Im currently doing a psychometric evaluation of Ryff's scale and my professor would want me to do an interview and see if the respondents' answers support the dimensions of the theory. Is it logical since psychometric evaluation generally gets the reliability and validity of the scale. Thanks!
I am a qualitative researcher. As I see it, no project that is tightly bound to quantitative research can ever see the rounded picture. Presumably you have done a qualitative assessment of the potential research area before starting? Even if you have, in my personal opinion, you need qualitative date (interviews are good!) to place your research results in a wider context. My field is military anthropology. If I were researching the use of the formally endorsed disciplinary system I could simply collect statistics about how many times formal disciplinary action was taken, and of what type (the options are administrative versus legal). However, I would miss the nuances of involved in taking the decision about what action to take, the relationships and negotiations between the authority figures, the effect of membership of different regiments/corps and so on.
In essence, for me at any rate, scale is not enough.
It would be unusual to do a qualitative evaluation of a scale that is as widely tested and used as Ryff's. If you feel required to do so, however, you might consider cognitive interviewing as a technique, since it is a version of qualitative research that is explicitly designed to evaluate survey items.
A good source of cognitive interviewing is willis, Cognitive Interviewing: A Tool for Improving Questionnaire Design.
Adding a qualitative part can be useful for at least two reasons:
1. To partly compensate for the shortcomings of self-reported scales
2. To know more about the reasons underlying the respondents' preferences and choices which can be reported in the discussion section of your work
There are other ways that we nearly "qualitatively" add to our psychometric analyses.
If we're to ask fellow researchers whether they believe the items cover the construct of interest, even if we use rating scales, we're using their judgement and while this isn't the same as qualitative data, it's traveling a similar path.
We'd call this a content analysis (Carpenter & Arthur, 2013), the link will get you a really bad scan of the chapter in which content analysis is discussed about halfway through.
http://people.tamu.edu/~w-arthur/611/Journals/Carpenter%20&%20Arthur%20%282013%29%20chpt,%20work%20performance.pdf
Hi researcher,
There is a little secret about using mixed methodologies in a research as the qualitative research will explore new dimensions about the topic studied in more depth compared to the quantitative method. Moreover, combining qualitative method with the quantitative one will assist you about whether you are on the right track in your project or not especially when you use semi-structured questions in the qualitative part.
Good luck,
Even though many attributes of psychometric scales - such as validity, reliability, internal consistency, and factor structure (dimensions) - can be represented as hard numbers, some elements - such as face validity and construct validity ultimately include subjective judgements that are at least in part qualitative. Do all of the questions on a scale, for example, appear to relate strongly to the construct which you are attempting to measure? If a factor analysis has been performed and certain dimensions are suggested, then are these dimensions correctly interpreted and labelled, or might there be alternative concepts accounting for the observed factor structure?
Including a qualitative dimension to your evaluation of Ryff's scale, therefore, is perfectly logical as it will enable you to more thoroughly appraise the dimensions which this scale purports to measure. People using it might have different interpretations of the questions being asked than their intended meaning, and a different idea about what the scale is measuring. If you have the opportunity to examine this, then you should certainly do so. The results will be richer and more informative and make for a more publishable report.
- Badges
- Science topic
- Similar topics
- Psychology
- Psychometrics