[IPAC-List] Rescaling exam scores

Reed, Elizabeth EReed1 at Columbus.gov
Tue Oct 6 10:42:39 EDT 2015


Steven,

The city of Columbus standardizes using z-scores for the purpose of combining multi-phases exams and for the purpose of the addition of points (seniority or vet preference) so that the impact is similar across exams.  The image below is a sample of the scoring process we use for the promotional exams. We provide this as a handout upon request. Please note, in multi-phase exams we re-z the combined z scores. The re-z is for the purposes you cite.

In the example we use reported  score of: reported score = (z x 8.5714) + 74.2857

We may use other formulas for the reported scores depending upon test plan. Our policies provide guidance to determine when to use each reported score formula are used under various circumstances. Ranked, Banded, hurdles present, are some of the considerations. Other samples used are:

Reported score = (z x 10) + 80
Reported score = (z x 12) + 76
Reported score = (z x 8.5714) + 78.5714

FYI: For exams that we use again and again, thus we have a great deal of data, we have used a 4-point rescale based upon previous test results.
Best wishes.
[cid:image002.png at 01D10023.B8F6C350]

Elizabeth A. Reed
Public Safety Assessment Team Manager
Columbus Civil Service Commission
______________________________

Direct: 614.645.6032
http://www.columbus.gov/

From: IPAC-List [mailto:ipac-list-bounces at ipacweb.org] On Behalf Of Partain, Steven C.
Sent: Monday, October 05, 2015 5:08 PM
To: 'ipac-list at ipacweb.org'
Subject: [IPAC-List] Rescaling exam scores

Our agency has recognized the variation of impact of veterans' preference (for our state: 5 points for veterans and 10 points for disabled veterans) on civil service exams depending on the distribution of scores.  I'm toying with the idea of rescaling all exam scores to standardize distributions and thus standardize the impact of preference.  My statistical background, unfortunately, is a few decades stale.

Does anyone have experience with this sort of rescaling?  I'll eventually figure it out with my google research, but thought maybe one of you does this routinely and has a suggested method/formula.

I'm also interested in any feedback on the concept as it relates to preference.

Thanks,

____________________________________________________
Steven Partain | HR Manager
Tualatin Valley Fire & Rescue
Direct: 503-259-1292
www.tvfr.com<http://cp.mcafee.com/d/1jWVIe3xEqdEIccnuhhsvdTdETu7fcLfc8CQrL3DCnDDAPqaabwVNVB5cSyyyeh7cL9CzATYvZ145mGuSGgrFN6FATqFXqF1KD4qCjq5_Zd5N_HYMqem1PfnKnjoooV6X5SbZh5dqWqJXIYG7DR8OJMddECQjt-h7cffT73DPhOrKr01jcKNVv3qQjd3HzAXSpptgQWiSSbDOFcKgGT2TQ1kBJI0bw09J6X3bNIQJxEm418K5pkCqnjh02_FrjVg8Cy3M_ljh1iRo_gSOOr8VIA>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist9.pair.net/pipermail/ipac-list/attachments/20151006/1b4b152a/attachment.html>


More information about the IPAC-List mailing list