Nonsampling errors

Survey, clinical trial, and experimental data may contain errors as a result of data collection and data processing mistakes, nonresponse and missing data values, the biasing influences of interviewers and clinicians, or respondent memory problems. Nonsampling errors are often more important than sampling errors in the analysis and interpretation of results.

ARCPOH researchers have made substantial efforts to increase response rates through better survey and experimental design, improved interviewer training and supervision, and the use of incentives and more private data collection modes. ARCPOH's statisticians employ various procedures to statistically control for biases introduced by sampling limitations or necessities.

ARCPOH can evaluate nonsampling error by employing designs and statistical methods for comparing alternative modes of data collection