As a research scholar it is my observation that generally scholars use scale or index for measuring a construct or a concept in social science research. When asked why did u use scale why not index? for a particular construct/ concept under study or vice-versa, almost everyone couldn't answer me. So I am interested to know those conditions for using or developing a scale /index. Kindly throw some light on this question.
Integrative validity framework (See, Embretson 1983, 2006, 2017)
Hope this helps,
Matthew
All scales are indices but the reverse is not true. This might be helpful:
http://www.analytictech.com/mb313/indices.htm
Multi-item scales typically have higher reliability than single-item measures. Higher reliability essentially means less random error, so these scales have higher potential correlations with other items. So, any time you think all your separate items measure a single underlying construct, you would be wise to create a scale.
I think there is more of social desirability and usage in scales and for indices preference, it is a more standardized abstract. Indices are more relevant in structural or factorial and predictive validity. For scales, they may or may not necessarily be that. Scales have more room for bias than indices. So if the study is intended to get more competence/ability based results rather than self report tests and results.
- Badges
- Science topic
- Similar topics
- Social Science