Originally Posted by raphael
From your answer, I understand that I was overzealous.

As you say, developing a new tool takes quite some effort (before it reaches a point where it is truly convincing). Maybe this would be more for a PhD project. Especially given that the data I currently have is not suited for my idea, I would need to go look for other data sets..

Thanks for answering!

I have a meeting appointed with a supervisor on monday so I can discuss further things

Apologies for the delay before my response.

I don’t think you’re overzealous at all. I think it would be fantastic if your research were to lead you to branch out successfully into new areas of statistics.

My comment was to urge you to evaluate your findings with whatever appropriate statistics tools are currently available, because if you do find a tool that is appropriate but not optimal, a bird in hand is worth two in the bush and it may be a better path to publish findings and interpretations (if you believe the conclusions are valid) and then return later to this area, with the perspective of validating your new statistical tool (which may take a while if you have to negotiate with other researchers to obtain their raw data sets). Then you could eventually have several birds securely in hand.

From your previous exchanges in this thread, I think one of the themes which definitely apply is ‘done is better than perfect’.

How did your meeting go?