ABSTRACT
For instructional purposes, the example in the previous topic illustrated a meta-analysis on four studies that were as similar as possible (i.e., a study by Researcher W plus three strict replications). In practice, it would be difficult to find even one perfectly strict replication of a study. One very important way that various studies on a given topic often differ is that various researchers frequently use different measures of the same variable. To see the importance of this issue when conducting and interpreting a meta-analysis, consider the values in Table 1, which are based on a study in which Experimenter A used a test with possible score values from 200 to 800 (like the SAT), while Experimenter B used a test with possible score values from 0 to 50. Table 1 Results Used in Conducting a Meta-Analysis
puted (100.00 + 2.00 = 102.00/2 = 51.00), the average lacks meaning because the results are expressed on different scales. In other words, the answer of 51.00 does not refer specifically to the scale that goes from 200 to 800 nor does it refer specifically to the scale that goes from 0 to 50. When locating additional studies on the same topic, other tests with other score-point ranges are likely to be found, making a simple average of the mean differences entirely uninterpretable.
