Math

1
You have been hired to evaluate the success of a
6-week adult education program for the state. A
sample of the pre-post raw scores are below.
After conducting an ANOVA, you find that the is
no significant difference in change scores.
You are asked by the client to analyze the data
in a different way to count the percent of
people who improved from their pre-test score.
The results are graphically shown on the next
page.
Case 1
Math Pretest 55 64 44 33 28 63 48 38 46 47
Math Posttest 56 66 46 38 29 63 50 40 48 47
English Pretest 33 35 43 36 20 60 40 31 52 64
English Posttest 35 37 47 36 21 62 40 31 56 66
2
The client wants you to include this analysis in
your final report. What should you do?
increase
100 90 80 70 60 50 40 30 20 10 0
The Product
Math
English
3
A company hires you to analyze customer
complaints as part of a larger evaluations study
of their performance. You produce a preliminary
graph (below). The organization believes that
this could be misinterpreted to indicate that
customer complaints are higher in the summer when
they often cut back on temporary staff. So, they
request that you represent the data in a
different form (see next page)
Case 2
10 9 8 7 6 5 4 3 2 1 0
J F M A M J Jul
Aug S O N D
4
Case 2 (cont.)
50 45 40 35 30 25 20 15 10
5 0
How would you respond to their request?
J F M A M J Jul
Aug S O N D
5
Case 3
As a consultant you are ask to evaluate the
effectiveness of an organizations recent drug
testing program. The company provides you with
data that suggests the program was a success.
That is, the year before drug testing began
(1995), the number of accidents was 50. In 1996,
the year testing was initiated, the amount
dropped to 40. In 1997, the year following the
program, the number of accident dropped to 29.
Your preliminary analysis to the company is
summarized below, a finding which is widely
touted by the organization (e.g., company CEO,
local newspaper, trade magazine).
55 50 45 40 35 30 25 20 15 10 5



1995
Drug Testing
1997
6
Case 3 (cont.)

65 60 55 50 45 40 35 30 25 20 15
After further analysis of company data, you
develop the graph below. How would you handle
this situation?








1992 1993 1994 1995 1996
1997 1998 1999 2000
7
Case 4
You have been hired to perform a validation study
to assess the predictive nature of a large-scale
assessment center for a multi-national
corporation. The results of your work are
summarized below.
Correlations with job performance (corrected for
attenuation range restriction)
Correlations with job performance
Correlations with job performance (corrected for
attenuation)
Assessment Center Exercise Scores Leadership Deci
siveness Planning Organization Resilience Verbal
Communication Written Communication Initiative To
tal Score
.16 .12 .11 .05 .22 .20 .14 .24
.18 .14 .13 .06 .24 .23 .17 .27
.20 .16 .13 .07 .32 .30 .21 .31
8
Case 4 (cont.)
After seeing the results of your investigation,
the company wants you to produce an executive
report (summary) highlighting the twice corrected
correlations, and to not include the lower
(uncorrected, less-corrected ones). The ask that
these correlations be included in a lengthy final
report, perhaps in an appendix. What would you
do in this instance?