Examination of variable errors of measurement in a survey-based social indicator

The sample survey, involving the use of specially developed questionnaires and instruments, is an important source of data regarding individual experiences and social-psychological characteristics for use in social indicator research. Variable errors of measurement are often given only cursory attention in presentations of survey sampling and social indicator theory and in the analysis of such data. These errors may, however, result in reduced precision of estimates and may affect statistical inferences drawn from the data. Some procedures for assessing the extent of variable measurement error are described and examples of their use in evaluating a survey instrument and data collected are discussed.