Statistical Clearing House

Tables

• Table 1: Completed Reviews, Impact of Statistical Clearing House Review Process
• Table A1: Annualised Respondent hours
• Table A2: Completed Reviews

Introduction

1    This report describes the impact the Statistical Clearing House review process has had on the quality of Commonwealth Government surveys and on respondent burden during the period 1 July 2002 to 30 June 2003.

Return to top

Purpose of the Statistical Clearing House process

2    The Statistical Clearing House is responsible for reviewing all surveys involving 50 or more businesses conducted by or on behalf of Commonwealth Government departments and agencies. The primary purpose of the Statistical Clearing House is to reduce the burden of Commonwealth Government surveys on businesses by ensuring such surveys do not duplicate existing collections and are of sufficient quality to warrant the burden imposed.

3    A summary of the business surveys the Statistical Clearing House reviewed during the 2002-03 financial year is included in Appendix 1, and Appendix 2 summarises the business surveys specific to your agency, which were reviewed during this period.

Return to top

Impact of Statistical Clearing House review process

4    The Statistical Clearing House review process assesses surveys for potential improvements that would assist in reducing the burden on businesses and/or raising the quality of surveys being conducted. Where quality improvements will impact directly on the fitness for purpose of the data, the Statistical Clearing House requires that changes be made to the survey methodology. The Statistical Clearing House review also highlights where changes can be made to improve data quality incrementally.

5    Table 1 summarises the number of completed reviews where the Statistical Clearing House review process has impacted on surveys. Reviews completed in 2002-03 have resulted in the most improvements to surveys to date, with improvements made to 64 surveys and improvements suggested to 78 surveys. This increased impact is due largely to the higher numbers of Once Only Public Interest surveys and Program or Product Evaluation surveys being reviewed. Often these types of surveys benefit most from the review process as they tend to have been developed with the assistance of less experienced survey managers. Respondent burden was also decreased in 19 surveys.

Table 1: Completed Reviews, Impact of Statistical Clearing House Review Process

Impact of Review	1998-99	1999-00	2000-01	2001-02	2002-03
Respondent Load Reduced	9	10	19	6	19
Improvements Made	17	35	45	44	64
Improvements Suggested	39	35	61	51	78
Total number of reviews	141	119	126	122	169

6    Considerable effort is expended to ensure that the review process provides a learning opportunity for survey managers who are less experienced with survey design and development. The Statistical Clearing House provides information to survey managers to enable them to implement recommended changes based on best practice survey design. Common quality improvements made as a result of the Statistical Clearing House review process have included:

• changes to questionnaires
• changes to sample design methodology
• increasing follow-up procedures to improve response rates
• undertaking questionnaire testing
• addressing duplication with other surveys
• reporting data quality information to improve useability of data.

7    In 2002-03, changes made to sample design or questionnaire content saved 5223 hours of businesses time. 73% of this reduction was the result of one review where a census was changed to a sample survey. Other reviews still resulted in 1423 hours of respondent load reductions, representing a 79% increase in provider load savings compared with 2001-02 (794 hours).

8    Provider load reductions have also been achieved when Statistical Clearing House reviews resulted in surveys being cancelled in favour of qualitative research methods, such as focus groups. In 2002-03, four surveys were cancelled, which, if run, would have contributed approximately 1100 hours to respondent burden. In each of these cases, an assessment of the proposed survey methodology showed that data quality would be insufficient to meet the stated information needs, due to low response rates or inadequate frames. The Statistical Clearing House assists survey managers to identify when and how to use qualitative research methods appropriately. This ensures Agencies' research resources are efficiently and effectively directed towards achieving the required information outcome, without placing unnecessary burden on businesses.

9    Respondent burden has also been reduced indirectly through changes made to questionnaires during the review process which simplify the respondents' task. These include changes to question order which improve the flow, improvements to the format which make questionnaires easier to use and clarification of question wording and response categories.

10    During 2002-03, the Statistical Clearing House has maintained and developed relationships with the network of Survey Liaison Officers. Survey Liaison Officers are a primary communication link between the Statistical Clearing House and survey managers.

11    Survey Liaison Officers encourage Survey Managers to notify the Statistical Clearing House early in the survey development phase. As a result, there has been an increase in early involvement of the statistical Clearing House in the survey development process. These early discussions with Statistical Clearing House staff have resulted in most survey managers adopting best practice in survey design, without disrupting the survey schedule.

Return to top

Commonwealth Business Surveys Register

12    The Statistical Clearing House has continued to maintain the Commonwealth Business Surveys Register on the Statistical Clearing House website. The Commonwealth Business Surveys Register is provided for survey managers and researchers as a resource tool to increase awareness of the range of statistical data generated by these surveys, and to reduce the likelihood of survey duplication. It also acts as an accessible reference guide to existing survey designs. There are currently metadata relating to 717 surveys on the Register. For each of these surveys, there is a description of the survey design and contact details for the survey manager.

13    In 2002-03, the Statistical Clearing House has added a quality view to the Register which enables survey managers to assess the quality of survey data using an international framework. In addition, the Statistical Clearing House has introduced a new procedure which allows more timely access to information about Commonwealth surveys.

Return to top

Contact for Further Information

14    If you have any queries about this report or wish to advise the Statistical Clearing House of business surveys please contact the Statistical Clearing House by phone on 02 6252 5285 or by email at: statistical.clearing.house@abs.gov.au. Further information about the Statistical Clearing House can also be found on the Statistical Clearing House website.

Statistical Clearing House
July 2003

Return to top

Appendix A

15    Annualised Respondent Load

Table A1 shows that the number of surveys being notified to the Statistical Clearing House and the respondent burden placed on businesses as a result of these surveys have increased. Surveys conducted by the Australian Bureau of Statistics have contributed to over 90% (405,985 hours) of total annualised respondent load, (443,320 hours). The respondent load attributable to surveys conducted by other Agencies has increased by 38% in the last year from 27,124 hours to 37,335 hours. This is in large part due to the increasing number of surveys being notified to the Statistical Clearing House by these Agencies.

(a) Annualised load is an estimate provided by survey managers of the number of hours imposed on providers by surveys notified to the SCH in the year of data collection, regardless of the year in which the SCH reviewed the survey.
(b) Australian Bureau of Statistics.

Types of surveys reviewed.
Table A2 below shows the types of surveys reviewed by the Statistical Clearing House over a six year period. Of particular note is that whilst the number of Public Interest surveys reviewed has reached a plateau, the number of Once Only Public Interest surveys has increased more than three and a half times from 15 in 1998/1999, to 53 in 2002/2003.

(a) Large surveys are surveys which planned to involve 2,000 or more respondents.
(b) Small surveys are surveys which planned to involve fewer than 2,000 respondents.

Return to top

(approx 89kB)

The complete document in Portable Document Format to view online or download.

Return to top