Grouping Colleges by Changes in Enrollment Volume

1
Grouping Colleges by Changes in Enrollment Volume

• Willard Hom
• Director, Research Planning
• Chancellors Office,CCC

2
Study Question

• Are there distinct patterns in the changes to
  enrollment volume among CCCs during recent years?

3
Uses for the Analysis

• Projecting enrollments may depend upon certain
  numeric patterns of enrollment change.
• Understanding and managing enrollment trends can
  benefit from a grouping of change patterns that
  can help us see common contributing factors.

4
Data

• Annual fall term credit enrollment counts for 113
  reporting public 2-year institutions.
• Years of 1991 through 1999.

5
The Construct of College-Level Enrollment Change

• Relative variability.
• Slope or direction of change, if any.
• Association with overall change in the state.

6
Relative variability

• Standard deviation of the percent change from the
  prior year.

7
Slope or direction of change

• Coefficient for the time parameter in a simple
  (one predictor) trend regression equation.

8
Association with overall change in the state

• Spearman correlation coefficient between college
  time series and state time series.

9
Summary of Indicators

• Relative variability.
• Slope or direction of change, if any.
• Association with overall change in the state.

• S.D. of change in counts.
• Coefficient of the regression line for the time
  series.
• Spearman correlation with state totals.

10
Graph of State Trend, 1991-99
11
Relative Variability
12
Slope or direction of change
13
Association with overall change in the state
14
Correlation Matrix for Variables
15
Clustering Approaches Used

• Wards Method with Euclidean Squared Distances.
• Average Linkage Method with Euclidean Squared
  Distances.
• Average Linkage Method with Pearson Similarity
  Metric.

16
Results of Ward/Euclidean, Part 1
17
Results of Ward/Euclidean, Part 2
18
Results of Ward/Euclidean, Part 3
19
Next Steps

• Do further tests on the quality of the cluster
  structures found.
• Try to interpret the cluster structures.
• Test alternative indicators of change.
• See if the cluster findings can lead to some
  theory.
• See if the enrollment planning can benefit.

20
Basic References

• Lorr,M. (1987). Cluster Analysis for the Social
  Sciences. S.F. Jossey-Bass.
• Hair, J.F. Black, W.C. (2000). Cluster
  Analysis. In L.G.Grimm P.R.Yarnold (eds.).
  Reading and Understanding More Multivariate
  Statistics. Washington, D.C. American
  Psychological Association.