In questionnaire data analysis, if one as a statistician received data after designing and filling the questionnaire by respondents, and encounters pre-determined groups of questions that are claimed each are measuring common aspects of the research issue, how can we statistically examine or reason the correctness of this let's say partitioning.
Actually, I need a numerical index that somehow shows why the set of questions has to be selected to measure certain unobserved variables.
I know my answer partly is related to designing the questionnaire and background information in the scientific field and but I mean after filling the questionary how is it possible to assess if the variables measure what they proposed. Does my question have something to do with Cronbach's alpha?
alpha.
Many thanks.
Hamid
hi,
Generally on the surface you can use data analyses like normality test (deciding to use parametric / non-parametric statistics), descriptive statistics, reliability test (Cronbach Alpha / Composite Reliability), Pearson / Spearman correlational test etc.
Quality Criteria of a Survey for Market Research
- Objectivity. Data Collection. Evaluation. Interpretation.
- Reliability.
- Validity.
for more info:
https://home.ubalt.edu/ntsbarsh/business-stat/stat-data/Surveys.htm
Best wishes..
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology