Why were changes made to the 2005 Benchmark Report?
There are two major reasons for the changing NSSE Benchmark Report from institutional level comparison to student level comparison. Prior to 2005, NSSE benchmark comparison group scores represented institutional level scores. By taking the average of the institutional scores (i.e., the mean of the school means), we reduce the comparison group scores to between institution variance only. Over the years, our analyses have shown that the between institutional variance is consistently much smaller than the within institution variance. In other words, the largest differences in student engagement occur inside a given school, not between institutions. The phenomenon is not unique to student engagement results, but is typical of measured variables at all educational levels. The second reason for this change is the presence of standard error problem for a number of institutions, particularly those with small numbers of respondents. Standard errors, an estimate of random fluctuations in samples
