Methodological issues in the calculation of enrolment retention rates

This study examines the methods used in calculating age- and grade- retention rates for secondary schools in the Victorian education system. It focusses on the methodology' used to obtain these statistics rather than on applications of the generated statistic. However, the administrative uses for which retention rates are calculated have been considered. As a consequence of applications of the retention statistic being diverse and requiring considerable detail to be fully understood, their discussion will be reserved to a separate appendix (see Appendix I, The Collection and Use of Enrolment Data in the Victorian Education Department). The use of school data to calculate retention may be regarded as an attempt to overcome the problems faced in collecting individual pupil data. The latter would necessitate the tracing of all individuals in each age-cohort separately over that time, a task requiring data-collection methods which are yet to be made available. ''Apparent" retention rates, as they are usually called, are used on the assumption that they are unlikely to differ from "exact" retention rates, those rates obtained when individual data is used, at least in the case of aggregates greater than the individual school level. Any errors resulting from this assumption can possibly be reduced by altering the method of calculating apparent retention rates. This thesis investigates a number of different approaches of calculating retention rates with this aim in view. This thesis does not attempt to reveal anything new about the factors which affect retention. However, it is clear that in exploring new methods of calculating retention, the influence of these factors must be kept in mind. Such factors may be of long-term rather than short-term influence or they may be an aggregate one rather than one which is exhibited at a particular school. The literature on the subject is equivalent to much of the work on social differences in educational achievement, since retention rates are in one respect the most rudimentary index of school achievement, the length of schooling. Thus, no attempt has been made to offer a review of the general literature. Instead, monographs or reports most relevant to the methodological aspects are cited in the footnotes at the appropriate points, or discussed within the text when necessary