|
A Review of High
Stakes Testing:
Who is Smarter––Academic or Vocational Students?
(Submitted at the editor’s request by: Dr.
Jack Elliot,
Associate Professor of Agricultural Education,
The University of
Arizona)
 The question should actually be: "Is a raw score
comparison a fair and accurate measure between the two
groups?" The answer is "No!" On a recent high
stakes test in Arizona, Career and Technical Education (CTE)
students scored lower than other students. Yet, when other
influences were reviewed and factored into the statistical
analysis, the following factors were found to have a significant
influence on the score.
• All 5 "special population" areas
(handicapped, limited English proficiency, economically
disadvantaged, academically disadvantaged and being a single
parent) were associated significantly with lower test scores and
were found predominantly in the CTE population.
• Scores of higher visual learners (those who
learn by seeing) were associated significantly with higher test
scores and were found predominantly with non-CTE students.
• Scores of higher kinesthetic learners (those
who learn by doing) were associated significantly with lower test
scores and were found predominantly with CTE students.
• Neither gender nor race/ethnicity had an
effect on test scores.
Therefore, after controlling for these other influences
(extraneous variables), no difference was found between the two
groups. This means that CTE and non-CTE students are just
different groups of people and a raw score comparison of these two
groups is not an appropriate comparison.
Does grouping students according to their special population
status affect the comparison? Yes. A higher proportion of CTE
students received special population services than non-CTE
students. Further, special population student categories were
associated with lower test scores.
Does this sound like statistical gibberish? Not at all––the
conclusions are simple! Career and technical education students,
for the most part, will perform worse than academic students on
raw score comparisons. However, when we control for the
appropriate extraneous variables in the statistical model (that is
when we include these variables into the analysis), there is no
difference between CTE and other students on standardized test
scores. Comparisons of the raw scores on standardized tests are
inappropriate because the groups are different. The differences in
raw scores between the two groups can be attributed to the effects
of the extraneous variables. The raw score differences between the
groups are not due to curriculum choice!
The implications for educators from these findings are
important. If extraneous effects are not recognized, understood,
and controlled for in statistical analyses, decision makers may
view career and technical education students as "not as
smart" as academic students.
In today’s world where high stakes testing is becoming more
prevalent, CTE teachers and administrators must understand the
problems associated with raw score comparisons on standardized
tests. |
