[Rasch] standardized scores

Jim Sick jimsick at gmail.com
Mon Nov 28 21:03:12 EST 2011


Standardized scores, by definition, have a mean of zero, or, if a 
transform such as a T-score is used, a constant such as 50. Both groups 
will have equivalent means if scores are standardized so it is thus 
impossible to compare the two groups with the data given. I hope, for 
the sake of the textbook users, that that was the author's intended 
solution.

Best,

Jim

On 11/27/11 11:30 PM, Anthony James wrote:
> Dear all,
> In an introductory research and statistics book there's a chapter on 
> standard scores (z score and T score) and there are some activities at 
> the end. One of the activities reads something like this:
> There are two groups, one experimental and one control. After the 
> treatment we give them different tests which have different scales for 
> scoring. The scores of Group 1 are 3, 3, 5, 8, 9, 10, etc. and scores 
> of Group 2 are 25, 87, 55, 78, 94, 36, etc. Has the treatment been 
> effective?
> I believe the author expects the students to convert the scores to 
> z-scores or any other standard score and then compare the means of the 
> standardized scores. Am I right?
> Does transforming scores to z scores make scores across tests and 
> groups comparable? Z scores are scores in standard deviation units and 
> depend on the ability of the population. Can we compare two groups in 
> this way?
> Cheers
> Anthony
>
>
> _______________________________________________
> Rasch mailing list
> Rasch at acer.edu.au
> Unsubscribe: https://mailinglist.acer.edu.au/mailman/options/rasch/jimsick%40iname.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20111128/abde656f/attachment.html 


More information about the Rasch mailing list