I currently conducted my pilot study, my total items is 70 but my total respondents are 50. So in this case, can I use PLS to do my pilot study or just conduct in SPSS?
Why not. As your respondents are more than 50. The minimum sample size is 50 in Smart PLS program.
Good luck for your research.
Hello Gangga,
You didn't indicate the type(s) of analysis you intend to conduct, so I'm not really sure what the competing model methods might be here.
Generally, 50 cases would be considered far too few to derive trustworthy estimates of relationships among 70 variables (= 2415 relationships or covariances, plus 70 variances, plus 70 means---if scale is important; that's a lot of parameters being estimated!). As well, I'm sure you understand that the assumptions and aims of the partial least squares method (as used in the SmartPLS program) are different from those of traditional least squares or maximum likelihood methods.
Good luck with your work!
Thank you for the answers.
- Dr Halima Begum. I do tried doing measurement model in pls which leads to losing more than 20% of items. So i kinda confused when i referred to some experts they mentioned that doing a measurement model in pls with low sample size (not to forget to mention I have 7 constructs), the results will be doubtful.
- Dr David Morse, I wants to do reliability test since it's my pilot. Since doing in pls seems hustles. will you advice me to do in spss ? can i test EFA and cronbach alpha?
Still you can do as you have seven constructs. Btw, your open ended questionnaire is reflective or formative? I believe that Smart PLs can work with limited data and constructs.
Good luck
- Badges
- Science topic
- Similar topics
- Instrumentation Engineering
- Instruments