When is a Sample Size too Small?
The issue of small number of respondents becomes apparent when combining low incidence and small target samples. Low incidence and/or small sample sizes may result in limited data for analysis. A minimum sample size of 40 is considered stable and usable information.
"Without an adequate number of respondents, the ability to break down the data and make meaningful conclusions is compromised".
Samples of low-incidence populations are a cause for concern. It is important to understand what approaches are available to deal with this issue.
The following are key recommendations and insights when working with small samples:
- Be cautious when conducting analysis on a small group of respondents.
- A small number of respondents does not allow you to make sound decisions or conduct detailed analysis.
- When you have only a few respondents it is possible to make some conclusions (e.g. the group is less likely to occur in the population).
Estimated Number of Respondents Based on Sample Size and Reach
|Target Sample||Readership Reach|
For more information on sample size and related topics read [ Sample Size Booklet. ]
- Newspaper remains an effective vehicle for advertisers
- Who is Online? A Profile of Canadian Internet Users
- Canadian Internet Project Findings supports NADbank Study