Supporting Statement: CES Confidence Interval Chart Survey

OMB Approval No. 1225-0059

CUSTOMER SATISFACTION SURVEY AND CONFERENCE EVALUATION CLEARANCE FORM

A. SUPPLEMENTAL SUPPORTING STATEMENT

A.1. Title:	CES Confidence Interval Chart Survey
A.2. Compliance with 5 CFR 1320.5: Yes ___X__ No _____				A.3. Assurances of confidentiality: No pledge of confidentiality will be given.
A.4. Federal cost: $ 5,922 (90 hours of BLS employee work)				A.5. Requested expiration date (Month/Year): 08/2016
A.6. Burden Hour estimates: a. Number of Respondents: CES homepage: 300 Customer email response: 100 CES NR Release Subscriber: 200 Total = 600 b. Frequency: One-time c. Average Response Time: 5 minutes d. Total Annual Burden Hours: 50 hours				A7. Does the collection of information employ statistical methods? ____X___ No Yes (Complete Section B and attach BLS review sheet).
A.8. Abstract: In an effort to graphically display the reliability of the Current Employment Statistics (CES) survey estimates, the Bureau of Labor Statistics would like to provide its users with a single chart showing the 90-percent confidence intervals for each major industry sector. Significant over-the-month changes are calculated at a 90-percent confidence level. The standard error is used for a 1-month change. The 90-percent confidence interval represents the symmetric range of values around the estimate for which there is 90-percent probability that the actual over-the-month change is contained within that range of values. The variance for total nonfarm employment is an approximation because government and rail transportation are not based on a probability sample. Federal government and rail transportation are essentially full population counts and do not contribute to variance. Data users’ reactions to two alternative graphical approaches for presenting the confidence interval data will be compared. Both charts (see Attachment 1) show the over-the-month change in employment for total nonfarm and major industry sectors (represented by the red dot) and the range for the 90-percent confidence interval (represented by the blue bar). In Chart 1, the confidence interval is computed around the mean over-the-month change for an industry. If the over-the-month change is statistically significant, the blue bar will not cross the black zero line. The red dot representing the over-the-month change lies on the middle (mean) value of the blue bar that shows the confidence interval for the specific industry. For example, the over-the-month change for construction is 18,000 with a 90-percent confidence interval of plus or minus 25,900 so the range is -7,900 to 43,900. These figures do not mean that the sample results are off by these magnitudes, but rather that there is about a 90-percent chance that the true over-the-month change lies within this interval. Since this range includes values of less than zero, we could not say with confidence that construction employment had, in fact, increased that month. For manufacturing, the over-the-month change is 29,000 with a 90-percent confidence interval of plus or minus 18,700 so the range is 10,300 to 47,700. In this case, it is likely (at least a 90-percent chance) that manufacturing employment had, in fact, risen over the month. In Chart 2, the confidence interval for each industry is computed assuming that the mean over-the-month change is zero. So if the over-the-month change is statistically significant, the red dot falls outside of the blue bar (in a positive or negative direction). Based on past experience, we assume that one percent of the visitors to the CES homepage will compare the graphical approaches and take the survey. Therefore, we expect that 300 people will complete the evaluation survey over a two month period. The evaluation questionnaire will take an average of 5 minutes to complete, resulting in 25 burden hours (300 x 5 min divided by 60 min). The CES program will also ask a convenience sample of data users who have sent CE staff an email question to compare the two graphical approaches. This will be done by including a survey-invitation link at the bottom of email responses. We hope to obtain 100 responses using this approach. Finally, we will send a survey email invitation to CES NR news release subscribers in the hope of obtaining another 200 respondents. These latter two approaches will result in an additional 25 burden hours (300 x 5 min divided by 60 min), and a total of 50 burden hours. All survey responses will be collected using SurveyMonkey, and no pledge of confidentiality will be given. We are only using the results of this survey internally for planning purposes. We are not employing statistical methods because we don’t have a sample frame of all data users who visit the CES website and, therefore, can’t extrapolate the results to all data users. A copy of the survey is attached.
Program Official		Date	Departmental Clearance Officer		Date

B. SURVEYS AND EVALUATIONS EMPLOYING STATISTICAL METHODS

B.1

Sample Frames:

Limitations of the Sample Frames:

Expected Response Rates:

B.2

Sample Size:

Sample Allocation:

Sample Selection:

Procedure for Sample Selection:

B.3

Methods to Reduce Non-Response:

Survey Distribution Procedures:

B.4

Test of Procedures:

B.5

Name	Agency/Company/Organization	Number Telephone
Bill Mockovak	DOL/BLS/OSMR	202-691-7414
Karen Ransom	DOL/BLS/CES	202-691-7625