Discussion: Variable Definition and Measurement

undefined

There are two familiar sayings in the research community. First, We live in a multivariate world, so we need to conduct multivariate research, and second, garbage in, garbage out. Thefirst statement is a reflection of the normal state of the humancondition, in which multiple factors usually influence a givenphenomenon. Stated in research terms, at least several variables arerelated to the primary variable of interest. How, then, do you knowwhich variables to study? For example, when conducting depressionresearch with the elderly, some of the highly related variables thatalso should be included in a primary depression study might includephysical health status, socialization, stress, coping, and access toresources. No one variable on its own can explain depression in thispopulation.

The second statement—garbage in, garbage out—iscommon in the quantitative statistical community and isself-explanatory. If the quality of the measures and collected data arepoor, every action and result that follows will also be poor, thusaffecting reliability, validity, and credibility in a negative way.Optimally designed and executed survey studies prioritize measurement.They are reliable, meaning measurement is consistently applied. They arealso valid, meaning truth and meaning in the measurement is applied.Finally, they are credible, meaning the results are subjectively, aswell as objectively, believable.

Refer back to the example in the Introductionfor this week. Using a depression scale that consistently measures thespecific construct of depression similarly with multiple measurementpoints and samples strengthens reliability. Using an instrument thatmeasures the construct of depression—and not a similar construct such asgrief, stress, or anxiety—strengthens validity. Ensuring that yourresults relate to similar findings supported in the literature as wellas within the community that works with depression in older adults alsostrengthens credibility.

For this Discussion, use your topic from Week 3,and consider three key variables in the research study you proposed forthat topic. Think about how you would operationally define and measureeach, using theory as the basis of your decisions. In addition, reflecton how you would justify your choice of measurement, including specificscales or measures you would use, and why. Finally, consider how amultivariate approach to variable selection and measurement wouldoptimally assess your research problem.

With these thoughts in mind: