Skip Navigation
small NCES header image
Projections of Education Statistics to 2019

NCES 2011-017
March 2011

Introduction to Projection Methodology: Enrollment in Postsecondary Degree-Granting Institutions


Projections in this edition

This edition of Projections of Education Statistics presents projections of enrollment in degree-granting institutions for fall 2009 through fall 2019. Three different models were used to produce these enrollment projections:

  • The Enrollment in Degree-Granting Institutions Model produced projec-tions of enrollments by attendance status, level of student, type of institution, control of institution, sex, and age. It also produced projections of full-time-equivalent enrollments by level of student, type of institution, and control of institution.
  • The Enrollment in Degree-Granting Institutions by Race/Ethnicity Model produced projections of enrollments by race/ethnicity.
  • The First-Time Freshmen Model produced projections of enrollments of first-time freshmen by sex.

Overview of approach

Basic features of the three degree-granting enrollment models

The Enrollment in Degree-Granting Institutions Model is the primary model for projecting enrollment in degree-granting postsecondary institutions. For this model, enrollment rates by attendance status and sex are projected for various age categories using either the pooled seemingly unrelated regression method or the pooled seemingly unrelated regression method with a first-order autocorrelation correction. These rates are applied to projections of populations of the same sex and age to produce projections of enrollment by attendance status, sex, and age. To project enrollments by level of student, type of institution, and control of institution, rates for these characteristics are projected using single exponential smoothing and applied to enrollment projections previously produced by the model.

The Enrollment in Degree-Granting Institutions by Race/Ethnicity Model takes an approach similar to that of the Enrollment in Degree-Granting Institutions Model. Enrollment rates by attendance status, sex, and race/ethnicity are projected for the age categories using either the pooled seemingly unrelated regression method or the pooled seemingly unrelated regression method with a first-order autocorrelation correction. The resulting rates are iteratively corrected to ensure consistency with those projected by the Enrollment in Degree-Granting Institutions Model. The adjusted rates are then applied to projections of populations of the same sex, age, and race/ethnicity.

The First-Time Freshmen Enrollment in Degree-Granting Institutions Model uses single exponential smoothing to project the ratio of freshmen enrollment to undergraduate enrollment separately for males and for females. It then applies the projected ratios to the projections of undergraduate enrollment by sex that were produced by the Enrollment in Degree-Granting Institutions Model.

The Enrollment in Degree-Granting Institutions Model

The Enrollment in Degree-Granting Institutions Model produces projections of enrollment counts by six levels of detail, as well as projections of full-time-equivalent enrollments by level of student, type of institution, and control of institution.

Steps used in the Enrollment in Degree-Granting Institutions Model

Step 1. Adjust age-specific enrollment counts from the U.S. Census Bureau to make them agree with the more highly aggregated NCES enrollment counts that do not include age. The Enrollment in Degree-Granting Institutions Model projects enrollments by six levels of detail: attendance status, level of student, type of institution, control of institution, sex, and age. While NCES does produce enrollment counts by the first five levels of detail, it does not produce data by the sixth level of detail, age, every year. However, the U.S. Census Bureau does produce age-specific enrollment counts.

In step 1, the age distributions from the Census Bureau counts for 1980 to 2008 were applied to the NCES counts to produce a set of enrollment data that breaks enrollments down by age while being consistent with NCES counts. Specifically, the most detailed level of Census Bureau data (by attendance status, level of student, type of institution, control of institution, sex, and age) was iteratively changed using proportions based on the more highly aggregated NCES enrollment numbers to ensure that all sums across this most detailed level of Census enrollment data equaled the more highly aggregated NCES enrollment totals that did not include age.

Step 2. Calculate enrollment rates by attendance status, sex, and age category. The enrollment data were broken up into 14 age categories, with separate age categories for individual ages 14 through 24 as well as for the age groups 25 to 29, 30 to 34, and 35 and over. For each of the 14 age categories, 4 enrollment rates were calculated—part-time male, full-time male, part-time female, and full-time female—resulting in a total of 56 enrollment rates. Each of the 56 enrollment rates was calculated by dividing the enrollment count for that combination of attendance status, sex, and age category by the total population for the corresponding combination of sex and age category. For each combination of attendance and sex, the enrollment rate for the oldest age category was calculated by dividing the enrollment count for those 35 and over by the total population for those 35 to 44.

Step 3. Produce projections of enrollment rates by attendance status, sex, and age category. Enrollment rates for most of the age groups were projected using multiple linear regression. However, because enrollment at degree-granting institutions is negligible for ages 14, 15, and 16, these ages were not included in the multiple linear regression models. Instead, projections for individual ages 14, 15, and 16 were produced by double exponential smoothing.

The following 11 age categories were modeled: individual ages 17 through 24 and age groups 25 to 29, 30 to 34, and 35 and over. For each of these age categories, enrollment rates by attendance status and sex were produced using 4 pooled time-series models—one for each combination of attendance status and sex. Each model was pooled across age categories. Each equation contained two independent variables, which were measures of

  • disposable income; and
  • the unemployment rate.

Either the pooled seemingly unrelated regression method or the pooled seemingly unrelated regression method with a first-order autocorrelation correction was used to estimate each equation.

For more details on the equations used in step 3, the data used to estimate these equations, and their results, see tables A-14 through A-16 on pages 123–125.

Step 4. Produce projections of enrollments by attendance status, sex, and age category. For each combination of attendance status, sex, and age category, enrollment projections were produced by multiplying the projected enrollment rate for that combination by projections of the total population with the corresponding combination of sex and age category.

Step 5. Add two additional levels of detail—level of student and type of institution—to the projected enrollments by attendance status, sex, and age category. For this step, the 14 age categories used in the previous steps were collapsed into the following 8 categories: ages 14 to 16, 17, 18 and 19, 20 and 21, 22 to 24, 25 to 29, 30 to 34, and 35 and over. Step 5 can be broken into three parts:

First, the historic data were used to calculate the percentage distribution of enrollment by level of student and type of institution for each combination of attendance status, sex, and age category. Because it was assumed that there was no enrollment in 2-year institutions at the postbaccalaureate level, three combinations of student level and institution type were used: undergraduates at 4-year institutions, undergraduates at 2-year institutions, and postbaccalaureate students at 4-year institutions.

