017 letterhead,
images of which can be found on pages D-2 and D-3.
[DATE], 2017
Dr. [FIRST NAME] [MIDDLE INITIAL] [LAST NAME]
[ORGNAME]
[STREET ADDRESS 1]
[STREET ADDRESS 2]
[CITY], [STATE] [ZIP CODE]
Dear Dr. [LAST NAME],
Please participate in the 2017 Survey of Doctorate Recipients (SDR). Your participation helps us to
accurately continue to report on the career outcomes of U.S.-trained doctorate recipients.
To complete the survey online, please go to the following secure URL and enter your PIN and password:
https://survey.norc.org/2017sdr
PIN: [WEBPIN]

Password: [WEBPWD]

Results from the 2013 SDR show that 86% of U.S.-trained doctorates living in the U.S. were employed full
or part time; 58% worked in science occupations, 13% in engineering; and 46% worked at educational
institutions. The figure below shows 66% of employed science, engineering, and health doctorates had jobs
closely related to their degree field.

Thank you in advance for your participation in the 2017 SDR,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

SDR ID: [SDR ID]

Page E-34

SDR OMB Package for the 2017 Cycle

Postcard Text: All (Thank_you_Postcard_text.docx)
Last week you should have received a survey in the mail from NORC at the University of Chicago,
who has been contracted to conduct a survey sponsored by the National Science Foundation (NSF)
and the National Institutes of Health. Many thanks if you have already completed the survey, we
look forward to receiving it shortly. If you have not completed the survey, we hope you will
complete it as soon as you can.
If you did not receive the materials, or if you would like to complete the survey online or over the
telephone, please contact NORC via email at [email protected] or toll-free at
1-800-685-1663 between 9 a.m. and 9 p.m. (U.S. Central Time).
Your participation is important for the success of this study.
With appreciation,

John R. Gawalt, Director
National Center for Science and Engineering Statistics
National Science Foundation

Survey of Doctorate Recipients

Page E-35

SDR OMB Package for the 2017 Cycle

Attachment F
2017 SDR Sample Allocation
and Selection Table
At the summary level and by strata defined by fine field of degree

Survey of Doctorate Recipients

Page F-1

SDR OMB Package for the 2017 Cycle

2017 Survey of Doctorate Recipients Sample Allocation and Selection Table
Stratum
2017

Fine Field of Doctorate Degree

000
003
005
010
012
014
019
020
025
030
039
043
044
046
049
050
055
066
068
070
072
074
079
080
081
098
099
100
102
103
104
105
107
110
115
120
125

Agricultural Economics
Natural Resource/Environmental Economics
Agricultural Animal Breeding
Animal Nutrition
Dairy Science
Animal Science, Poultry (or Avian)
Animal Science, Other
Agronomy & Crop Science
Agricultural & Horticultural Plant Breeding (2010 & 2011)
Plant Pathology/Phytopathology
Plant Sciences, Other
Food Science
Food Science & Technology, Other
Soil Chemistry/Microbiology
Soil Sciences, Other
Horticulture Science
Fishing & Fisheries Sciences/Management
Forest Sciences & Biology
Forest Engineering
Forest/Resources Management
Wood Science & Pulp/Paper Technology
Natural Resources/Conservation
Forestry & Related Science, Other
Wildlife/Range Management
Environmental Science
Agriculture Sciences/Natural Resources, General
Agriculture Sciences/Natural Resources, Other
Biochemistry
Bioinformatics
Biomedical Sciences
Computational Biology
Biophysics
Biotechnology
Bacteriology
Plant Genetics
Plant Pathology/Phytopathology
Plant Physiology

Survey of Doctorate Recipients

Panel
Cohort
Count
720
51
379
372
197
380
377
340
375
354
384
385
384
382
381
358
378
383
40
384
377
380
380
380
372
230
342
1,867
411
506
341
515
395
406
437
397
447

Panel
Cohort
Population
5,469
51
528
2,638
197
389
2,377
4,232
1,695
2,939
1,001
1,154
2,689
638
1,539
2,366
1,279
635
40
679
426
1,084
1,116
1,226
3,446
230
1,523
29,993
964
4,372
341
5,439
444
465
1,124
876
2,006

New
Cohort
Count
198
67
0
90
0
56
192
161
82
131
172
218
86
37
99
101
112
59
0
64
19
157
44
61
396
46
35
1,568
358
853
224
367
83
47
105
49
35

Total
Frame
Count
918
118
379
462
197
436
569
501
457
485
556
603
470
419
480
459
490
442
40
448
396
537
424
441
768
276
377
3,435
769
1,359
565
882
478
453
542
446
482

Total
Desired
Panel
New
Final New
Final
Overall
Total
Stratum
New
Cohort
Cohort
Cohort
Stratum
Sampling
Population Allocation
Cohort
Sampling Sampling
Allocation Allocation
Rate
Target
Allocation
Rate
Rate
5,667
673
22
22
742
13%
11%
13%
118
393
221
67
118
100%
100%
100%
528
387
0
0
379
72%
0%
72%
2,728
387
12
12
384
14%
13%
14%
197
387
0
0
197
100%
0%
100%
445
387
48
48
428
98%
86%
96%
2,569
387
28
28
405
16%
15%
16%
4,393
387
14
14
354
8%
9%
8%
1,777
387
17
17
392
22%
21%
22%
3,070
387
16
16
370
12%
12%
12%
1,173
387
56
56
440
38%
33%
37%
1,372
387
61
61
446
33%
28%
32%
2,775
387
12
12
396
14%
14%
14%
675
387
21
21
403
60%
57%
60%
1,638
387
23
23
404
25%
23%
25%
2,467
387
15
15
373
15%
15%
15%
1,391
387
30
30
408
30%
27%
29%
694
387
33
33
416
60%
56%
60%
40
387
0
0
40
100%
0%
100%
743
387
33
33
417
57%
52%
56%
445
387
16
16
393
89%
84%
88%
1,241
387
48
48
428
35%
31%
34%
1,160
387
14
14
394
34%
32%
34%
1,287
387
18
18
398
31%
30%
31%
3,842
387
39
39
411
11%
10%
11%
276
387
63
46
276
100%
100%
100%
1,558
387
8
8
350
22%
23%
22%
31,561
1,804
87
87
1,954
6%
6%
6%
1,322
413
111
111
522
43%
31%
39%
5,225
488
79
79
585
12%
9%
11%
565
398
157
157
498
100%
70%
88%
5,806
502
31
31
546
9%
8%
9%
527
397
62
62
457
89%
75%
87%
512
410
37
37
443
87%
79%
86%
1,229
441
37
37
474
39%
35%
39%
925
405
20
20
417
45%
42%
45%
2,041
482
8
8
455
22%
23%
22%

Page F-2

SDR OMB Package for the 2017 Cycle

Stratum
2017
129
130
133
134
136
137
139
142
145
148
151
152
154
155
157
158
160
163
166
167
168
169
170
175
180
185
188
189
198
199
200
207
210
211
212
215
217
222
227
230

Fine Field of Doctorate Degree
Botany/Plant Biology
Anatomy
Biometrics & Biostatistics
Epidemiology
Cell/Cellular Biology & Histology
Evolutionary Biology
Ecology
Developmental Biology/Embryology
Endocrinology
Entomology
Immunology
Marine Biology and Biological Oceanography
Molecular Biology
Structural Biology
Microbiology
Cancer Biology
Neurosciences
Nutrition Sciences
Parasitology
Environmental Toxicology
Virology
Toxicology
Genetics/Genomics, Human & Anima
Pathology, Human & Animal
Pharmacology, Human & Animal
Physiology, Human & Animal
Wildlife Biology
Zoology, Other
Biology/Biomedical Sciences,General
Biology/Biomedical Sciences,Other
Speech-Language Pathology & Audiology
Oral Biology/Oral Pathology
Environmental Health
Environmental Toxicology
Health Systems/Service Administration
Public Health
Health Policy Analysis
Kinesiology/Exercise Science
Gerontology
Nursing Science

Survey of Doctorate Recipients

Panel
Cohort
Count
484
481
453
407
607
450
665
466
401
507
628
211
925
248
1,002
490
1,123
513
403
147
417
477
722
565
900
901
0
516
608
531
377
72
378
230
387
403
110
398
50
424

