Re: Experimental Psychology-Some dangers?

[Marie Helweg-Larsen]

Like other TIPSters I make it clear that results can be interesting whether they are significant or not. I also give much anecdotal examples of my own difficulty in finding significant results over the years (plenty of those!). However, I also explain how serious of a violation of academic integrity it is to invent data and describe the horrible circumstances that may result from such an action (e.g., getting dispelled).
My point is that students may make up data (consistent with their hypotheses or not). So I think emphasizing that significance is not "required" will not prevent all students from making up data. It is also necessary to add it to the list of "academic integrity violations" that include cheating, plagiarism, etc. and explain why it is not a "small deal" to invent data.
Marie

don allen wrote:

Hi Tricia-

I think that you have asked a very important question. In my Methods course
I make a major point of explaining that "no difference" and "did not support
my hypothesis" are NOT the same as "no meaning" and "no value". Two examples
that I use to drive the point home:

1. I hold an eraser above the desk. I hypothesize that if I let go of the
eraser then gravity will cause it to fall toward the earth. I then ask the
class to think about which would be the more interesting finding, a. one
that confirms the hypothesis or b. one that disconfirms the hypothesis.

2. I cite a study of the nutritional analysis of meals prepared by a.
mothers working outside the home and b. stay-at-home mothers. The authors
report that there was no significant difference in the nutritional quality
of the meals. I then ask the class to think about whether such a study would
be of no interest to a working-outside-the-home mother who was feeling
guilty about her conflicting roles.

I think that these examples help students to understand that the data are
ALWAYS meaningful. Personally, I think that it is often the most meaningful
when it disconfirms our hypotheses.

All the best,

-Don.
----- Original Message -----
From: "Tricia Keith-Spiegel" <[EMAIL PROTECTED]>
To: "Teaching in the Psychological Sciences" <[EMAIL PROTECTED]>
Sent: Friday, October 11, 2002 8:15 AM
Subject: Experimental Psychology-Some dangers?


  

Hello all,

I have never taught experimental psychology, but I have recently gotten
    

hold
  

of a few articles (thanks to help from a couple of Tipsters, among others)
suggesting that undergraduate students often make up (or cook or trim)
    

their
  

data rather than actually collecting it.  One professor told me that when
she made it clear at the onset that the grade was not affected by whether
the results were "significant" (a term that may also be widely
misinterpreted by undergraduates), the number of students papers reporting
statistical significance dropped from almost 100% to less than 50%.

Do any of you think that courses involving the independent collection,
analysis, and reporting of of data in undergraduate courses can
MIS-socialize students into what is expected of a responsible scientist?
    

Do
  

you engage in any special lectures or assign any readings to help minimize
that risk?

I am currently working on a project on responsible conduct of research and
am very interested in views about how the research process is introduced
    

to
  

students.   I should add that I believe that the undergraduate years is a
VERY appropriate time for students to begin collecting actual data,
analyzing, and reporting it because it is the best way to connect them to
our science.  I think waiting until graduate school is too late (and those
who do not go on will have never have experienced how their major field of
study goes about making its discoveries).

Tricia Keith-Spiegel, PhD

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

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

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