Removing all bias from sample is almost impossible. However you can take steps to avoid sample bias and validate respondents through asking screening questions before interviewing commences.
|
Sample |
Probable Bias |
Reason |
|
Customers |
Favourable |
A random sample should be invited to participate in customer research. |
|
Lapsed Customers |
Favourable |
Important to understand why they are lapsed customers |
|
Prospects |
Favourable |
Important to understand what their requirements are to win them as customers |
The consequences of bias depend on the nature of the survey. For example, a survey for a product aimed at a retired respondent group will not be as biased by daytime interviews as will a public opinion survey. A survey about Internet products can safely ignore people who do not use the Internet.