Second, for each combination of attendance status, sex, and age category, the percentage distribution by level of student and type of institution was projected using single exponential smoothing with a smoothing constant of 0.4 and then adjusted so the sum of the categories by attendance status, level of student, type of institution, sex, and age category would equal 100 percent.

For the projected percentage distributions from step 5 and the actual 2008 distributions, see tables A-17 and A-18 on pages 126–127.

Third, the projected distributions by level of student and type of institution were applied to the projected enrollments by attendance status, sex, and age category from step 4 to obtain the enrollment projections by attendance status, level of student, type of institution, sex, and age category.

Step 6. Add the sixth level of detail—control of institutions—to the projected enrollments in degree-granting institutions. In this step, the data on enrollment by age category were not used. Control of institutions was added in the following manner:

First, the historic data were used to calculate the percentage of enrollment in public institutions for each combination of attendance status, level of student, type of institution, and sex.

Second, the percentages of enrollment in public institutions were projected using single exponential smoothing with a smoothing constant of 0.4.

For the projected percentages from step 6 and the actual 2008 percentages, see table A-19 on page 127.

Third, the projected percentages were applied to the projected enrollments in each corresponding enrollment combination to obtain projections for public institutions by attendance status, level of student, type of institution, and sex.

Fourth, the projected enrollments for public institutions were subtracted from the total to produce the projected enrollments for private institutions.

Step 7. Produce projections of full-time-equivalent enrollment by level of student, type of institution, and control of institution. Full-time-equivalent enrollment represents total full-time and part-time enrollment as if it were enrollment on a full-time basis. It equals the sum of full-time enrollment plus the full-time-equivalent of part-time enrollment. Full-time-equivalent enrollment projections were produced in the following manner:

First, for each combination of level of student, type of institution, and control of institution, the historic data were used to calculate the full-time-equivalent of part-time enrollment as a percentage of part-time enrollment.

Second, for each combination of level of student, type of institution, and control of institution, the full-time equivalent of part-time enrollment as a percentage of part-time enrollment was projected using single exponential smoothing with a smoothing constant of 0.4.

Third, for each combination of level of student, type of institution, and control of institution, the projected percentages were applied to the projections of part-time enrollment to project the full-time equivalent of the part-time enrollment.

Fourth, the projections of full-time equivalents of part-time projections were added to projections of full-time enrollment to obtain projections of full-time-equivalent enrollment.

Data and equation results for the Enrollment in Degree-Granting Institutions Model

Enrollment data for degree-granting institutions. Enrollment data for 1981 to 2008 by attendance status, level of student, type of institution, control of institution, and sex came from the NCES Integrated Postsecondary Education Data System (IPEDS). These are universe counts. The U.S. Census Bureau was the source for enrollment estimates for 1981 to 2008 by the characteristics listed above plus age.

Population data and projections. Population counts for 1980 to 2008 came from the U.S. Census Bureau. Population projections for 2009 to 2019 are the Census Bureau’s 2008 National Population Projections of the population by sex and age (August 2008). For more information, see Section A.0. Introduction, earlier in this appendix.

Projections for economic variables. The economic variables used in developing these projections were from the “U.S. Monthly Model: November 2009 Short-Term Projections” from the economic consulting firm, IHS Global Insight.

Data and results for the equations. The following details for the equations are shown on pages 123–127:

  • Table A-14 shows enrollment rates by sex, attendance status, and age for fall 2008 and projected enrollment rates for fall 2014 and fall 2019.
  • Table A-15 shows the estimated equations and model statistics used to project enrollments for men by attendance status, and table A-16 shows the estimated equations and model statistics used to project enrollment rates for women by attendance status. The particular 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.
  • Table A-17 shows actual and projected percentage distributions of full-time students, and table A-18 shows actual and projected percentage distributions of part-time students.
  • Table A-19 shows actual and projected data for enrollment in public degree-granting institutions as a percentage of total enrollment.

Accuracy of projections for the Enrollment in Degree-Granting Institutions Model

Mean absolute percentage errors (MAPEs) for enrollment in degree-granting institutions were calculated using the last 12 editions of Projections of Education Statistics. Exhibit A-7, below, shows MAPEs for key projections of the Enrollment in Degree-Granting Institutions Model.

Exhibit A-7. Mean absolute percentage errors (MAPEs), by lead time for enrollment in degree-granting institutions: 2010

Statistic Lead time (years)
1 2 3 4 5 6 7 8 9 10
Total enrollment 1.4 2.4 2.9 3.4 4.6 6.1 8.0 9.8 10.1 10.4
Men 1.5 2.9 3.5 4.1 5.2 6.6 8.1 9.8 10.2 10.3
Women 1.5 2.4 3.0 3.2 4.2 5.8 7.9 9.8 9.9 10.4
4-year institutions 1.5 2.5 3.3 4.1 5.3 6.8 8.8 10.9 11.5 12.4
2-year institutions 2.0 3.4 3.8 4.0 4.8 5.0 6.6 7.8 7.6 6.9
NOTE: MAPEs for degree-granting institution enrollments were calculated using the last 12 editions of Projections of Education Statistics. Calculations were made using unrounded numbers. Some data have been revised from previously published numbers.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Projections of Education Statistics, various issues. (This table was prepared February 2010.)

 

For more information about MAPEs, see Section A.0. Introduction, earlier in this appendix.

The Enrollment in Degree-Granting Institutions by Race/Ethnicity Model

The Enrollment in Degree-Granting Institutions by Race/Ethnicity Model projects enrollments in degree-granting institutions by attendance status, sex, age, and race/ethnicity. The following groups are projected in this model:

  • White;
  • Black;
  • Hispanic;
  • Asian/Pacific Islander;
  • American Indian/Alaska Native; and
  • nonresident alien.

See the Glossary for definitions of the five racial/ethnic categories and the nonresident alien category. (The race/ethnicity of nonresident aliens is unknown, but they are considered a racial/ethnic group for purposes of this analysis.)

