Note to Reviewer

January 11, 2008

MEMORANDUM FOR : Reviewer of 1220-0045

FROM : William McCarthy, Chief

Division of Safety and Health Statistics

Office of Compensation and Working Conditions

Bureau of Labor Statistics

SUBJECT : Survey of Occupational Injuries and Illnesses
Quality Assurance Study and Public Sector data

Attached is a description of a proposed quality assurance study of the Survey of Occupational Injuries and Illnesses (Survey) to be conducted for survey year 2007, which began January 1, 2008. Last year for survey year 2006, a quality assurance survey was conducted to ensure that the Survey correctly captures the data that employers have recorded on their OSHA logs. During that survey, some incorrect data collection practices were discovered and were corrected. Based on these findings, it was decided that an ongoing quality assurance program was necessary to insure proper data collection methods were being followed.

This follow-back study is described in the attachment. BLS plans to conduct this study from date of approval to July 17, 2008, when the current 2007 Survey data collection activities end. It is our intention to continue this quality assurance program, with needed modifications, in future years.

BLS is requesting an increase in burden hours for this quality assurance survey and for the collection of public sector data. The collection of public sector data was included in the original clearance submission from June 2007. The burden hours increase for public sector data collection for 2008 is 9,433 hours and for the quality assurance survey is 578 hours.

OMB approval for SOII is not scheduled to expire until September 30, 2010.

If you have any questions about this request, please contact William McCarthy by telephone at 202-691-6163 or by e-mail at [email protected].

Note to Reviewer of 1220-0045

I.Introduction and Purpose

The Bureau of Labor Statistics (BLS) conducts the Survey of Occupational Injuries and Illnesses (Survey) in conjunction with participating State statistical agencies. The Survey is an annual, mandatory survey of work-related injuries and illnesses experienced by establishments in the United States. The survey captures data from Occupational Safety and Health Administration (OSHA) logs of workplace injuries and illnesses maintained by employers. Traditionally, these data have been obtained from survey respondents by means of a hardcopy survey form.

In survey year 2008, BLS will collect public sector data from all States. Currently, 27 States collect public sector data. All but one of those States have made the collection of public sector data mandatory. BLS will expand the public sector data collection to the remaining States and produce national estimates and State estimates for the participating States. This collection of public sector data was included in the original clearance submission from June 2007.

In survey year 2006, the Survey program conducted a one-year quality assurance study that had primarily a focus on addressing the undercount issue. Specifically, the study was to determine the magnitude of employer error in recording data from their OSHA forms to the different types of BLS collection forms and methods. The sample size was approximately 3,600 employers, or 600 per month from January 2007 to June 2007. The sample size was determined by power analysis to support findings by establishment size class, by month of reporting, and by collection method – Internet e-mail, or paper form. This program was voluntary and the final response rate was 27.5%. The callbacks were conducted by approximately 25 national office staff.

The results showed no systematic under-reporting or over-reporting by employers. There was no strong dependence between error rates and collection methods. The larger employer size classes showed more errors than the smaller size classes. There was also no strong correlation between error rates and month of response.

During this study, certain incorrect collection methods were uncovered. One involved a State employee systematically zeroing out the counts for cases with only days of job transfer or restriction and other recordable cases. Another involved a State employee changing the month the injury or illness occurred to eliminate case edits. These edits arose for employers who were supposed to subsample their cases (usually large employers with many cases) but instead submitted all their cases. These incorrect collection methods were corrected before close of the survey but did underline the importance of an ongoing quality assurance process.

The goal of this project is to create an ongoing Quality Assurance (QA) program for the Survey. The QA program will review data in the Survey collection database against originally reported data from employers. Recontacts of employers using BLS Office of Field Operations (OFO) personnel will be necessary for some establishments.

II. Scope and Approach

This proposed QA program will be conducted for the 2007 survey year and will optimally start in January 2008. The goal of the QA program is to evaluate the quality of the data collected in terms of proper collection methods. If improper collection methods or procedures are uncovered, they will be corrected. A byproduct of this program would be the knowledge that each data collector would have that any form they have processed could be selected for the program.

The maximum sample size will be 30 QA schedules per person per month. There are approximately three persons in each of the six regional offices who will conduct this program plus three national office employees from the BLS Office of Field Operations (OFO). QA of non-participating States whose data are collected by BLS Regional Office staff will be conducted by the national OFO. QA data collection in the regional and national offices will involve a maximum sample size of 3,780 schedules. Stratification variables for the sample will be State, data collection method used by the employer, and size of establishment. The data collection methods used by the employer will be classified as:

• Electronic response - IDCF or the Adobe fillable e-mail form

• Paper response – data collection booklet or FAX

• Telephone response.

For this QA program, schedules eligible for sampling must be usable for estimation. In addition, the clean and final data on the collection database must be different from the employer submitted data and/or have bypassed bypassable edits. The sample will be drawn monthly from January through May and bi-weekly from June through collection closeout.

Electronic responses that are selected will show the original data submitted by the employer and the final data on the collection database. These employers will be recontacted by QA personnel to verify the data and to verify recontact by data collection personnel.

Paper responses that are selected will show the final data on the collection database. QA personnel will need to obtain the original paper response from the State data collection personnel, and if the final data on the data collection database are different from the original data on the paper response, QA personnel will recontact the employer. This recontact will be performed by QA staff to verify the data and to verify recontact by data collection personnel. Also, during the final two weeks of the survey, recontact of all employers submitting paper responses will be performed by QA staff to verify the data.

Telephone responses that are selected will show the final data on the collection database. QA personnel will need to recontact the employer to verify the data and to verify contact by data collection personnel.

The completed forms QA staff use to document the results of the recontacts will be transmitted by the QA staff to the BLS national OFO for analysis on a flow basis. Analysis will include State data collection agency, employer choice of response method, size of establishment, and month of response. Regional office staff will maintain their responsibility to follow-up with State data collection staff on data discrepancies discovered during this QA program.

III. Burden Hours

Additional burden hours of 578 are requested for the QA program. Employers will need to answer a recontact telephone call from BLS QA staff. The call should not exceed 10 minutes. The employer will be asked to verify previous contact by data collection personnel and to verify their data from the summary. BLS estimates that a maximum of 75 percent of the paper responses will need recontact to the employers. This will result in a maximum of 3,465 recontacts. BLS estimates that 366 recontacts will be made to public sector employers for the QA program. At 10 minutes per establishment, this results in a total of 578 hours for the study – 517 hours for private employers and 61 for public sector.

Additional burden hours of 9,433 are requested for the collection of public sector data for survey year 2008. These burden hours are for the recording of the data by the employer. These burden hours are broken down as follows:

• 7,000 public sector establishments at 50 minutes each for recording the summary information = 5,833 hours

• 6,000 public sector cases at 14 minutes each for the log = 1,400 hours

• 6,000 public sector cases at 22 minutes each for the log = 2,200 hours.

The total burden hours request is for 10,011 hours.

IV. Payments to Respondents

The SOII does not provide any payments to the respondents.

U .S. Department of Labor Bureau of Labor Statistics

2 Massachusetts Ave., N.E.

Washington, D.C. 20212

Dear Employer:

The U.S. Bureau of Labor Statistics (BLS) thanks you for your recent response to the 2007 Survey of Occupational Injuries and Illnesses (Survey). The 2007 Survey continues to feature the use of electronic methods of data collection, including the Internet and a new form that can be completed by e-mail. As an ongoing effort to assure the quality and accuracy of our data for all methods of response – including electronic, paper form, and telephone - BLS has randomly selected your survey response for recontact. We would like to insure the quality and accuracy of the different methods of data collection for the 2007 Survey and to verify that our data accurately reflect what you have recorded at your worksite. The telephone call needed for this recontact should take less than ten minutes and any information you provide will remain confidential and will be used for statistical purposes only.

You will be asked questions regarding contacts by staff members of the State agency or BLS regional office responsible for collecting these survey data in your State. You may also be asked questions about changes made (if any) to your data from what your company may have originally reported.

The BLS will use the information you provide for statistical purposes only and will hold the information in confidence to the full extent permitted by law. In accordance with the Confidential Information Protection and Statistical Efficiency Act of 2002 (Title 5 of Public Law 107-347) and other applicable Federal laws, your responses will not be disclosed in identifiable form without your informed consent. The OMB control number for this survey is 1220-0045, expiring September 30, 2010. Without this number, the BLS would not be able to conduct this survey.

After the telephone recontact for this quality assurance survey is complete and if you filled out only one SOII Survey for a single location, you will not be recontacted by the BLS for the purposes of this quality assurance survey. If you submitted SOII surveys for multiple locations, we may recontact you about one of the other locations. It is also possible that someone from the State agency or regional office may contact you regarding your survey response for further data clarification purposes.

Your report and response to this quality assurance survey are very important to us. They provide us the information we need to produce reliable and accurate statistics of the number and rate of injuries and illnesses in the various industry categories and in turn allows establishments to compare their own injury and illness record with the average for their particular industry. These statistics will impact government policy and allow safety and health professionals to make informed decisions about workplace safety throughout the United States.

Again, thank you for helping us collect accurate information and for helping in the effort to make America’s work places safer and healthier.

John W. Ruser, Ph.D,

Assistant Commissioner
Office of Safety, Health and Working Conditions
Bureau of Labor Statistics

4