Panel
Cohort
Population
4,668
2,918
2,935
5,586
8,165
1,274
10,334
3,242
714
5,400
8,920
211
19,537
248
12,398
2,103
14,815
4,800
754
147
636
3,433
6,822
3,947
10,573
10,801
0
6,064
8,558
6,069
4,083
72
1,734
230
1,341
5,714
110
3,156
50
10,757

New
Cohort
Count
190
40
330
673
651
424
904
385
48
222
928
196
1,302
123
945
910
2,137
361
57
84
326
222
757
186
534
418
83
78
506
139
230
34
158
0
159
835
132
513
27
1,115

Total
Frame
Count
674
521
783
1,080
1,258
874
1,569
851
449
729
1,556
407
2,227
371
1,947
1,400
3,260
874
460
231
743
699
1,479
751
1,434
1,319
83
594
1,114
670
607
106
536
230
546
1,238
242
911
77
1,539

Total
Desired
Panel
New
Final New
Final
Overall
Total
Stratum
New
Cohort
Cohort
Cohort
Stratum
Sampling
Population Allocation
Cohort
Sampling Sampling
Allocation Allocation
Rate
Target
Allocation
Rate
Rate
4,858
484
18
18
502
10%
10%
10%
2,958
526
7
7
488
16%
18%
16%
3,265
451
45
45
498
15%
14%
15%
6,259
664
70
70
477
7%
10%
8%
8,816
560
41
41
648
7%
6%
7%
1,698
461
115
115
565
35%
27%
33%
11,238
610
48
48
713
6%
5%
6%
3,627
458
48
48
514
14%
13%
14%
762
402
25
25
426
56%
52%
56%
5,622
498
19
19
526
9%
9%
9%
9,848
581
53
53
681
7%
6%
7%
407
395
190
190
401
100%
97%
99%
20,839
796
49
49
974
5%
4%
5%
371
394
131
123
371
100%
100%
100%
13,343
975
68
68
1,070
8%
7%
8%
3,013
519
156
156
646
23%
17%
21%
16,952
1,133
141
141
1,264
8%
7%
7%
5,161
488
34
34
547
11%
9%
11%
811
424
28
28
431
53%
49%
53%
231
391
142
84
231
100%
100%
100%
962
429
145
145
562
66%
45%
58%
3,655
458
27
27
504
14%
12%
14%
7,579
721
71
71
793
11%
9%
10%
4,133
571
25
25
590
14%
14%
14%
11,107
884
41
41
941
9%
8%
8%
11,219
892
32
32
933
8%
8%
8%
83
389
389
83
0
0%
100%
0%
6,142
512
6
6
522
9%
8%
8%
9,064
569
31
31
639
7%
6%
7%
6,208
511
11
11
542
9%
8%
9%
4,313
387
20
20
397
9%
9%
9%
106
387
124
34
106
100%
100%
100%
1,892
387
31
31
409
22%
20%
22%
230
387
0
0
230
100%
0%
100%
1,500
387
41
41
428
29%
26%
29%
6,549
387
48
48
451
7%
6%
7%
242
387
211
132
242
100%
100%
100%
3,669
387
53
53
451
13%
10%
12%
77
387
134
27
77
100%
100%
100%
11,872
387
36
36
460
4%
3%
4%

Page F-3

SDR OMB Package for the 2017 Cycle

Stratum
2017
240
245
250
280
298
299
300
303
306
309
312
315
316
318
321
324
327
330
333
336
337
339
342
345
348
351
357
360
363
366
369
372
373
376
398
399
400
410
415
418

Fine Field of Doctorate Degree
Pharmaceutical Sciences
Rehabilitation/Therapeutic Services
Veterinary Sciences
Health and Behavior
Health Sciences, General
Health Sciences, Other
Aerospace, Aeronautical & Astronautical
Agricultural Engineering
Bioengineering & Biomedical Engineering
Ceramic Sciences Engineering
Chemical Engineering
Civil Engineering
Structural Engineering
Communications Engineering
Computer Engineering
Electrical, Electronics & Communications Engineering
Engineering Mechanics
Engineering Physics
Engineering Science
Environmental Health Engineering
Geotechnical and Geoenvironmental Engineering
Industrial & Manufacturing Engineering
Materials Science Engineering
Mechanical Engineering
Metallurgical Engineering
Mining & Mineral Engineering
Nuclear Engineering
Ocean Engineering
Operations Research
Petroleum Engineering
Polymer & Plastics Engineering
Systems Engineering
Transportation and Highway Engineering
Engineering Management & Administration
Engineering, General
Engineering, Other
Computer Science
Information Science/Systems
Robotics
Computer/Information Sciences, General

