Measuring psychological constructs

PDF
Publication

FORS Guide Nº 22

DOI
https://doi.org/10.24449/FG-2023-00022
How to cite

Ryser, V.-A. (2023). Measuring psychological constructs. FORS Guide No. 22, Version 1.0. Lausanne: Swiss Centre of Expertise in the Social Sciences FORS. doi:10.24449/FG-2023-00022

Abstract

Over the past years, psychological constructs have become increasingly important also for researchers outside the field of psychological research. This guide will provide an overview of the most used as well as the most useful psychological constructs in large scale surveys and will discuss the implications and difficulties related to psychological scales namely their length, validity, and reliability, the use of the same questionnaires across different data collection modes, as well as their translation for the Swiss context.

Recommendations

  • Before introducing psychological constructs in a large panel survey, the question of the transversality of the dimension to be introduced should be debated. To what extent is the dimension based on a well-accepted psychological framework and might be useful in various fields of research? In addition, in longitudinal surveys, such as household panels and some large interdisciplinary surveys, it is better to introduce such constructs from a long-term perspective. Consequently, care must be taken in the choice of scale(s) and the wording of the items.
  • An important question is whether a short scale’s profile matches the necessities of the assessment setting in which it is applied and whether the short scale corresponds to the survey’s target population.
  • Quantitative psychological constructs have been mainly developed for the CAPI, PAPI, and CAWI modes. Are the scales of interest compatible with the data collection mode of the survey in which the construct should be introduced?
  • Long and complex questions in general, but in particular for CATI, must be avoided as well as negations in questions. Simple, efficient, and straightforward wording of questions must be favored. In addition, surveys based on general random samples of a population comprise individuals from the whole spectrum of the social strata, including individuals with a migration background. All these subpopulations need to master the written or oral questionnaire language in a basic way. The wording of the questions is a key issue regarding high-quality data.
  • Try to use short and validated standardized scales. Many psychological constructs have been introduced in Swiss samples of international surveys. Examining what has been done and exists in the Swiss context might help researchers select a scale that has demonstrated its value. In case of the need for an adaptation of the scale in Swiss-German, cognitive testing should be systematically undertaken. The adaptation of the German version into Swiss-German is, however, hampered by the fact that some questionnaires have a copyright. It is not possible to make even minimal changes to this type of questionnaire. Without any adaptation, such scales are often not suitable for the Swiss-German context.
  • Try to base questionnaires on harmonized scales to allow for international comparisons.
  • For longitudinal surveys, the data collection modes are likely to change over the waves. Similarly, some surveys combine several data collection modes simultaneously, which means that the choice of questionnaires must be compatible with these various data collection modes.

    • Copyright

    © the author 2023. This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0)

    Publication year

    2023

    Data Sharing in the Social Sciences
    Quantitative data anonymisation: practical guidance for anonymising sensitive social science data