Okay, I just took a look at 2 reviews from the Mental Measurements Yearbook about the MBTI Form M. Here's a brief summary about reliability and validity...

The manual reports studies of reliability/validity based on continuous scoring of the 4 dimensions, however the manual also states that the dimensions should not be looked at as continuous scores but rather leading to the various types. The one study that did look at the reliability of the types found that (quoting from the review):

"After a mere 4 weeks, only 65% of the 424 respondents in the retest sample scored the same. Most of the remaining 35% showed consistency in three of the four scales, which is somewhat comforting, but less than adequate given the emphasis on holistic interpretation of the four preferences. Thus, the typical estimates of reliability are quite high, but they provide an inappropriate estimate for this scoring system; the appropriate reliability estimate shows consistent classification for only 65% of respondents, and that is less impressive."

Discussions of validity also followed this pattern... if you violate the assumptions/model of what it is supposed to do (put people into "types") and use the continuous scores then it works okay...

Of course, all of this is for the Myers-Briggs, and not the Kersey Temperament Sorter which is so often used in place of the MBTI but is not the same thing. One difference is that, if I'm remembering correctly, the Kersey allows for the X to represent a tie, but the MBTI uses various rules to force people one direction or the other based on things, such as gender. Form M uses revised scoring based on IRT that helps prevent these ties though.

Just some added information for the on-going debate...

- Marc


=============================================
G. Marc Turner, MEd, Network+, MCP
Instructor & Head of Computer Operations
Department of Psychology
Southwest Texas State University
San Marcos, TX 78666
phone: (512)245-2526
email: [EMAIL PROTECTED]


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

Reply via email to