ED Response to OMB's Passback

April 29, 2010

MEMORANDUM

To: Shelly Martinez, Kashka Kubzdela

From: Gail Mulligan, Jill Carlivati, Chris Chapman

Subject: Responses to Passbacks for ECLS-K:2011 National Data Collection Clearance Package (OMB# 1850-0750 v.8)

This memo contains responses to the questions received from OMB in the first passback on 4/22/2010 and in the second passback on 4/26/2010.

We also want to provide updated information about the vision and hearing screenings for which we had indicated previously we would be sending a change memo. We have decided to field these screenings in the fall first grade collection. This decision made the most sense because the federal cosponsors of the screenings wanted to conduct them on a subsample and fall first grade was already planned as a subsample. We will submit an official request for approval for the screenings with the fall first grade clearance package. We have revised the current submission (Parts A and B) to reflect this. The burden for the screenings has been added to table A-6, which outlines estimated burden for upcoming activities.

Please let us know if you have further questions on the screenings or the responses to your questions below.

Q. Page A-27, please either create two lists and/or denote which folks were consulted for which round.  As written, we find it a little misleading.  For example, several other folks listed are no longer affiliated as listed, so if they haven’t been consulted since they had those affiliations, it should be so noted.  Further, two of the three folks listed under OMB do not currently work here and in fact, current employees do not know who they are.  Finally, Brian Harris-Kojetin’s last name is misspelled.

NCES: We have addressed this issue by adding a note indicating which staff provided input on the ECLS-K but not the ECLS-K:2011. For these staff, we also noted that the affiliation listed is the person’s affiliation at the time he or she provided input on the study.

We apologize for misspelling Brian’s name and have fixed the error.

Q. Page A-41, please clarify both in the table title and in the “study total” row that this is just a year 1 burden estimate.  Similarly for the table on page A-45, it is only of year 2, right?

NCES: On page A-41, table A-5, we have clarified that the burden is just for the collections in the base year (i.e., kindergarten). The burden included in table A-6 cuts across two survey years (first and second grade), but it is only a portion of the burden for those years, because national data collection for the springs of first and second grade are not included. Please let us know if the way in which we have clarified this is not sufficient.

Q. Page B-1, we wanted to clarify that “is” (rather than “is not”) is correct in the sentence that reads: “Although they will be oversampled as one group, the number of completed interviews for children in each of these groups is expected to be large enough….”

 NCES: Yes, this is correct. We expect to be able to report out overall distributions of race, as well as outcomes by race, for Asians separately from Native Hawaiians and Other Pacific Islanders. However, it likely will not be possible to conduct more detailed analyses with finer breakdowns on these two groups separately, for example reporting outcomes by mother’s education by race. (For an example of reporting by race/ethnicity in the ECLS-K, see America’s Kindergartners at http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2000070).

Per another question below, we have revised this sentence to clarify that we will report on Asians separately from Native Hawaiian and Other Pacific Islanders.

Q. Page B-11, are there really still pending issues or are these resolved at this point?  If they can’t be submitted now, they would require a change request later.

NCES: These issues have been resolved. A memo regarding assessment routing was forwarded to OMB on 3/4/2010.We have revised this text in Part B with information contained in that memo. We also have updated Part A to reflect these routing decisions (see pages A-6 and A-7).

Q. Page B-14, Please clarify if the translators are being asked to sign an affidavit of nondisclosure or how confidentiality will otherwise be assured in those cases. 

NCES: Translators will be required to sign the NCES Affidavit of Nondisclosure and a Confidentiality Pledge. We have clarified this in the package on page B-15.

Q. Page B-16, the information on how nonresponse bias will be assessed is insufficient.  Please provide more detail.

NCES: We have added more information on statistical assessment of nonresponse bias on page B-30.

Q. Throughout, we appreciate that NCES updated that references to Asian/Pacific Islanders, but we wanted to clarify that “Asian” is one category and that “Native Hawaiian and Other Pacific Islander” is a second.  There are not three categories, which is how it is written.  Also, these are all “race” categories.  They are not “race/ethnicity” categories.

NCES: We apologize for the confusion that arose as a result of the way we wrote this. We do realize that Native Hawaiian and Other Pacific Islander is one group, and that these are races, not ethnicities. We have revised the text on page B-1 to make this more clear. Specifically, at the bottom on page B-1, we have revised the text to the following:

Although they will be oversampled as one group, the number of completed interviews for children in these groups is expected to be large enough in the kindergarten year to produce estimates for Asians separately from Native Hawaiians and Other Pacific Islanders.