Survey of Doctorate Recipients

Panel
Cohort
Count
399
383
366
0
370
381
698
467
840
391
1,345
1,144
296
445
962
3,634
505
372
435
514
100
658
975
1,534
459
388
526
411
380
439
447
546
156
291
428
556
1,844
527
185
0

Panel
Cohort
Population
5,729
1,171
1,811
0
1,128
3,969
7,998
2,554
11,261
1,028
23,662
18,999
296
863
7,284
40,799
3,892
937
1,269
3,532
100
6,897
14,572
28,398
2,968
625
3,976
663
2,785
1,483
1,589
2,150
156
291
1,647
4,826
26,764
2,845
185
0

New
Cohort
Count
550
170
98
214
75
191
746
147
2,174
0
1,975
1,255
226
59
877
3,956
136
66
132
553
140
541
1,707
2,798
58
0
286
66
205
212
143
200
207
89
82
327
3,321
309
161
211

Total
Frame
Count
949
553
464
214
445
572
1,444
614
3,014
391
3,320
2,399
522
504
1,839
7,590
641
438
567
1,067
240
1,199
2,682
4,332
517
388
812
477
585
651
590
746
363
380
510
883
5,165
836
346
211

Total
Desired
Panel
New
Final New
Final
Overall
Total
Stratum
New
Cohort
Cohort
Cohort
Stratum
Sampling
Population Allocation
Cohort
Sampling Sampling
Allocation Allocation
Rate
Target
Allocation
Rate
Rate
6,279
387
33
33
432
7%
6%
7%
1,341
387
48
48
431
33%
28%
32%
1,909
387
19
19
385
20%
19%
20%
214
387
387
129
129
0%
60%
60%
1,203
387
23
23
393
33%
31%
33%
4,160
387
17
17
398
10%
9%
10%
8,744
688
57
57
755
9%
8%
9%
2,701
480
26
26
493
18%
18%
18%
13,435
844
135
135
975
7%
6%
7%
1,028
423
0
0
391
38%
0%
38%
25,637
1,269
96
96
1,441
6%
5%
6%
20,254
1,080
66
66
1,210
6%
5%
6%
522
405
175
175
471
100%
77%
90%
922
452
28
28
473
52%
48%
51%
8,161
951
101
101
1,063
13%
12%
13%
44,755
3,481
306
306
3,940
9%
8%
9%
4,028
528
17
17
522
13%
13%
13%
1,003
422
27
27
399
40%
41%
40%
1,401
435
41
41
476
34%
31%
34%
4,085
528
70
70
584
15%
13%
14%
240
395
230
140
240
100%
100%
100%
7,438
641
46
46
704
10%
9%
9%
16,279
941
98
98
1,073
7%
6%
7%
31,196
1,461
128
128
1,662
5%
5%
5%
3,026
492
9
9
468
15%
16%
15%
625
409
0
0
388
62%
0%
62%
4,262
535
35
35
561
13%
12%
13%
729
412
37
37
448
62%
56%
61%
2,990
393
26
26
406
14%
13%
14%
1,695
445
55
55
494
30%
26%
29%
1,732
446
36
36
483
28%
25%
28%
2,350
552
46
46
592
25%
23%
25%
363
399
228
207
363
100%
100%
100%
380
400
93
89
380
100%
100%
100%
1,729
447
20
20
448
26%
24%
26%
5,153
566
35
35
591
12%
11%
11%
30,085
1,551
169
169
2,013
7%
5%
7%
3,154
511
49
49
576
19%
16%
18%
346
400
186
161
346
100%
100%
100%
211
395
395
132
132
0%
63%
63%

Page F-4

SDR OMB Package for the 2017 Cycle

Stratum
2017
419
420
425
430
435
440
445
450
455
460
465
498
499
500
505
509
510
512
514
518
519
520
522
524
526
528
530
532
534
538
539
540
542
544
546
548
550
552
558
559

Fine Field of Doctorate Degree
Computer/Information Sciences, Other
Applied Mathematics
Algebra
Analysis & Functional Analysis
Geometry/Geometric Analysis
Logic
Number Theory
Statistics
Topology/Foundations
Computing Theory & Practice
Operations Research
Mathematics/Statistics, General
Mathematics/Statistics, Other
Astronomy
Astrophysics
Astronomy, Other
Atmospheric Chemistry & Climatology
Atmospheric Physics & Dynamics
Meteorology
Atmospheric Science/Meteorology, General
Atmospheric Science/Meteorology, Other
Analytical Chemistry
Inorganic Chemistry
Nuclear Chemistry
Organic Chemistry
Medicinal Chemistry
Physical Chemistry
Polymer Chemistry
Theoretical Chemistry
Chemistry, General
Chemistry, Other
Geology
Geochemistry
Geophysics & Seismology
Paleontology
Mineralogy & Petrology
Stratigraphy & Sedimentation
Geomorphology & Glacial Geology
Geological & Earth Sciences, Genera
Geological & Earth Sciences, Other

Survey of Doctorate Recipients

Panel
Cohort
Count
451
388
357
361
373
351
373
389
357
332
374
376
368
405
424
47
377
377
345
381
373
1,052
964
350
1,706
511
1,152
578
495
950
611
376
369
379
349
353
347
359
351
367

Panel
Cohort
Population
1,261
8,809
3,950
5,145
2,345
986
1,609
8,528
2,712
1,500
1,034
5,771
2,670
2,818
3,948
47
1,047
959
909
966
705
12,206
10,779
458
24,199
2,700
14,291
3,624
2,337
10,492
4,276
4,487
2,152
3,273
1,055
1,001
1,020
684
974
1,194

New
Cohort
Count
148
912
319
298
228
47
170
782
164
43
62
478
180
198
345
15
78
110
62
117
46
804
726
0
1,232
165
708
262
214
816
420
265
179
273
68
36
32
52
114
112

Total
Frame
Count
599
1,300
676
659
601
398
543
1,171
521
375
436
854
548
603
769
62
455
487
407
498
419
1,856
1,690
350
2,938
676
1,860
840
709
1,766
1,031
641
548
652
417
389
379
411
465
479

Total
Desired
Panel
New
Final New
Final
Overall
Total
Stratum
New
Cohort
Cohort
Cohort
Stratum
Sampling
Population Allocation
Cohort
Sampling Sampling
Allocation Allocation
Rate
Target
Allocation
Rate
Rate
1,409
442
46
46
497
36%
31%
35%
9,721
408
37
37
425
4%
4%
4%
4,269
396
29
29
386
9%
9%
9%
5,443
399
21
21
382
7%
7%
7%
2,573
393
34
34
407
16%
15%
16%
1,033
389
17
17
368
36%
36%
36%
1,779
391
37
37
410
23%
22%
23%
9,310
407
33
33
422
5%
4%
5%
2,876
393
22
22
379
13%
13%
13%
1,543
390
10
10
342
22%
23%
22%
1,096
389
21
21
395
36%
34%
36%
6,249
401
30
30
406
7%
6%
6%
2,850
393
24
24
392
14%
13%
14%
3,016
401
26
26
431
14%
13%
14%
4,293
407
32
32
456
11%
9%
11%
62
387
94
15
62
100%
100%
100%
1,125
387
26
26
403
36%
33%
36%
1,069
387
39
39
416
39%
36%
39%
971
387
24
24
369
38%
39%
38%
1,083
387
41
41
422
39%
35%
39%
751
387
23
23
396
53%
50%
53%
13,010
1,026
62
62
1,114
9%
8%
9%
11,505
960
58
58
1,022
9%
8%
9%
458
411
0
0
350
76%
0%
76%
25,431
1,660
77
77
1,783
7%
6%
7%
2,865
529
30
30
541
19%
18%
19%
14,999
1,135
52
52
1,204
8%
7%
8%
3,886
577
38
38
616
16%
15%
16%
2,551
514
42
42
537
21%
20%
21%
11,308
948
67
67
1,017
9%
8%
9%
4,696
616
54
54
665
14%
13%
14%
4,752
387
21
21
397
8%
8%
8%
2,331
387
29
29
398
17%
16%
17%
3,546
387
29
29
408
12%
11%
12%
1,123
387
23
23
372
33%
34%
33%
1,037
387
13
13
366
35%
36%
35%
1,052
387
11
11
358
34%
34%
34%
736
387
27
27
386
52%
52%
52%
1,088
387
39
39
390
36%
34%
36%
1,306
387
32
32
399
31%
29%
31%

