Related Experiment Videos
Evaluating the results of mail survey research
1Clinical Investigation Regulatory Office, Ft. Sam Houston, Tex. 78234-6125, USA. LTC_Donald_L_Harrison@smtplink.medcom.army.nil
Objective:
To present overall guidance enabling the reader to critically evaluate survey research conducted via mail questionnaire. Also considered are the sources of error that may operate during the conduct of survey research and the threats they pose to both the internal rigor of the study and the generalizability of the results.
Study Selection:
Not applicable.
Data Extraction:
Not applicable.
Data Synthesis:
Not applicable.
Conclusion:
Survey research has been used so commonly and for such a wide range of purposes in a large number of fields, that some people may mistakenly conclude that surveys are easy to conduct. Survey research involves much more than generating a set of questions to ask a group of subjects. To draw valid conclusions about the topic of a survey, the pharmacist must be prepared to evaluate the results in light of how well the investigators approached four sources of error--coverage, sampling, measurement, and nonresponse. The evaluation of survey administration procedures primarily concerns determining whether or not a pilot study was conducted, whether proper statistical analyses were conducted, and whether multiple mailing procedures were used. Failure to account for these sources of error, or inappropriate survey administration, could seriously affect both the generalizability and validity of the results.