Steps used in the Degree-Granting Institutions by Race/Ethnicity Model

Step 1. Adjust U.S. Census Bureau enrollment counts by attendance status, sex, age, and race/ethnicity to make them sum to NCES enrollment counts by attendance status, sex, and race/ethnicity. For 1981 to 2008, the most detailed levels of Census Bureau enrollment data (by enrollment status, sex, age, and race/ethnicity) were iteratively changed using proportions that were based on the more highly aggregated NCES enrollment numbers to ensure that the sums across these most detailed levels of enrollment data equaled the more highly aggregated NCES enrollment numbers that did not include age.

Step 2. Calculate enrollment rates by attendance status, sex, age category, and race/ethnicity. The enrollment data were broken up into 14 age categories, with separate age categories for individual ages 14 through 24 as well as for the age groups 25 to 29, 30 to 34, and 35 and over. For each of the 14 age categories, enrollment rates were calculated for each combination of attendance status, sex, and the six racial/ethnic groups, resulting in a total of 336 enrollment rates. Each of the 336 enrollment rates was calculated by dividing the enrollment count for that combination of attendance status, sex, age category, and race/ethnicity by the total population for the corresponding combination of sex, age category, and race/ethnicity. For each combination of attendance status, sex and racial/ethnic group, the enrollment rate for the oldest age category was calculated by dividing the enrollment count for those 35 and over by the total population for those 35 to 44.

Step 3. Produce projections of enrollment rates by attendance status, sex, age category, and race/ethnicity. Enrollment rates for most of the age groups and racial/ethnic groups were projected using multiple linear regression. However, there were several exceptions:

  • Due to negligible enrollments for ages 14, 15, and 16, these ages were not included in the multiple linear regression models. Instead, projections for individual ages 14, 15, and 16 were produced by single exponential smoothing.
  • Due to the nature of the historical data, American Indian/Alaska Native enrollments were projected using single exponential smoothing.
  • Due to the nature of the historical data, non-resident alien enrollments were projected using patterns in recent historical growth.

Four racial/ethnic groups were modeled: White, Black, Hispanic, and Asian/Pacific Islander. Eleven age categories were modeled: individual ages 17 through 24 and age groups 25 to 29, 30 to 34, and 35 to 44. For each of the age categories, projected enrollment rates by attendance status, sex, and race/ethnicity were produced using 16 pooled time-series models—one for each combination of attendance status, sex, and the four racial/ethnic groups. Each equation included variables measuring

  • recent trends;
  • economic conditions (such as disposable income); and
  • demographic changes.

For more information on the equations used to project enrollment rates for the combinations of attendance status, sex, and race/ethnicity, see tables A-20 through A-27, under “Data and equations used for the Enrollment in Degree-Granting Institutions by Race/Ethnicity Model,” below.

The final set of projected rates by attendance status, sex, age, and race/ethnicity were controlled to enrollment rates by attendance status, sex, and age produced by the Enrollment in Degree-Granting Institutions Model to ensure consistency across models.

Step 4. Produce projections of enrollments by attendance status, sex, age category, and race/ethnicity. For each combination of attendance status, sex, age category, and race/ethnicity, enrollment projections were produced by multiplying the projected enrollment rate for that combination by projections of the total population with the corresponding combination of sex, age category, and race/ethnicity.

Data and equations used for the Enrollment in Degree-Granting Institutions by Race/Ethnicity Model

Enrollment data for degree-granting institutions by race/ethnicity. Enrollment data for 1981 to 2008 by attendance status, sex, and race/ethnicity came from the NCES Integrated Postsecondary Education Data System (IPEDS). These are universe counts. The U.S. Census Bureau, Current Population Survey was the source for enrollment estimates for 1981 to 2008 by the characteristics listed above plus age.

Population data and projections by race/ethnicity. Population counts for 1981 to 2008 came from the U.S. Census Bureau, Population Estimates series. Population projections for 2009 to 2019 are the Census Bureau’s 2008 National Population Projections of the population by sex, age and race/ethnicity (August 2008).

Projections for economic variables. The economic variables used in developing these projections were from the “U.S. Monthly Model: November 2009 Short-Term Projections” from the economic consulting firm, IHS Global Insight.

Estimated equations and model statistics. Tables A-20 through A-27 show the estimated equations and model statistics used to project enrollment rates for the various combinations of attendance status, sex, and race/ethnicity.

Accuracy of projections for the Degree-Granting Institutions by Race/Ethnicity Model

Because this is the fifth edition of Projections of Education Statistics to include enrollment projections by race/ethnicity projections, there are too few years of data to present the MAPEs.

The First-Time Freshmen Enrollment in Degree-Granting Institutions Model

The First-Time Freshmen Enrollment in Degree-Granting Institutions Model produced projections of first-time freshman enrollment in degree-granting institutions by sex.

Steps used in the First-Time Freshmen Enrollment in Degree-Granting Institutions Model

The projections were produced in the following manner:

Step 1. Calculate the ratio of first-time freshman enrollment to undergraduate enrollment. For 1975 to 2008, the ratio of first-time freshmen enrollment to undergraduate enrollment was calculated for males and females.

Step 2. Project the ratio of first-time freshmen enrollment to undergraduate enrollment. Separately for males and for females, the ratio was projected using single exponential smoothing with a smoothing constant of 0.4, yielding a constant value for males and a constant value for females over the forecast period.

Step 3. Apply the projected ratio to projected undergraduate enrollment. For each sex, the projected ratio was applied to projections of undergraduate enrollment by sex produced by the Enrollment in Degree-Granting Institutions Model to yield projections of first-time freshman enrollment.

Assumptions underlying this method

This method assumes that the future pattern in the trend of first-time freshmen enrollment will be the same as that for undergraduate enrollment.

Data used in the First-Time Freshmen Enrollment in Degree-Granting Institutions Model

Undergraduate and freshmen enrollment data for degree-granting institutions. Undergraduate and freshmen enrollment data by sex for 1975 to 2008 came from the NCES Integrated Postsecondary Education Data System (IPEDS).

Projections of undergraduate enrollment. Projections of undergraduate enrollment by sex came from the Enrollment in Degree-Granting Institutions Model, discussed earlier in this section of appendix A.

Accuracy of projections for the First-Time Freshmen Enrollment Model

Because this is the second edition of Projections of Education Statistics to include projections of first-time freshmen, there are too few years of data to present the MAPEs.

Top

Table A-14. Actual and projected numbers for college enrollment rates, by sex, attendance status, and age: Fall 2008, 2014, and 2019

Sex, attendance status, and age Actual 2008   Projected
  2014 2019
Men        
Full-time        
16 years old 0.5   0.4 0.4
17 years old 2.7   2.5 2.7
18 years old 30.2   31.3 32.9
19 years old 40.2   41.4 43.2
20 years old 35.5   36.4 38.1
21 years old 30.9   32.5 34.1
22 years old 20.2   22.7 23.9
23 years old 15.2   15.7 16.7
24 years old 13.3   12.8 13.7
25 to 29 years old 5.2   5.7 6.1
30 to 34 years old 2.3   2.5 2.6
35 to 44 years old 1.2   1.4 1.5
         
Part-time        
16 years old #   0.1 0.1
17 years old 0.8   0.9 0.9
18 years old 6.1   6.3 6.3
19 years old 6.4   7.0 6.9
20 years old 9.4   8.7 8.7
21 years old 6.3   6.9 6.9
22 years old 7.3   7.7 7.7
23 years old 7.6   7.7 7.8
24 years old 7.1   7.5 7.6
25 to 29 years old 5.5   5.3 5.4
30 to 34 years old 4.1   4.1 4.2
35 to 44 years old 3.8   4.0 4.1
         
Women        
Full-time        
16 years old 0.6   0.5 0.5
17 years old 3.6   4.2 4.8
18 years old 42.8   44.5 47.7
19 years old 46.2   49.5 52.3
20 years old 43.6   46.1 49.0
21 years old 40.8   41.6 44.5
22 years old 25.8   28.1 30.6
23 years old 18.7   19.7 21.6
24 years old 15.0   15.1 16.8
25 to 29 years old 7.1   7.8 8.7
30 to 34 years old 3.1   3.5 4.0
35 to 44 years old 2.6   3.0 3.4
         
Part-time        
16 years old 0.1   0.2 0.2
17 years old 0.4   0.8 0.9
18 years old 6.3   6.4 6.7
19 years old 10.6   11.1 11.5
20 years old 10.6   11.2 11.6
21 years old 9.4   9.9 10.4
22 years old 10.9   11.5 12.3
23 years old 12.0   12.8 13.8
24 years old 11.4   11.6 12.7
25 to 29 years old 7.9   8.3 9.2
30 to 34 years old 5.7   5.7 6.3
35 to 44 years old 7.3   8.2 9.1
# Rounds to zero.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Integrated Postsecondary Education Data System, Spring 2009; Enrollment in Degree-Granting Institutions Model, 1980–2008; and U.S. Department of Commerce, Census Bureau, Current Population Reports, "Social and Economic Characteristics of Students," 2008. (This table was prepared February 2010.)

 

Top

Table A-15. Estimated equations and model statistics for full-time and part-time college enrollment rates of men

Independent variable Coefficient Standard error t-statistic R2 D.W. statistic
Full-time          
Intercept term for 17-year-olds -7.59 0.287 -26.42 1.00 1.9*
Intercept term for 18-year-olds -4.75 0.284 -16.71    
Intercept term for 19-year-olds -4.34 0.214 -20.24    
Intercept term for 20-year-olds -4.43 0.216 -20.49    
Intercept term for 21-year-olds -4.59 0.216 -21.24    
Intercept term for 22-year-olds -5.05 0.228 -22.17    
Intercept term for 23-year-olds -5.51 0.217 -25.47    
Intercept term for 24-year-olds -5.88 0.223 -26.35    
Intercept term for 25- to 29-year-olds -6.60 0.219 -30.09    
Intercept term for 30- to 34-year-olds -7.59 0.238 -31.83    
Intercept term for 35- to 44-year-olds -8.11 0.251 -32.36    
Log of three-period weighted average of per capita          
disposable income in 2000 dollars, using the present          
period and the previous two periods 0.69 0.037 18.75    
Log of age-specific unemployment rate for men 0.17 0.018 9.07    
Autocorrelation coefficient for 17-year-olds 0.88 0.040 22.30    
Autocorrelation coefficient for 18-year-olds 0.91 0.043 21.31    
Autocorrelation coefficient for 19-year-olds 0.47 0.122 3.83    
Autocorrelation coefficient for 20-year-olds 0.59 0.123 4.84    
Autocorrelation coefficient for 21-year-olds 0.57 0.122 4.71    
Autocorrelation coefficient for 22-year-olds 0.83 0.121 6.88    
Autocorrelation coefficient for 23-year-olds 0.64 0.113 5.65    
Autocorrelation coefficient for 24-year-olds 0.75 0.125 5.98    
Autocorrelation coefficient for 25- to 29-year-olds 0.69 0.069 9.98    
Autocorrelation coefficient for 30- to 34-year-olds 0.85 0.077 11.05    
Autocorrelation coefficient for 35- to 44-year-olds 0.87 0.068 12.73    
Part-time          
Intercept term for 17-year-olds -7.66 0.364 -21.06 0.99 1.9*
Intercept term for 18-year-olds -5.25 0.299 -17.57    
Intercept term for 19-year-olds -4.79 0.394 -12.17    
Intercept term for 20-year-olds -4.71 0.320 -14.71    
Intercept term for 21-year-olds -4.90 0.286 -17.13    
Intercept term for 22-year-olds -5.04 0.292 -17.28    
Intercept term for 23-year-olds -5.08 0.284 -17.89    
Intercept term for 24-year-olds -5.13 0.285 -18.01    
Intercept term for 25- to 29-year-olds -5.58 0.301 -18.54    
Intercept term for 30- to 34-year-olds -5.97 0.296 -20.19    
Intercept term for 35- to 44-year-olds -5.94 0.285 -20.82    
Log of three-period weighted average of per capita          
disposable income in 2000 dollars, using the present          
period and the previous two periods 0.40 0.045 8.86    
Log of unemployment rate 0.15 0.024 5.98    
Autocorrelation coefficient for 17-year-olds 0.73 0.114 6.39    
Autocorrelation coefficient for 18-year-olds 0.78 0.078 10.01    
Autocorrelation coefficient for 19-year-olds 0.94 0.056 16.75    
Autocorrelation coefficient for 20-year-olds 0.84 0.104 8.08    
Autocorrelation coefficient for 21-year-olds 0.57 0.094 6.01    
Autocorrelation coefficient for 22-year-olds 0.72 0.079 9.20    
Autocorrelation coefficient for 23-year-olds 0.42 0.097 4.36    
Autocorrelation coefficient for 24-year-olds 0.54 0.110 4.87    
Autocorrelation coefficient for 25- to 29-year-olds 0.89 0.049 18.02    
Autocorrelation coefficient for 30- to 34-year-olds 0.88 0.044 20.01    
Autocorrelation coefficient for 35- to 44-year-olds 0.60 0.063 9.48    
* p <05.
R2 = Coefficient of determination.
D.W.statistic = Durbin-Watson statistic, a test for autocorrelation among regression residuals. For more details, see Johnston, J., and Dinardo, J. (1996). Econometric Methods, New York: McGraw-Hill.
NOTE: The regression method used to estimate the full-time and part-time equations was the pooled seemingly unrelated regression method with a first-order autocorrelation correction. The time period used to estimate both equations is from 1980 to 2008, and the number of observations is 308 after the correction for autocorrelation. For additional information, see Intriligator, M.D. (1978). Econometric Models, Techniques, & Applications. New Jersey: Prentice-Hall, Inc., pp. 165–173.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Enrollment in Degree-Granting Institutions Model, 1980–2008. (This table was prepared January 2010.)

 

Top

Table A-16. Estimated equations and model statistics for full-time and part-time college enrollment rates of women

Independent variable Coefficient Standard error t-statistic R2 D.W. statistic
Full-time          
Intercept term for 17-year-olds -9.30 0.163 -57.05 1.00 1.31**
Intercept term for 18-year-olds -6.56 0.143 -45.81    
Intercept term for 19-year-olds -6.44 0.138 -46.54    
Intercept term for 20-year-olds -6.62 0.137 -48.26    
Intercept term for 21-year-olds -6.84 0.137 -49.82    
Intercept term for 22-year-olds -7.50 0.139 -54.04    
Intercept term for 23-year-olds -8.01 0.141 -57.00    
Intercept term for 24-year-olds -8.39 0.142 -59.16    
Intercept term for 25- to 29-year-olds -9.07 0.141 -64.39    
Intercept term for 30- to 34-year-olds -9.77 0.139 -70.34    
Intercept term for 35- to 44-year-olds -9.99 0.139 -71.78    
Log of three-period weighted average of per capita          
disposable income in 2000 dollars, using the present          
period and the previous two periods 1.11 0.029 38.30    
Log of age-specific unemployment rate for women 0.16 0.039 4.00    
Part-time          
Intercept term for 17-year-olds -11.09 0.436 -25.42 0.99 1.91*
Intercept term for 18-year-olds -8.53 0.344 -24.80    
Intercept term for 19-year-olds -7.99 0.327 -24.43    
Intercept term for 20-year-olds -8.13 0.324 -25.09    
Intercept term for 21-year-olds -8.16 0.320 -25.52    
Intercept term for 22-year-olds -8.32 0.320 -25.97    
Intercept term for 23-year-olds -8.39 0.322 -26.07    
Intercept term for 24-year-olds -8.44 0.325 -25.97    
Intercept term for 25- to 29-year-olds -8.91 0.335 -26.61    
Intercept term for 30- to 34-year-olds -9.38 0.334 -28.07    
Intercept term for 35- to 44-year-olds -9.09 0.345 -26.37    
Log of three-period weighted average of per capita          
disposable income in 2000 dollars, using the present          
period and the previous two periods 1.01 0.050 20.05    
Log of unemployment rate 0.14 0.026 5.59    
Autocorrelation coefficient for 17-year-olds 0.76 0.080 9.42    
Autocorrelation coefficient for 18-year-olds 0.78 0.076 10.29    
Autocorrelation coefficient for 19-year-olds 0.76 0.073 10.30    
Autocorrelation coefficient for 20-year-olds 0.65 0.101 6.38    
Autocorrelation coefficient for 21-year-olds 0.36 0.129 2.79    
Autocorrelation coefficient for 22-year-olds 0.47 0.098 4.80    
Autocorrelation coefficient for 23-year-olds 0.52 0.076 6.88    
Autocorrelation coefficient for 24-year-olds 0.73 0.064 11.33    
Autocorrelation coefficient for 25- to 29-year-olds 0.89 0.034 26.37    
Autocorrelation coefficient for 30- to 34-year-olds 0.91 0.025 36.32    
Autocorrelation coefficient for 35- to 44-year-olds 0.92 0.025 36.80    
* p <05.
** Inconclusive.
R2 = Coefficient of determination.
D.W.statistic = Durbin-Watson statistic, a test for autocorrelation among regression residuals. For more details, see Johnston, J., and Dinardo, J. (1996). Econometric Methods. New York: McGraw-Hill.
NOTE: The regression method used to estimate the full-time equation was the pooled seemingly unrelated regression method. The regression method used to estimate the part-time equation was the pooled seemingly unrelated regression method with a first-order autocorrelation correction. The time period used to estimate both equations is from 1980 to 2008. The number of observations for the full-time equation is 319 and the number of observations for the part-time equation, after the correction for autocorrelation, is 308. For additional information, see Intriligator, M.D. (1978). Econometric Models, Techniques, & Applications. New Jersey: Prentice-Hall, Inc., pp. 165–173.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Enrollment in Degree-Granting Institutions Model, 1980–2008. (This table was prepared January 2010.)

