Annual Survey of Manufactures MA-10000 Pretesting Plan

This letter is to inform you of our plans to conduct additional research under the generic clearance for questionnaire pretesting research (OMB number 0607-0725). The Census Bureau plans to conduct respondent debriefings for the Annual Survey of Manufactures (ASM; form MA-10000). The ASM is conducted on an annual basis, normally as a sample survey, but the MA-10000 form is also used for the Economic Census in years ending in “2” and “7.” The form used for the 2012 reference year census is the focus of this study. The MA-10000 is a self-administered mail-out/mail-back questionnaire with a Web-reporting option, and is used to collect information on manufacturing establishments in the U.S., including sales, shipments, inventories, employment, payroll, fringe benefits, expenditures, etc.

The purpose of the respondent debriefings is to understand how respondents answered a complex of items related to inventories when they completed their 2012 MA-10000 form and to identify the causes of misreporting for this item, particularly the LIFO (“last-in, first-out,” an inventory valuation method) items. Specifically, we intend to address the following questions: Is “LIFO” a familiar concept to MA-10000 respondents? Do respondents have issues with comprehending the terminology used on the questionnaire? What terms for LIFO-related concepts are used within companies that use it? Does respondents’ understanding of LIFO depend on their position, training, or location within the company (i.e., at a corporate headquarters versus at a manufacturing establishment)? What role does the visual design of the inventory complex play in causing incorrect reporting? Are the requested data available in respondent’s records? Findings will be used to inform any future changes to the MA-10000 form.

Starting as early as April 2013 and continuing as late as September 2013, we plan to conduct a maximum of 50 interviews by telephone with companies throughout the U.S. that complete the MA-10000 form. The sample will be selected on a rolling basis as responses are received. Cases will be selected from industries in which LIFO is commonly used. These include manufacturing industries that require market-volatile commodities -- such as petroleum, steel, and rubber -- as inputs. Most cases will be selected on the basis of apparent reporting errors, but a small number of interviews (up to 10) will be conducted with cases whose data appear correct, in order to understand what characteristics of respondents or companies, of the item complex, and of the LIFO concept contribute to correct reporting. After companies are recruited, Census Bureau staff will provide follow-up reminders by telephone and/or email.

Debriefings will be conducted by members of the Response Improvement Research Staff of the Census Bureau. Subject matter experts from the Manufacturing and Construction Division will likely participate in most or all of the interviews to observe and to clarify survey content as needed.

We will record the interviews with the participants’ permission, to aid researchers in accurately reporting findings and recommendations. We also will inform the participants that the information they provide is confidential and will not be shared with anyone not directly involved in this research project. Respondents will be told that their participation is voluntary. Monetary incentives for participation will not be offered.

We expect that each interview will last approximately 30 minutes. Thus, the total estimated reporting burden for this research is approximately 25 hours (.5 hours X 50 interviews).

Enclosed are a copy of the MA-10000 questionnaire and a draft of the respondent debriefing protocol.

The contact person for questions regarding data collection and analysis of the respondent debriefing is listed below:

Alfred D. Tuttle

Response Improvement Research Staff

Office of Economic Planning and Innovation

U.S. Census Bureau,

Washington, D.C. 20233

(301) 763-7809

Alfred.D.Tuttle@census.gov