We generally have left references to the three groups in other locations but have made a few revisions. Since we cannot oversample Asians separately from Native Hawaiians and Other Pacific Islanders in the initial stage of sampling, we believe it is appropriate in instances where we are talking about the oversample to refer to the three groups in the following way: “Asians, Native Hawaiians, and Other Pacific Islanders” (e.g., see highlighted section on page B-1). Our thought is that if we were to refer to the groups as “Asians and Native Hawaiians and Other Pacific Islanders” this would give the incorrect impression that we can oversample Asians separately from Native Hawaiians and Other Pacific Islanders.”(Note, we found an instance where we used “Asian and Native Hawaiian and Other Pacific Islander” when discussing sampling, so we changed that to “Asian, Native Hawaiian, and Other Pacific Islander” (see highlighted section on page B-5). We also changed “Asians, Native Hawaiians and Other Pacific Islanders” to” Asians, Native Hawaiians, and Other Pacific Islanders” on page B-4.

When we refer to these groups as analytic subgroups, we generally refer to them in the following ways: “Asians, Native Hawaiians and other Pacific Islanders, …” (e.g., see highlighted section on page B-2)

“Asian or Native Hawaiian and Other Pacific Islander groups” (e.g., see highlighted section on page B-4)

We found one instance in which we did not on page B-5 (in the discussion of findings from the ECLS-K) and have corrected it.

Hopefully the original text on page B-1 that suggested we would report out on all three groups separately was the main problem and that the revision helps to make this issue clear when reading the rest of the section. Please let us know if you would like further changes.

We have reviewed the text for use of the term race/ethnicity and have revised three instances in which “Asian” and “Native Hawaiian and Other Pacific Islander” were referred to incorrectly as race/ethnicity groups (see page B-5 and footnote 6 on page B-7). Please note that the term “race/ethnicity” is still used in the second full paragraph on page B-4, because this text describes the distribution by race/ethnicity in the first ECLS-K study based on a set of mutually exclusive race/ethnicity categories that includes Hispanic.

Q. Based on discussions between OMB and the school meal evaluation staff at FNS, we would suggest two changes that impact  questions A-20 and A-24 on the school administrator questionnaire. 

1. The concept of “eligible” children is misused here.  Instead, the questionnaire should ask for the number of children “approved” for free and reduced priced meals.  “Eligibility” doesn’t work because school administrators would have no way of knowing a child’s family income unless the student applied or was directly certified for meal assistance.  And you only need to ask this question once – as students do not apply separately for breakfast and lunch assistance.

2. Rather than ask the number of participating children on a given day, a better way to calculate participation rates in the meal programs is to ask the total number of meals served free, at reduced price, or at paid rates over the course of a month (such as for the month of October.)  With the enrollment numbers, researchers would be able to calculate the average daily participation over a month, which may be more robust.

OMB staff have mocked up some suggested alternative questions below. 

++++++++++++++++++++++++++++++++++++++++++

Original Questions:

A20.     How many children in your school were (a) eligible for and (b) participating in the school breakfast program as of October 1, 2010, or the date nearest to that for which data are available? WRITE IN NUMBERS BELOW. IF NONE, WRITE “0.” THE NUMBER OF CHILDREN RECEIVING FREE BREAKFAST AND THE NUMBER OF CHILDREN RECEIVING REDUCED-PRICE BREAKFAST SHOULD NOT ADD TO MORE THAN THE TOTAL NUMBER OF PARTICIPATING CHILDREN.

	Eligible Children	AND	Participating Children
a... Any school breakfast	ALL ELIGIBLE
b... Free school breakfast
c... Reduced-price breakfast

A24.     How many children in your school were (a) eligible for and (b) participating in the school lunch program as of October 1, 2010, or the date nearest to that for which data are available? WRITE IN NUMBERS BELOW. IF NONE, WRITE “0.” THE NUMBER OF CHILDREN RECEIVING FREE LUNCH AND THE NUMBER OF CHILDREN RECEIVING REDUCED-PRICE LUNCH SHOULD NOT ADD TO MORE THAN THE TOTAL NUMBER OF PARTICIPATING CHILDREN.

	Eligible Children	AND	Participating Children
a... Any school lunch	ALL ELIGIBLE
b... Free school lunch
c... Reduced-price lunch

++++++++++++++++++++++++++++++++++++++++++

Suggested Alternatives:

Q#.  How many children in your school were certified for free or reduced price meals as of October 1, 2010, or the date nearest to that for which data are available?  WRITE IN NUMBERS BELOW.  IF NONE, WRITE “O”. 

	Children Approved for Free/Reduced Priced Meals
b... Free school meals
c... Reduced-price meals

Q#.  How many federally-reimbursable school lunches did you serve at free, reduced price, and paid rates over the entire month of October?  WRITE IN NUMBERS BELOW.  IF NONE, WRITE “O”. 

		Total meals served in October
a... Paid school lunches
b... Free school lunches
c... Reduced-price lunches

Q#.  How many federally-reimbursable school breakfasts did you serve at free, reduced price, and paid rates over the entire month of October?  WRITE IN NUMBERS BELOW.  IF NONE, WRITE “O”. 

		Total meals served in October
a... Paid school breakfasts
b... Free school breakfasts
c... Reduced-price breakfasts

NCES: Thank you for the new items. All of your suggested revisions to this set of questions have been incorporated into the revised Spring 2011 Kindergarten School Administrator Questionnaire, in Appendix C of this clearance submission package.

Page 6 of 6