Survey Reliability & Instrument Calibration
10 Pages Posted: 1 Mar 2013
Date Written: February 27, 2013
Abstract
This purpose of this paper is to introduce researcher to the tool used to determine the reliability of a survey. A survey is a research instrument. As a research instrument, a survey is use to gather data. If the survey is defective, the data it gathered is said to be unreliable. Such unreliability leads begs the question of reliability and efficacy of the research. Since 1951, survey reliability measurement had been centered around the calculation of the Cronbach alpha. The Cronbach alpha is a calculation of correlation ratio between series of survey questions in relations to the total questionnaire score. While this inter-correlation ratio analysis allows the researchers to see the consistency of answer scores among questions in a survey, it does not answer the question of how reliable is the data. This section introduces the idea of population projection analysis as a tool for determining the reliability of the survey, and thus, answering the question of whether the research finding is reliable. The decision rule of the population projection analysis lies in the supposition that if the individual survey question is defective, the entire research instrument is also defective. Therefore, any data collected from such an instrument cast a shadow of doubt. Any conclusion drawn from a defective research instrument also begs the question of reliability and validity.
Keywords: Cronbach’s alpha, data integrity, dichotomous data, explanatory power, instrument
JEL Classification: C10, C12, C19, C42
Suggested Citation: Suggested Citation