CSFI NORMING CALCULATIONS
Scale Averages
The norming of the CSFI began in 1994 with50 graduate students in counseling. Since that datethe development of the LP forms has continued, andmany persons have been added to the norm group.
For a combined sample of 2256 persons, the scalemeans were:
However, these means included samples fromseveral special groups with acknowledged learningdeficits or special other characteristics. Therefore, forcomparisons with your college sample, use Table 3,or Table 2 for comparisons to specific samples.
The means for 1485 regular admissions collegefreshmen are:
The standard deviations for the CSFI Total Scorevary between groups, and range from 26 to 30 formost colleges. With a few exceptions, subscale standarddeviations between groups, and between subscales,are typically 4.5, plus or minus 5.5. Refer toTables 2 and 3 for the samples and their means.
Recent mean calculations include data from 28colleges and universities, resulting in an overall meanof 171.8 and 165.6. (See Tables 4 and 5)
Since the onset of a contract, the CSFI has beenadministered to upwards of 100,000 students. Over
the last seven years, approximately 17,000 studentscores have added each year. Scores and means havechanged little since 2001.
Gender Differences
Analyses suggest that there was a significant differencebetween male and females in the Competitionscale, males proving to endorse a higher level ofcompetition. In that first comparison, males averaged19.7, females averaged 22 .3. The difference wassignificant at p=.003. No other differences werenoted.
A replication in 1991-92 with 860 enteringfreshmen at a Midwestern university confirmed thedifference. In this group, the average for males was18.5, and for females, 19.9. The difference of 1.4 wassignificant at p<.0001. In addition, females in thisgroup were more inclined to endorse Time Managementitems, with a scale average of 23.2, comparedto the male average of 24.9. The 1.7 point differencewas significant at p<.0001.
The overall results indicate little differences initem response or total score between males and females.Thus, a single continuum of scores is presentedfor both male and female, keeping in mind the differencesnoted above.
Cultural Diversity
Cultural diversity must be supported in any representative test. So the authors took as a diversity modelthe California Community Colleges, and some state universities. The percentage breakdown of four ethnicgroups (excluding no response and other) was 53.6% Caucasian, 24.7% Hispanic, 8.9% African-American,and 12.9% combined Asian students. From the origination of the test, two studies have indicated similarrepresentation.
The standard deviations for the CSFI Total Score vary between groups, and range from _6 to _6 for most colleges. With a few exceptions, subscale standard deviations between groups, and between subscales, are typically 4.5, plus or minus.
Special Sample Representation
In the development of the CSFI, several special college populations were included in the developmentplan. These were the following:
Learning Disabled
Re-Entry Students
E.O.P.S.
Career Exploration
College Transfer Psychology Courses
M.S. Candidates in Counseling
Students in Study Skill Classes
Students on Probation
College Orientation
These groups and sample sizes are shown in Tables 2 and 3. Heterogeneous college student groups arerepresented in Tables 4 and 5.
Table 4 continued.
Table 5 continued.