Skip Navigation
small NCES header image
Projections of Education Statistics to 2018

NCES 2009-062
September 2009

Projection Methodology: Degrees Conferred

Projections of associateís, bachelorís, masterís, and first-professional degrees for men and women were based on demographic models that relate degree awards to college enrollment by level enrolled and attendance status. Table A-22 describes the estimated equations used to calculate projections. The equations shown were selected on the basis of their statistical properties, such as coefficients of determination (R2s), the t-statistics of the coefficients, the Durbin-Watson statistic, the Breusch-Godfrey Serial Correlation LM test statistic, and residual plots.

The equations used to produce the degrees conferred projections for the previous edition of this report (Projections of Education Statistics to 2018) all used a similar form in which the log of the ratio of the number of degrees to the population of the relevant age group was regressed on the log of the ratio of enrollment in the relevant level to the population of the relevant age group. In the equations for associateís, bachelorís, masterís, and first–professional degrees, the number of degrees is expressed as either a first–difference or a percentage change. This value is regressed on the enrollment in the relevant level, again expressed as either a first–difference or a percentage change. The projections of the of doctorís degrees for men and women were produced using double exponential smoothing.

Associate's Degrees

Associateís degree projections for men and women were based on 2 years full–time undergraduate enrollment in 2–year institutions by sex. Menís projections were based on current and lagged 2 years full–time enrollment, and womenís projections were based on the current full–time enrollment and enrollment lagged 1 and 2 years. Results of the regression analysis used to project associate's degrees are shown in table A–22.

Bachelor's Degrees

Bachelorís degree projections for men and women were based on current and lagged 2 years full–time undergraduate enrollment in 4–year institutions by sex. Results of the regression analysis used to project bachelor's degrees are shown in table A–22.

Master's Degrees

Masterís degree projections for men and women were based on full–time graduate enrollment by sex. Menís projections were based on current and previous year enrollment, and womenís projections were based on current enrollment. Results of the regression analysis used to project master's degrees are shown in table A–22.

Doctor's Degrees

Doctorís degree projections for men and women were obtained by double exponential smoothing of the historical data with a smoothing parameter of 0.4.

First–Professional Degrees

First–professional degree projections were based on total full–time first–professional enrollment lagged 1 and 2 years by sex. Results of the regression analysis used to project first–professional degree are shown in table A–22.

Projection Accuracy

An analysis of projection errors from similar models used in the past twelve editions of Projections of Education Statistics indicates that mean absolute percentage errors (MAPEs) for associateís degrees were 2.1 percent for 1 year out, 3.3 percent for 2 years out, 6.0 percent for 5 years out, and 15.6 percent for 10 years out. For the 1–year–out prediction, this means that one would expect the projection to be within 2.1 percent of the actual value, on average. MAPEs for bachelorís degree projections were 0.9 percent for 1 year out, 1.9 percent for 2 years out, 6.0 percent for 5 years out, and 13.5 percent for 10 years out. MAPEs for masterís degrees were 1.5, 3.5, 12.4, and 25.0 percent, respectively. For doctorís degrees, the MAPEs were 3.4, 5.5, 6.1, and 11.9 percent, respectively. For first–professional degrees, the MAPEs were 1.3, 1.7, 5.1, and 13.8 percent, respectively. For more information on the MAPEs, see table A–2.

Top


Would you like to help us improve our products and website by taking a short survey?

YES, I would like to take the survey

or

No Thanks

The survey consists of a few short questions and takes less than one minute to complete.