SRS Responses

Non-substantive Change Request to the GSS OMB Clearance

OMB Q1: The recent ROCIS submission for this collection does not show any change for burden hours or number of respondents.  However, it looks like there are more respondents (i.e., the postdoc coordinators) and more questions, so we would like to hear more about what burden information was collected during the pilot and what the expected effect will be on the full collection.

NSF A1: We have reviewed the burden hour estimates based on the 2009 survey collection, and would like to revise our estimates as listed in table 1. The initial request for clearance for the 2010 GSS had estimated an increase in burden hours of 30 minutes for a subset of units that have postdocs for responding to expanded postdoc questions.  It is difficult to estimate overall average burden estimates for the expanded postdoc questions based on the postdoc pilot study because only small schools were asked to provide the burden estimates. Large schools in the pilot were not asked to estimate additional burden because they were only asked to fill out the postdoc pilot items for a few units, rather than all units with postdocs.

NSF estimates that the overall increase in burden in 2010 for the new postdoc questions to be 1.5 hours per unit based on the median burden hours for small schools from the postdoc pilot study. NSF expects that postdoc reporting will be more efficient in large schools.

In future years, NSF expects the time required for reporting about postdocs will be lower as institutions develop procedures to support expanded GSS reporting on postdocs. Such a phenomenon can be seen in the time required to complete the GSS which declined from 2.5 hours in 2007 (the first year GSS went to a two-stage reporting procedure and emphasized Part 1, the listing of eligible units) to 1.9 hours in 2009.

For calculating burden hours in GSS, respondents are the number of units for which data are being reported, not the number of individuals within the units that are actually providing the data. Institutions and units within institutions vary widely on the number of people providing GSS data for any single unit, depending on how centralized or decentralized the administrative records are in an institution and how the school coordinator and then unit respondents choose to handle the reporting.  A school coordinator may do it all themselves for many units, may call on people in several different administrative units (e.g., Registrar, Office of Institutional Research, Financial Aid, or Human Resources) for each unit, or may totally disperse the reporting to unit respondents, who themselves may do it all themselves or call on different administrative units to provide the data.  GSS has not attempted to estimate the number of individuals involved in providing GSS data but only the number of units for which data are reported and the average burden hours per unit.

Although the distribution of burden hours for 2010 in Table 1 differs from the original clearance package, the total estimated burden hours for 2010 (35,521) remains substantially below the number of hours (46,247) for which the 2010 GSS has been cleared. The estimated number of respondents (19,185) is also lower than previously approved for the 2010 GSS (22,635).

Table 1: Revised Burden Estimates for 2010 GSS
Category	# of Units	Burden/ unit (hours)	Total Burden (hours)	Notes
2009 existing units and 5% increase	13,931	1.9	26,469	Estimated average GSS response burden per unit has decreased from 2.5 hrs in 2007 to 1.9 hrs in 2009.
FFRDC study	40	4.7	188	Unchanged.
Postdoc Coordinators Part 1 (preparation)	5,214	0.2	1,043	School coordinators report 0.2 burden hours, and as a conservative estimate, the same level is estimated for postdoc coordinators. However, burden hours for Postdoc Coordinators are likely to lower than that because in general fewer units in an institution will have postdocs than have graduate students.
Additional postdoc questions Part 2 (reporting)	5,214	1.5	7,821	Median burden estimate of 1.5 hours from small schools in the postdoc pilot study is used
Estimated total	19,185		35,521
Total original clearance	22,635		46,247

OMB Q2: In addition, there was a Term of Clearance on this collection that said, “OMB appreciates that SRS/NSF has conducted systematic outreach as part of its updating of this collection. Since SRS/NSF has learned that a small number of institutions may have difficulty meeting the schedule for the first round of collection due to reprogramming requirements, SRS/NSF will offer technical assistance as well as alternative reporting options for year one, as needed. OMB looks forward to a report back on the number of institutions that required an alternative reporting option.”  Before we make additional changes to the survey, we would like to understand the effects of the last round.

NSF A2: In 2008, 19 schools submitted data by providing data files in the old format.