Sample is 130. How do you suggest to look at both reliability and validity? Would appreciate more insight into the concepts and calculations. Let's say if both are weak after the results? Then what happens?
The usual way to calculate reliability with Liker-scored items is to use Cronbach's alpha. If the items file to form a single scale, then you should consider Exploratory Factor Analysis.
Because there have been so many questions here about LIkert scales, I have compiled a set of resources on this topic:
https://www.researchgate.net/post/How_can_one_help_people_create_simple_scales_from_Likert-scored_items
I would recommend you to read COSMIN guidelines for any research on clinimetrics. You may follow the website www.clinimetrics.nl for all the resources, checklist, manual, and book (measurement properties in medicine). It is the current reference standard.
Briefly compute the following:
1. Internal consistency using Cronbach's alpha
2. Test-retest reliability using ICC
3. Smallest detectable change and standard error of measurement
4. Construct validity using hypothesis testing, factor analysis
5. Responsiveness.
I hope this helps.
Reliability can be easily determined with Cronbach's alpha and intraclass correlations (ICC). You can apply the latter method for test-retest assessment at the item level. Alternatively, applying Pearson's correlation to the correlation of the summed scale scores also works. The r-values should be comparable in both cases. The central challenge lies with ascertaining the validity of your questionnaire. You need to relate/correlate your scale scores with other surveys or behavioral measures that assess cultural group differences. Ideally, you would want to see some cultural group scoring high on one measure and low on the other and perhaps vice versa for another cultural group. In other words, you need to prove that your questionnaire has discriminant and non-discriminant validity with regard to all cultural groups under investigation. This should not be difficulty, presuming that you surveyed your participants with other questionnaires or have some culturally sensitive DVs at hand.
- Badges
- Science topic
- Similar topics
- Psychology