[Rasch] Rasch analysis of interval data

Timothy W. Pelton tpelton at uvic.ca
Thu Jul 17 08:34:41 EST 2008


We conducted an experiment where we had participants compare physical rods of different lengths (ranging from 3mm to 350mm) to mental images of appropriately comparable common objects (orange, pop can, pencil)

The resulting scale was not linear - but after a simple transformation, the results were very accurate.  We hypothesized that this was related to an innate perceptual distortion and the fact that we were asking the participants to to compare physical objects to recalled objects.  There is lots of early work on perception and we are still looking for appropriate references relevant to our experiment.

We also used two types of rods (wire and wooden) - and found a modest but consistent shift in estimated lengths of rods (and conducted a second experiment to confirm). From this experience, I would suggest that if you conduct your experiment, you might begin by picking 'object' people with similar body styles and gender or at least look at two groups.


> on 16/07/08 3:11 PM, Mike Linacre (RMT) at rmt at rasch.org wrote:


Yes, there have been several comparisons of Rasch measures and physical
measures. Here are two:
http://www.rasch.org/rmt/rmt71b.htm - physical distance
http://www.rasch.org/rmt/rmt112a.htm - physical weight

But would be delighted if you would conduct your height experiment and tell
us about it.

Mike L.

You wrote:
I'd love to see what happens. Pairing persons heights in cm and their Rasch
heights simultaneously on a vertical line and comparing the calibrations
give a good test of Rasch's interval scaling, I guess.

Rasch mailing list
Rasch at acer.edu.au

-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20080716/35fde5ba/attachment.html 

More information about the Rasch mailing list