I was wondering whether it would be possible to measure the same construct (i.e. the perception of training) in two samples by using different scales for each sample (i.e. employed and unemployed people)?
Thanks,
I would give the both scales to the both groups and then calculate how well they correlate. What you are doing depends on your research questions.
well .. I want to compare whether there is a difference in the construct when it comes to two samples. However, due to the nature of the measurement, some questions are more appropriate or required in one sample and not in the other .. so this is my concern mainly.
To compare things and obtain a VALID SCIENTIFIC result, not only should you have equal scales for both groups, but you should also prove that they are really equal. Otherwise, things cannot be compared. Or, as Béatrice Marianne Ewalds-Kvist, give both scales to both groups and then calculate how well they correlate. Thus it would be quite obvious which questions are more or less appropriate regarding each sample.
You could also convert the resulting values to z-scores and compare those. However, you have to ensure that both scales truly measure the same thing, and that their mins and maximum values are reflecting the same degree of "Perception of Training" in both scales. Max in Scale 1 = 100 (perfect perception), Max in Scale 2 = 100 (perfect perception).
But certainly run a collinearity check.
I'm not understanding how you end up in a situation where the "Employed" and "Unemployed" groups are giving you different scales on a common assessment. Did you specifically create one survey for employed people and another survey for unemployed people?
Hello, this is just an idea whether i could do it .. i know that most people avoid it .. so i am trying to get different opinions on this matter .. but training perception of emoloyed could be related to the company while for unemloyed could be more general?
This may sound remedial, but the best way to think about your idea is to clarify, what is the hypothesis? This question is key, because the methodology we (as researchers) use in any given project is specifically motivated by the kind of data that is needed to test the hypothesis. In fact, the research hypothesis will be so well defined that we already know how the data will be analyzed before any data are collected. So, start with a very clear (i.e., empirical and testable) research question, and then everything after that will fall into logical place for you.
No , it will be a wrong method. Have the same instrument for both the types of samples. The dimensions may be different in the two instruments, this will dilute the answer
- Badges
- Science topic
- Similar topics
- Organizational Psychology
- Training