On the sins of short-form development.

Abstract

The empirical short-form literature has been characterized by overly optimistic views of the transfer of validity from parent form to short form and by the weak application of psychometric principles in validating short forms. Reviewers have thus opposed constructing short forms altogether, implying researchers are succumbing to an inappropriate temptation by trying to abbreviate measures. The authors disagree. The authors do not oppose the development of short forms, but they do assert that the validity standards for short forms should be quite high. The authors identify 2 general and 9 specific methodological sins characterizing short-form construction and offer methodological suggestions for the sound development of short forms. They recommend a set of 6 a priori steps researchers should consider and 9 methodological procedures researchers can use to develop valid abbreviated forms of clinical-assessment procedures.

Confirmatory factor analysis of the Frommelt Attitude Toward Care of the Dying Scale (FATCOD–B) among Italian medical students

“…However, when attempting to measure a multidimensional construct, the content coverage of each dimension must be preserved. Each item deleted causes a loss of content and construct validity and exposes researchers to the risk of that the abbreviated scale does not measures the construct intended in the original scale (Smith et al, 2000). Furthermore, every data driven respecification of the model might result in attempts to solve, a posteriori and by means of statistical analyses, the problems regarding item wording and construct definition.…”

Psychometric Properties of the Abbreviated Perceived Motivational Climate in Exercise Questionnaire

“…Although applied researchers may understand the necessity for shortened instruments to better entice potential participants, the practice of developing short forms is not without critics (Marsh et al, 2005;Smith et al, 2000). Therefore, in an effort to avoid unnecessary truncation of existing instruments, Marsh et al (2005) proposed a number of guidelines to follow when creating a short form, several of which are used to frame this discussion.…”

“…Marsh and colleagues (2005) go on to suggest several guidelines that must be met for the short form to be a viable alternative instrument. Their recommendations expanded on those provided by Smith, McCarthy, and Anderson (2000) on the development of short-form evaluation tools. Marsh and colleagues (2005) suggested that items should measure the same factors on the long and short forms, the responses to both the short and long instruments should be stable over time, and age and gender differences should be consistent across both versions.…”

Validación en Chile de la escala de sobrecarga del cuidador de Zarit en sus versiones original y abreviada: corrección

“…Sus resultados ponen un referente importante para la calidad métrica del EZA, pues sus resultados parecen tener un buen nivel de validez y salvar tiempo en su uso 4 , ambos aspectos valorados en la investigación y en la práctica profesional.…”

“…El problema ocurre con la correlación lineal estimada entre la versión larga y la versión breve elegida para la validación (la correlación fue r = 0,92), resultado que demostraría la validez de criterio de la versión breve debido a la magnitud de la correlación. Sin embargo, el puntaje de la versión breve contiene ítems que están contenidos en la versión completa y ello produce que las correlaciones sean sobre-estimadas artificialmente 4,5 . El resultado así obtenido es una correlación espuria que en parte está influenciada por los elementos comunes que ambas versiones tienen y, por otra parte, por la verdadera covariación entre las variables (puntaje de la versión larga y corta).…”