Measurement issues in quantitative research

Measurement is central to empirical research whether observational or experimental. Common to all measurements is the systematic application of numerical value (scale) to a variable or a factor we wish to quantify. Measurement can be applied to physical, biological, or chemical attribute or to more complex factors such as human behaviors, attitudes, physical, social, or psychological characteristics or the combination of several characteristics that denote a concept. There are many reasons for the act of measurement that are relevant to health and social science disciplines: for understanding aetiology of disease or developmental processes, for evaluating programs, for monitoring progress, and for decision-making. Regardless of the specific purpose, we should aspire that our measurement be adequate. In this chapter, we review the properties that determine the adequacy of our measurement (reliability, validity, and sensitivity) and provide examples of statistical methods that are used to quantify these properties. At the concluding section, we provide examples from the physical activity and public health field in the four areas for which precise measurements are necessary illustrating how imprecise or biased scoring procedure can lead to erroneous decisions across the four major purposes of measurement.