IC Template

REQUEST FOR APPROVAL UNDER THE “CFPB GENERIC
INFORMATION COLLECTION PLAN FOR STUDIES OF CONSUMERS
USING CONTROLLED TRIALS IN FIELD AND ECONOMIC
LABORATORY SETTINGS” (OMB CONTROL NUMBER: 3170-0048)
PART A. GENERAL INFORMATION
1. Title of the Information Collection (Study): Economic Research on Choice Overload
2. Study Abstract:
Our goal is to study how individuals make choices among alternatives when each alternative varies
along several dimensions. Specifically, we will ask students at a college conducting the research
under contract to CFPB to choose among several pre-paid debit cards as their compensation for
participating. The cards differ in their attributes or features, such as how much they pay now vs.
later, fees, color, etc. By varying how many different attributes and cards the participant sees, such
that he/she faces either an easy decision (only a few features to keep track of) or a much harder
decision (cards that differ from each other in many ways), we will learn about whether the number
of features influences decision-making quality. In order to assess the quality of the decision, as
well as the cost of providing too much information, we first collect data on each individual’s risk
aversion and patience.
3. Type of Collection:
a.

Will there be an informed consent? [ X] Yes [ ] No [ ] N/A
Explain why or why not an informed consent is being used.

An informed consent will be given to participants as part of standard procedures required by the
contracted college’s Institutional Review Board (IRB).
b.

How will you collect the information? (Check all that apply)
[ ] Field Study [X ] Laboratory Trials
[ ] Other, Explain ______________________

c.

Will interviewers or facilitators be used? [ ] Yes [X ] No [ ] N/A

4. Personally Identifiable Information:
a. Is personally identifiable information (PII) collected? [X] Yes [ ] No
1. If yes, explain direct identifying PII and/or other PII and relevant uses.
PII will be collected to facilitate recruitment into the study; when participation is solicited, those
who are interested in participating will provide their name and email address so that they can be
contacted about the study if needed for scheduling purposes. This information will be used for
recruitment only and will not be connected to the participant’s responses to the study at any point.
Page 1 of 7

The contractor will not include any direct personally identifying information in the final data set
delivered to CFPB.
The contractor will also need to collect some information while the experiment is in progress for
the purposes of payment and linking participants’ decisions across the two stages. Because the
data collection will take place on two different occasions several months apart, the contractor will
ask participants to provide their college ID number in both stages. The ID number will be
necessary to ensure that responses can be linked across the two stages, and the college ID number
was selected because the participants will be more likely to remember and report the same
number than one that is randomly generated. In each stage, the computer program will add a
constant 6 digit number to each participant’s ID, which is known only to the contractor’s
researchers. Thus the ID number would be masked if anyone gained access to the data. The ID
numbers will be stripped from the data as soon as payment is finished and before delivery to
CFPB.
Additionally, the contractor will ask a series of questions about participant demographics and
backgrounds that may be jointly considered PII (see the attached questionnaire). These questions
are included either to a) allow us to compare our sample to the total population in order to
determine the external validity of our results, b) to allow us to determine, within our sample,
whether different groups behave differently, and, c) to reduce standard errors, and, thus, sample
size. The participant’s five-digit zip code will be collected for the purposes of linking to U.S.
Census data on household income in the zip code. This will be necessary to estimate income
because the participants, as students, may not be able to provide a reliable self-report of their
family’s income level. After the participant’s zip code has been matched to Census income
information, the zip code will be stripped from the dataset and will not be delivered to the CFPB.
b. If Yes, is the information that will be collected included in records that are subject to the
Privacy Act of 1974? [X ] Yes [ ] No [ ] Not Applicable
1. If Applicable, has a System or Records Notice (SORN) been published?
[X ] Yes [ ] No
2. If Yes, provide SORN title and Federal Register citation for the SORN _77 FR 67802_
Title: __CFPB.022 Market and Consumer Research Records ____________.
c. 1. Has the Privacy Impact Assessment (PIA) been published?
[X ] Yes [ ] No [ ] Not Applicable
2. If Yes, provide link to PIA. If No, please describe the status of the PIA:
http://files.consumerfinance.gov/f/201412_cfpb_market-research-in-the-field-v1.pdf

