convenience sampling and p-values

[purifyz]

Dear all,
I am in the middle of my research. I gathered data (via convenience sampling). I intend to run PCA/EFA and then use dimensions in logistic regression model. Since my sample is nonprobabilistic I can not formulate any inferences about population.
My question is - what do I do with p-values? Do I report them? Should I "care" if my model is insignificant?
I think that p-values only make sense with probability sampling...
Thanks in advance for any input.

 

[rogojel]

hi,
I think you are right about the p values, but I guess you should report them anyway with the appropriate caveat.
What is research question, and how can you answer it with an opportunistic sample?

regards

 

[purifyz]

Well, my main research question is how older customers perceive loyalty programs. As this research is part of my MSc thesis, I do not have resources to get representative sample (hardly a justification). I chose opportunistic sample because it's more of exploratory than confirmatory research.