Date of Award

1976

Document Type

Thesis

Degree Name

Master of Arts

College

College of Education and International Services

First Advisor

Robert J. Cruise

Second Advisor

Rudolf E. Klimes

Third Advisor

Marion J. Merchant

Abstract

Problem

While global forecasts of pupil populations are vital to educational planning, projections at micro-level are even more pertinent to detailed planning. In Jamaica, the need for local level forecasts becomes particularly apparent every September, when there is great pressure from would-be pupils on urban school accommodation.

Method

The review of literature was aimed at identifying projection methods feasible for micro-level applications. The following three methods, identified by Webster (1969) as about the most efficient of seventeen ratio and regression methods of pupil projection which he investigated, were selected for testing on data for the parish of Hanover, Jamaica: Transition Analysis, Time Analysis, and Cohort Survival. Transition Analysis and Cohort Survival utilize birth and enrollment data as predictors. Time Analysis utilizes time and enrollment. The research also involved the creation of fourteen "school districts" in the parish on the basis of physiographic features. Districts were also classified according to previous population change.

Results

Utilizing birth data for 1958/59-1966/67 and enrollment data for 1965/66-1973/74, two sets of projections based on time spans of 8-7 and 6-5 years were made for each of the three methods in seven districts. These projections were gauged against actual enrollment for 1973/74, the third succeeding and immediately succeeding year for the above periods. Both Time Analysis and Cohort Survival yielded better estimates than Transition Analysis with Time Analysis having the edge over Cohort Survival. Since Transition Analysis fell short of expectations and since it was suspected that small, erratic figures were the cause, the methods were tested on total figures for the parish. Here Transition Analysis performed best of the three methods.

Conclusion

Either Time Analysis or Cohort Survival could provide better estimates of future student flows than Transition Analysis in small districts with features like Hanover. On larger, more ordered figures, Transition Analysis might perform best of the three. Tests of the three methods should therefore precede application elsewhere. Statistical methods may usefully be supplemented by other means of estimating population change like the Delphi Technique described in Appendix E. Further work involving the above techniques is recommended.

Subject Area

Educational planning--Jamaica--Statistical methods; Education--Jamaica--Statistics

DOI

https://dx.doi.org/10.32597/theses/161

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS