1352.0.55.028 - Research Paper: Impact of non-response bias in business surveys (Methodology Advisory Committee), Nov 1999  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 01/11/2000  First Issue
   Page tools: Print Print Page Print all pages in this productPrint All
  • About this Release

About this Release

The paper "Low response rates and their effects on survey results" is the product of research undertaken to address a key issue within the Statistical Clearing House (SCH). Many non-ABS surveys suffer from low response rates and it is often difficult to convince survey managers of the importance in maximising response rates. Sampling error is usually addressed adequately, and it is believed that a primary reason for this is the relative ease in demonstrating the error introduced by taking samples as opposed to censuses. Demonstrating the effect of non-response bias is far more difficult, and this research was aimed to provide evidence of the deleterious effects on low response rates and consequential non-response bias.

There are primarily two issues we would like MAC feedback on this research.

An important aspect of this research is to achieve an effective outcome in demonstrating the deleterious effects on low response rates to a wide range of survey managers. The paper provides the basis for such a demonstration, but it now needs to be utilised to inform survey managers. We have made a start in this regard by presenting the paper (and the message) to the 28th National Conference of the Market Research Society of Australia in Adelaide, 3-6 October, 1999. We also intend to place the paper in our "Research documents" section of the SCH Web site (this section is yet to be developed). Other strategies for educating the survey manager community would be appreciated.

Secondly, there are a few limitations identified while conducting the research. These limitations are briefly discussed as an attachment to the paper. The views of MAC on these limitations and other limitations identified by MAC members would also be appreciated. MAC advice on alternative approaches to the problem of demonstrating the ill effects of non-response for business surveys would also be appreciated, as would any references to papers describing other relevant studies.