Karl (name dropper) :) 
I do agree that p is a probability distribution and is continuous. However, I think 
the correct reply to the student is also dependent on several things. First, how was 
the proposal written. If it was clear that you choose to establish a p <.05 criteria 
(and you had good reasons, I assume) then shouldn't you stick to that. Second, it 
would depend on how you wrote it up. If you consider power and discuss effect size 
etc. then I agree with Karl (and with Cohen). Some of how you'd do that would depend 
on the amount of research already published in the area of research, standards of 
publishing in the area, and consequences of error (as always). I think you might also 
share the "Fishscale Model of Omniscience" (Campbell) with your student. It includes 
some notions in addition to the overlap of disciplines that might help your student 
think it through as they write up the drafts. I think it is accessible even to most 
advanced undergrads and is a good place to start (though power and p aren't per se 
main topics). 
Tim Shearon
 
___________________________________________
Albertson College of Idaho
Caldwell, ID 83605
 
[EMAIL PROTECTED]
Teaching: History and systems, physiological, Intro to Neurpsychology, 
psychopharmacology (and Winter term courses in Blues and Film Theory)

        -----Original Message----- 
        From: Karl L. Wuensch [mailto:wuenschk@;mail.ecu.edu] 
        Sent: Mon 11/11/2002 9:45 PM 
        To: Teaching in the Psychological Sciences 
        Cc: 
        Subject: p is continuous, not dichotomous
        
        

        Last I checked, the significance level, p, was a probability (the
        conditional probability of obtaining results as more discrepant with the
        null than are those in the current sample), and probabilities vary
        CONTINUOUSLY from 0 to 1.  At least that is what Jack Cohen told me.
        
        I suggest that we simply treat p as a measure of how well the data fit with
        the null hypothesis.  P = .08 is very poor fit, p = .04 is not much poorer,
        and p = .80 tells me that we got just about what we would expect were the
        null true.
        
        Karl W.
        

<<winmail.dat>>

---
You are currently subscribed to tips as: [EMAIL PROTECTED]
To unsubscribe send a blank email to [EMAIL PROTECTED]


Reply via email to