I have a question regarding a small numbers of respondents in my model.
In my case I differentiated between three different types of attitude as an depended variable. Two of them were mentioned really often (more than 60 % of the respondents choose them), the other one was only chosen by less than 2 %. Hence, I decided to consider it not anymore because the quality criteria were not good enough. Finally, my model has only two different attitudes. Is there any statistic paper which supports the theory that to small sample can not be considered further (my behavior).
Frequently asked questions about PLS path modeling.
1 post • Page 1 of 1