 

Top

Table A-17. Actual and projected numbers for the percentage distribution of full-time students at degree-granting institutions, by sex and age group: Fall 2008, and 2009 through 2019

Age and institution type Men   Women
Actual 2008 Projected
2009 through 2019
Actual 2008 Projected
2009 through 2019
18 and 19 years old          
Undergraduate, 4-year institutions 66.4 65.8   67.4 67.7
Undergraduate, 2-year institutions 33.4 33.9   32.8 32.1
Postbaccalaureate, 4-year institutions 0.3 0.3   # 0.3
           
20 and 21 years old          
Undergraduate, 4-year institutions 77.6 77.5   79.7 79.5
Undergraduate, 2-year institutions 20.6 20.7   18.4 18.6
Postbaccalaureate, 4-year institutions 1.8 1.8   1.9 2.0
           
22 to 24 years old          
Undergraduate, 4-year institutions 62.1 64.1   60.3 60.4
Undergraduate, 2-year institutions 17.5 16.8   16.7 16.9
Postbaccalaureate, 4-year institutions 20.4 19.1   23.1 22.7
           
25 to 29 years old          
Undergraduate, 4-year institutions 44.9 43.5   39.7 40.2
Undergraduate, 2-year institutions 18.3 18.1   21.7 22.7
Postbaccalaureate, 4-year institutions 36.8 38.4   38.6 37.0
           
30 to 34 years old          
Undergraduate, 4-year institutions 38.7 36.9   39.7 39.4
Undergraduate, 2-year institutions 23.2 21.6   29.0 30.7
Postbaccalaureate, 4-year institutions 38.1 41.4   31.3 29.9
           
35 years and over          
Undergraduate, 4-year institutions 39.9 40.2   45.4 43.4
Undergraduate, 2-year institutions 20.0 23.2   28.7 30.1
Postbaccalaureate, 4-year institutions 40.1 36.6   25.9 26.5
# Rounds to zero.
NOTE: Detail may not sum to totals because of rounding.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Integrated Postsecondary Education Data System, Spring 2009; Enrollment in Degree-Granting Institutions Model, 1980–2008; and U.S. Department of Commerce, Census Bureau, Current Population Reports, "Social and Economic Characteristics of Students," 2008. (This table was prepared February 2010.)

 

Top

Table A-18. Actual and projected numbers for the percentage distribution of part-time students at degree-granting institutions, by sex and age group: Fall 2008, and 2009 through 2019

Age and institution type Men   Women
Actual 2008 Projected
2009 through 2019
Actual 2008 Projected
2009 through 2019
18 and 19 years old          
Undergraduate, 4-year institutions 14.9 17.6   20.4 20.4
Undergraduate, 2-year institutions 85.0 82.1   79.9 79.4
Postbaccalaureate, 4-year institutions 0.1 0.3   # 0.3
           
20 and 21 years old          
Undergraduate, 4-year institutions 33.3 30.9   30.4 31.5
Undergraduate, 2-year institutions 66.2 68.7   69.4 67.8
Postbaccalaureate, 4-year institutions 0.6 0.4   0.2 0.7
           
22 to 24 years old          
Undergraduate, 4-year institutions 31.8 32.5   31.0 29.6
Undergraduate, 2-year institutions 61.2 59.4   55.3 57.2
Postbaccalaureate, 4-year institutions 7.0 8.1   13.7 13.1
           
25 to 29 years old          
Undergraduate, 4-year institutions 25.3 26.3   26.0 24.7
Undergraduate, 2-year institutions 55.6 54.1   53.4 53.5
Postbaccalaureate, 4-year institutions 19.1 19.6   20.5 21.8
           
30 to 34 years old          
Undergraduate, 4-year institutions 25.5 24.5   26.4 25.1
Undergraduate, 2-year institutions 51.4 49.9   49.5 50.8
Postbaccalaureate, 4-year institutions 23.1 25.6   24.1 24.2
           
35 years and over          
Undergraduate, 4-year institutions 28.4 25.7   25.4 24.4
Undergraduate, 2-year institutions 44.7 47.3   50.0 50.9
Postbaccalaureate, 4-year institutions 26.9 26.9   24.6 24.6
# Rounds to zero.
NOTE: Detail may not sum to totals because of rounding.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Integrated Postsecondary Education Data System, Spring 2009; Enrollment in Degree-Granting Institutions Model, 1980–2008; and U.S. Department of Commerce, Census Bureau, Current Population Reports, "Social and Economic Characteristics of Students," 2008. (This table was prepared February 2010.)

 

Top

Table A-19. Actual and projected numbers for enrollment in public degree-granting institutions as a percent of total public and private enrollment, by sex, attendance status, level enrolled, and type of institution: Fall 2008, and 2009 through 2019

Enrollment category Men   Women
Actual 2008 Projected
2009 through 2019
Actual 2008 Projected
2009 through 2019
Full-time, undergraduate, 4-year institutions 65.1 65.5   61.1 62.2
Part-time, undergraduate, 4-year institutions 67.5 69.3   63.4 65.8
Full-time, undergraduate, 2-year institutions 92.2 92.0   88.3 89.0
Part-time, undergraduate, 2-year institutions 99.2 99.2   98.7 98.7
Full-time, postbaccalaureate, 4-year institutions 49.0 49.0   46.4 46.4
Part-time, postbaccalaureate, 4-year institutions 53.0 53.0   54.5 54.5
SOURCE: U.S. Department of Education, National Center for Education Statistics, Integrated Postsecondary Education Data System, Spring 2009; and Enrollment in Degree-Granting Institutions Model, 1980–2008. (This table was prepared February 2010.)

 

Top

Table A-20. Estimated equations and model statistics for full-time and part-time college enrollment rates of White men

Independent variable Coefficient Standard error t-statistic R2 D.W. statistic
Full-time          
Intercept term for 17-year-olds -8.46 0.166 -51.11 1.00 1.50*
Intercept term for 18-year-olds -5.45 0.145 -37.70    
Intercept term for 19-year-olds -5.23 0.140 -37.24    
Intercept term for 20-year-olds -5.42 0.141 -38.53    
Intercept term for 21-year-olds -5.57 0.141 -39.62    
Intercept term for 22-year-olds -6.05 0.141 -42.80    
Intercept term for 23-year-olds -6.61 0.141 -46.79    
Intercept term for 24-year-olds -7.04 0.143 -49.23    
Intercept term for 25- to 29-year-olds -7.88 0.141 -55.81    
Intercept term for 30- to 34-year-olds -8.93 0.144 -62.01    
Intercept term for 35- to 44-year-olds -9.55 0.146 -65.45    
Log of White per capita disposable income in current dollars 0.25 0.007 34.39    
Part-time          
Intercept term for 17-year-olds -5.24 0.153 -34.37 0.99 1.53*
Intercept term for 18-year-olds -1.91 0.077 -24.86    
Intercept term for 19-year-olds -1.50 0.092 -16.33    
Intercept term for 20-year-olds -1.52 0.077 -19.69    
Intercept term for 21-year-olds -1.56 0.078 -20.01    
Intercept term for 22-year-olds -1.72 0.076 -22.62    
Intercept term for 23-year-olds -1.76 0.071 -24.71    
Intercept term for 24-year-olds -1.79 0.070 -25.38    
Intercept term for 25- to 29-year-olds -2.10 0.070 -29.84    
Intercept term for 30- to 34-year-olds -2.55 0.073 -34.77    
Intercept term for 35- to 44-year-olds -2.60 0.069 -37.73    
Log of real total private compensation employment cost index 0.91 0.088 10.26    
* p <05.
R2 = Coefficient of determination.
D.W. statistic = Durbin-Watson statistic, a test for autocorrelation among regression residuals. For more details, see Johnston, J., and Dinardo, J. (1996). Econometric Methods. New York: McGraw-Hill.
NOTE: The regression method used to estimate the full-time and part-time equations was the pooled seemingly unrelated regression method. The time period used to estimate the equations is from 1980 to 2008. The number of observations is 319. For additional information, see Intriligator, M.D. (1978). Econometric Models, Techniques, & Applications. New Jersey: Prentice-Hall, Inc., pp. 165–173. Race categories exclude persons of Hispanic ethnicity.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Enrollment in Degree-Granting Institutions by Race/Ethnicity Model, 1980–2008. (This table was prepared January 2010.)

 

Top

Table A-21. Estimated equations and model statistics for full-time and part-time college enrollment rates of White women

Independent variable Coefficient Standard error t-statistic R2 D.W. statistic
Full-time          
Intercept term for 17-year-olds -13.29 0.227 -58.47 1.00 1.49*
Intercept term for 18-year-olds -10.35 0.207 -49.96    
Intercept term for 19-year-olds -10.24 0.204 -50.22    
Intercept term for 20-year-olds -10.48 0.204 -51.30    
Intercept term for 21-year-olds -10.74 0.205 -52.45    
Intercept term for 22-year-olds -11.47 0.206 -55.76    
Intercept term for 23-year-olds -12.05 0.207 -58.15    
Intercept term for 24-year-olds -12.45 0.207 -60.12    
Intercept term for 25- to 29-year-olds -13.25 0.205 -64.51    
Intercept term for 30- to 34-year-olds -13.96 0.205 -68.27    
Intercept term for 35- to 44-year-olds -14.15 0.205 -69.04    
Log of White per capita disposable income in current dollars 0.52 0.011 49.67    
Part-time          
Intercept term for 17-year-olds -8.96 0.277 -32.32 0.81 1.60*
Intercept term for 18-year-olds -5.78 0.220 -26.29    
Intercept term for 19-year-olds -5.34 0.223 -23.90    
Intercept term for 20-year-olds -5.41 0.222 -24.32    
Intercept term for 21-year-olds -5.49 0.220 -24.95    
Intercept term for 22-year-olds -5.67 0.219 -25.88    
Intercept term for 23-year-olds -5.73 0.219 -26.14    
Intercept term for 24-year-olds -5.75 0.218 -26.30    
Intercept term for 25- to 29-year-olds -6.07 0.217 -27.97    
Intercept term for 30- to 34-year-olds -6.40 0.219 -29.20    
Intercept term for 35- to 44-year-olds -6.08 0.217 -27.97    
Log of White per capita disposable income in current dollars 0.18 0.011 16.15    
* p <05.
R2 = Coefficient of determination.
D.W. statistic = Durbin-Watson statistic, a test for autocorrelation among regression residuals. For more details, see Johnston, J., and Dinardo, J. (1996). Econometric Methods. New York: McGraw-Hill.
NOTE: The regression method used to estimate the full-time and part-time equations was the pooled seemingly unrelated regression method. The time period used to estimate the equations is from 1980 to 2008. The number of observations is 319. For additional information, see Intriligator, M.D. (1978). Econometric Models, Techniques, & Applications. New Jersey: Prentice-Hall, Inc., pp. 165–173. Race categories exclude persons of Hispanic ethnicity.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Enrollment in Degree-Granting Institutions by Race/Ethnicity Model, 1980–2008. (This table was prepared January 2010.)

 

Top

Table A-22. Estimated equations and model statistics for full-time and part-time college enrollment rates of Black men

Independent variable Coefficient Standard error t-statistic R2 D.W. statistic
Full-time          
Intercept term for 17-year-olds -9.66 0.436 -22.16 0.97 1.58*
Intercept term for 18-year-olds -7.45 0.430 -17.32    
Intercept term for 19-year-olds -7.20 0.430 -16.74    
Intercept term for 20-year-olds -7.29 0.430 -16.93    
Intercept term for 21-year-olds -7.56 0.431 -17.57    
Intercept term for 22-year-olds -7.73 0.432 -17.89    
Intercept term for 23-year-olds -8.23 0.436 -18.91    
Intercept term for 24-year-olds -8.51 0.433 -19.68    
Intercept term for 25- to 29-year-olds -9.28 0.433 -21.45    
Intercept term for 30- to 34-year-olds -10.08 0.438 -23.02    
Intercept term for 35- to 44-year-olds -10.45 0.434 -24.07    
Log of Black per capita disposable income in current dollars 0.32 0.023 13.61    
Part-time          
Intercept term for 17-year-olds -9.85 0.334 -29.48 0.66 1.85*
Intercept term for 18-year-olds -9.01 0.365 -24.71    
Intercept term for 19-year-olds -8.17 0.342 -23.93    
Intercept term for 20-year-olds -8.12 0.339 -23.98    
Intercept term for 21-year-olds -8.08 0.329 -24.60    
Intercept term for 22-year-olds -8.15 0.345 -23.64    
Intercept term for 23-year-olds -8.37 0.352 -23.75    
Intercept term for 24-year-olds -8.38 0.354 -23.68    
Intercept term for 25- to 29-year-olds -8.47 0.327 -25.89    
Intercept term for 30- to 34-year-olds -8.68 0.326 -26.66    
Intercept term for 35- to 44-year-olds -8.75 0.323 -27.12    
Log of Black per capita disposable income in current dollars 0.28 0.017 16.10    
* p <05.
R2 = Coefficient of determination.
D.W. statistic = Durbin-Watson statistic, a test for autocorrelation among regression residuals. For more details, see Johnston, J., and Dinardo, J. (1996). Econometric Methods. New York: McGraw-Hill.
NOTE: The regression method used to estimate the full-time and part-time equations was the pooled seemingly unrelated regression method. The time period used to estimate the equations is from 1980 to 2008. The number of observations is 319. For additional information, see Intriligator, M.D. (1978). Econometric Models, Techniques, & Applications. New Jersey: Prentice-Hall, Inc., pp. 165–173. Race categories exclude persons of Hispanic ethnicity.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Enrollment in Degree-Granting Institutions by Race/Ethnicity Model, 1980–2008. (This table was prepared January 2010.)