Page F-5

SDR OMB Package for the 2017 Cycle

Stratum
2017
560
561
564
565
566
568
569
570
572
574
576
577
578
579
585
590
595
599
600
602
603
606
609
612
613
614
615
618
620
621
624
627
630
633
636
639
648
649
650
651

Fine Field of Doctorate Degree
Acoustics
Atomic/Molecular/Chemical Physics
Particle (Elementary) Physics
Biophysics
Fluids
Nuclear Physics
Optics/Phototonics
Plasma/Fusion Physics
Polymer Physics
Condensed Matter/Low Temperature Physics
Applied Physics
Medical Physics/Radiological Science
Physics, General
Physics, Other
Hydrology & Water Resources
Oceanography, Chemical & Physica
Marine Sciences
Ocean/Marine, Other
Clinical Psychology
Behavioral Analysis
Cognitive Psychology & Psycholinguistics
Comparative Psychology
Counseling
Developmental & Child Psychology
Human Development & Family Studies
Health and Medical Psychology
Experimental Psychology
Educational Psychology
Family Psychology
Industrial & Organizational
Personality Psychology
Physiological/Psychobiology Psychology
Psychometrics
Psychometrics & Quantitative Psychology
School Psychology
Social Psychology
Psychology, General
Psychology, Other
Anthropology, General
Gender and Women's Studies

Survey of Doctorate Recipients

Panel
Cohort
Count
359
399
441
399
344
380
417
396
391
512
397
308
461
431
373
364
378
383
2,582
98
567
349
1,103
728
514
110
714
510
414
637
397
536
350
405
562
706
814
646
1,274
0

Panel
Cohort
Population
647
4,016
7,375
1,082
532
3,748
3,676
2,447
635
13,255
1,257
308
8,502
5,815
1,182
3,421
1,200
788
46,057
98
3,973
388
15,380
7,526
2,778
110
7,660
3,403
819
5,536
1,076
3,750
404
577
4,072
7,240
9,451
6,005
15,124
0

New
Cohort
Count
39
246
487
258
0
188
410
170
70
827
302
187
514
258
139
155
83
36
2,398
114
421
0
796
435
275
181
285
124
97
424
42
251
0
81
225
461
565
322
285
74

Total
Frame
Count
398
645
928
657
344
568
827
566
461
1,339
699
495
975
689
512
519
461
419
4,980
212
988
349
1,899
1,163
789
291
999
634
511
1,061
439
787
350
486
787
1,167
1,379
968
1,559
74

Total
Desired
Panel
New
Final New
Final
Overall
Total
Stratum
New
Cohort
Cohort
Cohort
Stratum
Sampling
Population Allocation
Cohort
Sampling Sampling
Allocation Allocation
Rate
Target
Allocation
Rate
Rate
686
390
21
21
380
55%
54%
55%
4,262
407
23
23
422
10%
9%
10%
7,862
423
26
26
467
6%
5%
6%
1,340
393
76
76
475
37%
30%
35%
532
389
0
0
344
65%
0%
65%
3,936
405
19
19
399
10%
10%
10%
4,086
406
40
40
457
11%
10%
11%
2,617
399
25
25
421
16%
15%
16%
705
390
39
39
430
62%
56%
61%
14,082
452
26
26
538
4%
3%
4%
1,559
394
76
76
473
32%
25%
30%
495
389
147
147
455
100%
79%
92%
9,016
429
24
24
485
5%
5%
5%
6,073
415
17
17
448
7%
7%
7%
1,321
387
40
40
413
32%
29%
31%
3,576
387
16
16
380
11%
10%
11%
1,283
387
24
24
402
32%
29%
31%
824
387
17
17
400
49%
47%
49%
48,455
2,333
111
111
2,693
6%
5%
6%
212
395
212
114
212
100%
100%
100%
4,394
561
53
53
620
14%
13%
14%
388
403
0
0
349
90%
0%
90%
16,176
1,039
49
49
1,152
7%
6%
7%
7,961
707
37
37
765
10%
9%
10%
3,053
506
45
45
559
19%
16%
18%
291
398
248
181
291
100%
100%
100%
7,945
713
24
24
738
9%
8%
9%
3,527
534
18
18
528
15%
15%
15%
916
423
44
44
458
51%
45%
50%
5,960
624
43
43
680
12%
10%
11%
1,118
433
15
15
412
37%
36%
37%
4,001
547
33
33
569
14%
13%
14%
404
404
0
0
350
87%
0%
87%
658
413
50
50
455
70%
62%
69%
4,297
559
28
28
590
14%
12%
14%
7,701
699
40
40
746
10%
9%
10%
10,016
793
43
43
857
9%
8%
9%
6,327
642
31
31
677
11%
10%
11%
15,409
1,019
18
18
1,292
8%
6%
8%
74
388
388
74
74
0%
100%
100%

Page F-6

SDR OMB Package for the 2017 Cycle

Stratum
2017
652
655
656
658
662
665
667
668
670
674
676
678
682
684
685
686
690
694
698
699
710
770
773
822
930
DIS

Fine Field of Doctorate Degree
Area /Ethnic/Cultural Studies
Anthropology, Cultural
Anthropology, Physical and Biologica
Criminology
Demography/Population Studies
Natural Resource/Environmental Economics
Economics
Econometrics
Geography
International Relations/Affairs
Linguistics
Political Science & Government
Public Policy Analysis
Gerontology
Natural Resource/Environmental Policy
Sociology
Statistics
Urban Affairs/Studies
Social Sciences, General
Social Sciences, Other
History, Science &Technology & Society
American/U.S. Studies
Archaeology
Educational Psychology
Operations Research
Discontinued Fields
Overall

Survey of Doctorate Recipients

Panel
Cohort
Count
397
0
0
473
396
98
2,781
433
538
456
603
986
503
106
0
1,614
365
426
379
502
401
398
443
896
368
844
113,814

Panel
Cohort
Population
1,333
0
0
1,695
623
98
36,134
1,057
6,032
3,511
7,941
20,193
4,102
106
0
20,694
1,784
2,579
1,072
4,940
1,345
2,774
1,358
11,360
1,652
20,727
1,035,376

New
Cohort
Count
171
584
148
183
52
96
2,266
86
594
209
562
1,635
570
52
92
1,423
42
66
114
304
148
250
226
547
200
0
82,522

Total
Frame
Count
568
584
148
656
448
194
5,047
519
1,132
665
1,165
2,621
1,073
158
92
3,037
407
492
493
806
549
648
669
1,443
568
844
196,336

Total
Desired
Panel
New
Final New
Final
Overall
Total
Stratum
New
Cohort
Cohort
Cohort
Stratum
Sampling
Population Allocation
Cohort
Sampling Sampling
Allocation Allocation
Rate
Target
Allocation
Rate
Rate
1,504
408
44
44
441
30%
26%
29%
584
410
410
82
82
0%
14%
14%
148
393
393
79
79
0%
53%
53%
1,878
463
44
44
517
28%
24%
28%
675
396
30
30
426
64%
58%
63%
194
396
195
95
193
100%
100%
99%
38,400
2,286
131
131
2,912
8%
6%
8%
1,143
444
33
33
466
41%
38%
41%
6,626
480
42
42
580
9%
7%
9%
3,720
440
23
23
479
13%
11%
13%
8,503
507
32
32
635
8%
6%
7%
21,828
692
51
51
1,037
5%
3%
5%
4,672
451
55
55
558
12%
10%
12%
158
389
126
52
158
100%
100%
100%
92
388
388
92
92
0%
100%
100%
22,117
1,288
79
79
1,693
8%
6%
8%
1,826
391
9
9
374
20%
21%
20%
2,645
426
10
10
436
17%
15%
16%
1,186
404
36
36
415
35%
32%
35%
5,244
462
26
26
528
10%
9%
10%
1,493
408
39
39
440
30%
26%
29%
3,024
436
30
30
428
14%
12%
14%
1,584
452
62
62
505
33%
27%
32%
11,907
872
38
38
934
8%
7%
8%
1,852
391
40
40
408
22%
20%
22%
20,727
908
0
0
844
4%
0%
4%
1,117,898
120,519
13,916
10,766
124,580
11%
13%
11%

Page F-7