Purposes and practical considerations enter into choices of statistical analysis methods. This paper considers the application of multilevel modelling to data from a Performance Monitoring system in the United Kingdom known as the A-level Information System (ALIS). The practical question concerned the extent to which it was important to use this more sophisticated model rather than simply using standard OLS regressions, given the kind of data dealt with in this particular monitoring system. We begin with a consideration of dependent variables (outcome indicators) and present an example of their apparent relationship to a process variable. The example is a springboard for a brief discussion of the purpose of a monitoring system. The results of applying ML 2 2 to the analysis of four dependent variables are then presented. The paper concludes with a discussion of the limitations of the data and an attempt to assess the role of multilevel modelling for this particular indicator system. Finally, a suggestion is made that a desirable piece of statistical software might be one delivering rapid-fire graphical bootstrapping.
[1]
R. P. Carver.
The Coleman Report: Using Inappropriately Designed Achievement Tests1
,
1975
.
[2]
J. Willms.
Catholic-School Effects on Academic Achievement: New Evidence from the High School and Beyond Follow-Up Study.
,
1985
.
[3]
M. C. Wittrock,et al.
Generative processes in reading comprehension.
,
1978
.
[4]
B. Efron.
Nonparametric estimates of standard error: The jackknife, the bootstrap and other methods
,
1981
.
[5]
H. Dyer.
School Factors and Equal Educational Opportunity
,
1968
.
[6]
A. Kelly,et al.
Sex Differences in Science and Technology Among 11‐year‐old Schoolchildren: II‐‐affective
,
1984
.
[7]
R. Murnane.
Improving Education Indicators and Economic Indicators: The Same Problems?
,
1987
.
[8]
B. Efron.
Bootstrap Methods: Another Look at the Jackknife
,
1979
.