Test anxiety reduction research bloomed during the late 1960’s, with the introduction of relaxation and systematic desensitization.
High test anxious students were found to score almost half a standard deviation (SD) below their low anxiety peers, or 12 percentile below. Present overviews estimate that test anxiety accounts for about 7–8 percent of the variance across the whole range of students,* which is about equivalent to a 12 percentile span between the high vs. low anxiety students with the moderately-high and moderate anxiety students omitted. Thus, the 12 percentile high vs. low difference continues in present findings.
The consensus from the sixties into the eighties was that available treatment reduced test anxiety but produced little or no improvement in test performance. A 1980 review of almost 50 studies found only 18% showing a statistically significant improvement in performance, and concluded that performance improvement should not be expected. Hembree notes that many of these studies involved too few subjects to attain statistical significance.
Using a more complete sample of available studies, and the more advanced meta-analysis methods, Hembree observed that many of the better protocols produced about a .50 SD improvement on average in test scores and GPA, removing the handicap.
Current research appears to support grade improvement benefits somewhat below the earlier 12 percentile gain. We might estimate that a good test anxiety reduction protocol should produce about a 6–12 percentile grade benefit, which would be about 3–6 tenths of a letter grade benefit. Current interventions can now be measured against this established standard, and should be expected to produce the 6–12 percentile grade improvement.
McDonald, A. (2001). The Prevalence and Effects of Test Anxiety in Children. Educational Psychology. 21, 1, 89-10; and
Cassady J. & $. Johnson (2002). Cognitive Test Anxiety and Academic Performance. Contemporary Educational Psychology, 27, 270-295.