Page 2 of 7

PART B. JUSTIFICATION

1. Purpose of the Study and Intended Uses of the Data:
Our goal is to study how individuals make choices among alternatives when each alternative varies
along several dimensions. When shopping for financial products, consumers often need to
consider the tradeoffs between multiple attributes of a product for which they have alternatives in
terms of financial service providers. Regulators may require the disclosure of certain attributes of
a product to enable consumers to make comparisons across alternatives. However, little is known
about the welfare implications of providing information about increasing numbers of product
attributes. This study will identify whether there is a cost to providing too much information by
assessing whether consumers select dominated alternatives or alternatives that are contrary to
their stated preferences when the number of attributes and alternatives is sufficiently large. The
data that result from this study will be analyzed for research purposes only.
2. Payments or Gifts (Incentives) to Respondents:
The study will entail two stages, each of which will include payment to participants. In the first
stage of the experiment, participants will be asked a series of questions about how they would
allocate money between payment on a sooner date vs. a later date. One of these questions will be
randomly chosen to be “the decision that counts” and the participant will be paid according to the
choice they made in this decision. For this first stage, participants will spend less than one hour
and the average participant is expected to earn approximately $16, depending on their choices and
which decision is the one randomly chosen as “the one that counts” for that participant.
In order to pay participants for the amount they chose to receive today vs. later, they will selfaddress two envelopes with their campus mailbox number and include a randomly assigned
“subject ID number.” After the experiment, we will stuff the envelopes with the correct amount of
money for that subject number. We will deliver that day’s payment to their mailbox within 24
hours and the later payment to their mailbox on the stated day.
For the second stage, expected earnings for each subject from this part of the experiment are $5060. The amount of the payment will depend upon the payment method that the participant will
select during this stage, and participants will be aware of the amounts that they are selecting to
receive. This payment for the second stage is structured to provide adequate incentive to return to
the laboratory to complete the study, because data collection will take place during two stages. In
addition, the payment amount is intended to approximate an amount that a person might load on
to a prepaid card in real life, to make the payment decision at the end of the study approximate a
natural decision as closely as possible.
The cards used for payment will again be delivered to their mailbox via the same delivery method
(participants will self-address an envelope). Participants will receive a contact sheet for the
Page 3 of 7

contractor’s researchers in case there are any problems with delivery of payments. The contractor
will also retain a file with card account number and mailbox number until the payments are fully
made, so that it can resolve any problems that the participant reports. Card account numbers and
mailbox numbers will not be provided to the CFPB at any point.
3. Assurances of Confidentiality and Justification for Sensitive Questions:
Participants will be given Institutional Review Board (IRB) forms from the contracted college that
describe the study, including privacy information. Participants will also be provided a Privacy Act
Statement during the study. The questions we consider to be the most sensitive are the personality
and financial product usage questions, which are included to give us a sense of whether certain
personality types are associated with differences in credit choices, and whether those with more
experience with financial product use (credit cards, checking accounts, reloadable cards, loans)
are more likely to make optimal choices in the experimental credit choice.

4. Estimated Burden of Information Collection:
Information Collection

First Session
Second Session (Follow-up)
Totals:

No. of
Respondents

Frequency

Total
Annual
Responses

140
140
140*

1
1
//////////////

140
140
280

Average
Response
Time
(hours)
.5
.5
/////////////

Total
Burden
Hours
70
70
140

*Note: Respondents to the First Session are the same as those for the Follow-up Session.

5. Federal Costs (estimated annual cost to the Federal government):
The study will be conducted under a one-time, fixed-price contract with a total award amount of
$107,040.

Page 4 of 7

PART C. STATISTICAL METHODS
1. Respondent Universe and Selection Methods:
Respondents will be members of the contracted college’s pre-existing subject pool made up of
undergraduate students at Middlebury College. Members of the subject pool will be contacted
electronically to solicit participation. In order to participate in a specific study (including this
one), students voluntarily sign-up as a participant using an online portal managed by the college.
Respondents will be enrolled on a first-come, first-serve basis until all sessions are complete.
Individuals can only participate once. There are no additional restrictions on eligibility.
Therefore, the Bureau is not the sponsor of the mechanism use d for recruiting potential
respondents.
2. Information Collection Procedures:
Information will be collected in the contracted college’s experimental laboratory facilities on
campus over two sessions. During the first session, information will be collected via a
questionnaire to measure risk and time preferences as well as background information. The
second session will include collection of each participant’s choice for method of payment.
First Stage:
In the first stage, participants will come to the lab to fill out a questionnaire on demographic
information, typical weekly consumption, Big 5 personality measures, and their subjective
impressions of their own risk-aversion and patience. The questionnaire is attached.
In addition, they will be asked a series of questions about how they would allocate money between
payment on a sooner date vs. a later date (payment over time decisions), and whether they would
prefer sure payments vs. lotteries, where they have some probability of a low payment and some
probability of a high payment (lottery decisions). For example, participants may choose whether
they would prefer a safe lottery in which they will receive either $10.39 or $8.31 or a riskier lottery
in which they will receive either $20 or $0.52 (see attached questionnaire).
Second Stage:
In the second stage (approximately 2 months after the first), participants will again come to the
lab and chose their method of payment. They will make this decision in the experimental software
z-Tree, which will record the data in an Excel sheet accessible only to the researchers. They will
see several pre-paid debits cards that they can choose from. The cards will have an amount loaded
onto them that the participant can spend immediately. The cards might also be reloaded with
another amount one month from the experiment. The exact payment schedule for the cards will be
known to the participants and we will load the card chosen by the participant in exactly the way
described. The cards will vary in several ways, including the amount loaded on the card
Page 5 of 7

