Survey Research

You want your employee performance surveys, public opinion polls, and online market survey studies to produce valid data.  But what does validity mean exactly in the field of survey research?  Although there are actually many different types of validity, in general, designing valid survey instruments means that survey questions adequately measure (in reality) the concepts the researcher is attempting to measure.  As opposed to reliability, which refers to producing the same results with repeated measurements, validity involves designing legitimate, suitable survey questions that accurately address the ideas you’re seeking to evaluate. 

Survey researchers try to attain high levels of external and internal validity. External validity is the degree to which your study’s results are generalizable to individuals other than those in your survey sample.  If your results are externally valid, you can come to increasingly accurate conclusions regarding larger and different populations (people/markets across various time periods, geographical locations, etc.).  Obtaining external validity is a major goal of survey and market researchers, because most research is performed using a small sample of people from a larger population or target market in order to decrease costs and lessen workloads.

Internal validity, on the other hand, centers around estimating how valid your measures are based on an inward look at the research design and any other variables that might interfere with accurate results and conclusions.  To increase internal validity, survey researchers often like to implement experimental designs, as well as perform statistical analyses that control for variable effects.  Examining any additional measures and proposing alternative explanations for results both go a long way in judging the validity of any survey research project.