Relative standing: Where does a score reside compared to everyone else?
Percentiles or ordinal position or rank: is not sensitive to the variability of raw scores, just ranking.
Standard scores (z-scores): Statistical approach to standardizing scores in a standard scale of measurement. The relative standing of one score can be compared to another, even when they are measured on different scales (GMAT, CPA, GRE, etc.). A numerical index of relative standing expressed in standard deviation units. The mean has a z-score of 0 standard deviation units.
Calculated by (score - population mean)/population standard deviation also known as the distance from the mean expressed in standard deviation units.
Formula structure is "observed" minus "expected" divided by "error".
Was Wayne Gretsky a better scorer than Michael Jordan? (Compare z-scores)
If you don't know the population standard deviation, generally you don't calculate z-scores.
Z-distribution (Appendix B, pg B1-B5)
Table calculates the (a) area between the Z score and Mean and (b) area beyond the Z score in a normal distribution.
The "area beyond the Z score" is the probability score.
Wednesday, February 10, 2010
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment