Low Response Rates and Their Effects on Survey Results
Views expressed in this paper are those of the author(s) and do not necessarily represent those of the Statistical Clearing House. Where quoted or used, they should be attributed clearly to the author.
Fogliani, M. (1999) Low Response Rates and Their Effects on Survey Results, Methodology Advisory Committee paper November 1999 meeting.
This paper has two main aims. The first aim is to demonstrate the factors which influence non response bias and the relative magnitude of these influences. Simulated data is used to identify possible influential factors and demonstrate their effects on survey estimates. To do this, a plausible non response model was postulated, and a number of parameters were varied. These parameters represented factors that might be considered to have an effect on the magnitude of the non response bias. These were the sampling fraction, population standard deviation, response rate and population distribution. These factors were examined to determine the circumstances under which survey managers should be particularly wary of non response bias effects.
The second aim is to illustrate, using a real life case, how low response rates and any subsequent non response bias can affect the quality of survey results and to clarify why it is important to dedicate time and resources into trying to increase response rates. This is done through a case study of data obtained from the Year 2000 (Y2K) survey conducted by the ABS in November, 1998.
This paper first presents the data sources and methods used in demonstrating and evaluating non response bias effects. The results of this evaluation will then be discussed with some conclusions and recommendations made based on the results of the analysis undertaken.