immediately, amount that we will load onto the five weeks from the experiment, activation fee,
and monthly fees charged against the immediate payment and the later payment. Note that the
“fees” will be small amounts that are deducted from the amount loaded on the card and
participants will always receive significant positive earnings. Participants will select from among
different cards with differing attributes so that their choices can be connected with their responses
to the questionnaire in the first stage of the study, which will enable the researchers to assess
whether the choices participants make align with their stated preferences, and whether deviations
from their stated preferences are associated with the number of attributes the participants are
presented with during their payment choice.
The study’s incentives to respondents are structured to maximize retention of participants
between the first and second stage, by making their expected earnings from the second stage
significantly higher than in the first stage. Experimenters will stress the importance of returning
for the second stage upon recruitment and in the first stage of data collection, and will include a
box to check when signing up for the experiment where the participant affirms they will
participate in both stages. Students will also receive emails reminding them of the second stage.
3. Testing of Procedures or Methods:
Testing of the questionnaires and the first and second round will take place with a small pilot with
students. During this pilot the relevant computer code used to instruct subjects and record their
choices will be test run to ensure it operates smoothly (following tests among the research team as
well). In addition, the questions related to payments over time and lottery decisions have been
tested with students as part of a class, and analyzed for effectiveness in achieving the variation
necessary for analysis.
4. Contact Information for Statistical Aspects of the Design:
Dustin Beckett, PhD
Economist
Consumer Financial Protection Bureau
[email protected]
Heidi Johnson, MPP
Research Analyst
Consumer Financial Protection Bureau
[email protected]

Page 6 of 7

PART D. CERTIFICATION PURSUANT TO 5 CFR 1320.9, AND THE RELATED PROVISIONS
OF 5 CFR 1320.8(b)(3) :
By submitting this document, the Bureau certifies the following to be true:
(a)
(b)
(c)
(d)
(e)
(f)

(g)
(h)
(i)

It is necessary for the proper performance of agency functions;
It avoids unnecessary duplication;
It uses plain, coherent, and unambiguous terminology that is understandable to respondents;
Its implementation will be consistent and compatible with current reporting and recordkeeping
practices;
It indicates the retention period for recordkeeping requirements;
It informs respondents of the information called for under 5 CFR 1320.8(b)(3):
(i) Why the information is being collected;
(ii) Use of information;
(iii) Burden estimate;
(iv) Nature of response (voluntary);
(v) Nature and extent of confidentiality; and
(vi) Need to display currently valid OMB control number;
It was developed by an office that has planned and allocated resources for the efficient and effective
management and use of the information to be collected;
It uses effective and efficient statistical survey methodology; and
It makes appropriate use of information technology.

PART E. CERTIFICATION FOR INFORMATION COLLECTIONS SUBMITTED UNDER A
GENERIC INFORMATION COLLECTION PLAN
(a)
(b)
(c)
(d)

(e)
(f)
(g)
(h)

The collection is voluntary.
The collection is low-burden for respondents and low-cost for the Federal Government.
The collection is non-controversial and does not raise issues of concern to other federal agencies.
The collection is not intended to be published to the public as an official government statistic to be
externally valid and representative of a population of interest. The results are intended to be
internally valid, not necessarily externally valid.
Information gathered will not be used for the purpose of substantially informing influential policy
decisions.
The collection is targeted to the solicitation of opinions from respondents who have experience with
the topics or issues being studied.
The results will not be used to measure regulatory compliance or for CFPB program performance
evaluation.
The results are not intended to be generalizable or otherwise draw inferences beyond the survey
population

Page 7 of 7