Coming soon · Book Profile
Lib42a60ac68e311ec0
A concise introduction to measurement theory that explains how social scientists can assess whether their empirical indicators reliably and validly represent abstract theoretical concepts.
A profile of this book is on the way.
What it’s about
This classic monograph demystifies the twin pillars of sound social science measurement: reliability and validity. Written for researchers in sociology, political science, psychology, and beyond, it moves from the fundamental problem of linking abstract concepts to empirical indicators, through the three basic types of validity (criterion-related, content, and construct), into the logical and statistical foundations of classical test theory, and finally to four concrete methods for estimating reliability plus the correction for attenuation. Using accessible examples like Rosenberg's self-esteem scale and requiring only knowledge of simple correlation, Carmines and Zeller give practitioners a rigorous yet lucid toolkit for building and evaluating measurement instruments, including an advanced appendix on factor analysis. It remains the perfect starting point for anyone who wants their measures to withstand scientific scrutiny.
The through-line
- Who it’s for
- A social science researcher or student who wants their measures of abstract concepts to be scientifically credible and defensible.
- The problem
- Their empirical indicators may not consistently or accurately represent the theoretical concepts they intend to study. They feel uncertain whether their research conclusions rest on solid measurement or on flawed, error-ridden instruments.
- The plan
- Understand measurement as linking abstract concepts to empirical indicators and distinguish random from nonrandom error.
- Evaluate validity using criterion-related, content, and especially construct validity within a theoretical context.
- Learn the classical test theory model and the concept of parallel measurements.
- Estimate reliability using the retest, alternative-form, split-halves, or internal consistency (alpha) methods.
- Correct observed correlations for attenuation using reliability estimates and report reliability transparently.
- The payoff
- The researcher builds measures that produce consistent results and behave in accordance with theoretical expectations. · Their findings withstand scrutiny because measurement quality is documented and defensible. · They can distinguish true relationships from artifacts introduced by measurement error.