HomeRoast Digest


Topic: now OT: Re: +Donkey Blend Cappos (2 msgs / 42 lines)
1) From: Pecan Jim Gundlach
On Feb 15, 2005, at 9:42 AM, Jared Andersson wrote:
<Snip>
Later research shows that the attention effect has been greatly 
exaggerated and that the Hawthorn effect did not even exist once the 
Hawthorn data was subjected to modern statistical analysis techniques.
Jim Gundlach

2) From: Frank Haist
Agreed that Jared's concept of the Hawthorne effect of a 25% behavior 
changes a rule of thumb due to observation or manipulation is incorrect. 
I've never seen a reference to anything that high consistently.  
However, I am unaware of studies showing the original Hawthorne effect 
was an artifact of analysis methods. I took a brief look in the PsycLit 
and PubMed databases and couldn't find such a reference. Granted, a 
Hawthorne effect tends to be short-lived, but it does exist. It has been 
replicated, to paraphrase Yogi Bera, over and over. Indeed, I believe a 
follow-up to the original Hawthorne observation, done in the same plant, 
used an A-B-A design that showed the effect reversed itself, something 
you don't need modern statistical analyses to document the reliability 
of the effect (the argument then becomes one of practical significance 
rather than statistical significance). That is, they replaced the 
lighting with what was there originally and productivity again was 
boosted. I'd appreciate the reference you alluded to as I occasionally 
have to teach intro psych level and stats courses and would like to have 
this on hand. Wow, this *is* off topic. I apologize.
---Frank
Pecan Jim Gundlach wrote:
<Snip>


HomeRoast Digest