 

Top

Table A-23. Estimated equations and model statistics for full-time and part-time college enrollment rates of Black women

Independent variable Coefficient Standard error t-statistic R2 D.W. statistic
Full-time          
Intercept term for 17-year-olds -14.59 0.601 -24.27 0.97 1.77*
Intercept term for 18-year-olds -12.44 0.595 -20.90    
Intercept term for 19-year-olds -12.23 0.594 -20.58    
Intercept term for 20-year-olds -12.50 0.595 -21.02    
Intercept term for 21-year-olds -12.64 0.594 -21.26    
Intercept term for 22-year-olds -13.11 0.594 -22.05    
Intercept term for 23-year-olds -13.40 0.595 -22.52    
Intercept term for 24-year-olds -13.76 0.596 -23.07    
Intercept term for 25- to 29-year-olds -14.54 0.597 -24.38    
Intercept term for 30- to 34-year-olds -15.02 0.595 -25.24    
Intercept term for 35- to 44-year-olds -15.39 0.595 -25.85    
Log of Black per capita disposable income in current dollars 0.62 0.032 19.29    
Part-time          
Intercept term for 17-year-olds -14.11 0.622 -22.69 0.64 1.79*
Intercept term for 18-year-olds -12.69 0.622 -20.42    
Intercept term for 19-year-olds -12.31 0.621 -19.82    
Intercept term for 20-year-olds -12.35 0.620 -19.93    
Intercept term for 21-year-olds -12.26 0.620 -19.76    
Intercept term for 22-year-olds -12.25 0.620 -19.74    
Intercept term for 23-year-olds -12.24 0.620 -19.76    
Intercept term for 24-year-olds -12.43 0.620 -20.05    
Intercept term for 25- to 29-year-olds -12.60 0.615 -20.49    
Intercept term for 30- to 34-year-olds -12.72 0.616 -20.65    
Intercept term for 35- to 44-year-olds -12.59 0.615 -20.48    
Log of Black per capita disposable income in current dollars 0.53 0.033 15.95    
* p <05.
R2 = Coefficient of determination.
D.W. statistic = Durbin-Watson statistic, a test for autocorrelation among regression residuals. For more details, see Johnston, J., and Dinardo, J. (1996). Econometric Methods. New York: McGraw-Hill.
NOTE: The regression method used to estimate the full-time and part-time equations was the pooled seemingly unrelated regression method. The time period used to estimate the equations is from 1980 to 2008. The number of observations is 319. For additional information, see Intriligator, M.D. (1978). Econometric Models, Techniques, & Applications. New Jersey: Prentice-Hall, Inc., pp. 165–173. Race categories exclude persons of Hispanic ethnicity.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Enrollment in Degree-Granting Institutions by Race/Ethnicity Model, 1980–2008. (This table was prepared January 2010.)

 

Top

Table A-24. Estimated equations and model statistics for full-time and part-time college enrollment rates of Hispanic men

Independent variable Coefficient Standard error t-statistic R2 D.W. statistic
Full-time          
Intercept term for 17-year-olds -10.64 0.550 -19.34 0.94 1.92*
Intercept term for 18-year-olds -8.69 0.541 -16.05    
Intercept term for 19-year-olds -8.47 0.541 -15.66    
Intercept term for 20-year-olds -8.69 0.541 -16.07    
Intercept term for 21-year-olds -8.90 0.544 -16.34    
Intercept term for 22-year-olds -9.37 0.543 -17.25    
Intercept term for 23-year-olds -9.66 0.544 -17.77    
Intercept term for 24-year-olds -9.81 0.543 -18.08    
Intercept term for 25- to 29-year-olds -10.64 0.544 -19.57    
Intercept term for 30- to 34-year-olds -11.44 0.545 -20.99    
Intercept term for 35- to 44-year-olds -11.95 0.550 -21.73    
Log of Hispanic per capita disposable income in current dollars 0.37 0.030 12.46    
Part-time          
Intercept term for 17-year-olds -10.25 0.390 -26.29 0.73 1.79*
Intercept term for 18-year-olds -8.67 0.386 -22.45    
Intercept term for 19-year-olds -8.38 0.394 -21.27    
Intercept term for 20-year-olds -8.28 0.385 -21.48    
Intercept term for 21-year-olds -8.29 0.385 -21.54    
Intercept term for 22-year-olds -8.65 0.383 -22.56    
Intercept term for 23-year-olds -8.65 0.394 -21.96    
Intercept term for 24-year-olds -8.67 0.384 -22.59    
Intercept term for 25- to 29-year-olds -9.04 0.374 -24.17    
Intercept term for 30- to 34-year-olds -9.48 0.377 -25.18    
Intercept term for 35- to 44-year-olds -9.51 0.374 -25.44    
Log of Hispanic per capita disposable income in current dollars 0.31 0.020 15.12    
* p <05.
R2 = Coefficient of determination.
D.W. statistic = Durbin-Watson statistic, a test for autocorrelation among regression residuals. For more details, see Johnston, J., and Dinardo, J. (1996). Econometric Methods. New York: McGraw-Hill.
NOTE: The regression method used to estimate the full-time and part-time equations was the pooled seemingly unrelated regression method. The time period used to estimate the equations is from 1980 to 2008. The number of observations is 319. For additional information, see Intriligator, M.D. (1978). Econometric Models, Techniques, & Applications. New Jersey: Prentice-Hall, Inc., pp. 165–173.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Enrollment in Degree-Granting Institutions by Race/Ethnicity Model, 1980–2008. (This table was prepared January 2010.)

 

Top

Table A-25. Estimated equations and model statistics for full-time and part-time college enrollment rates of Hispanic women

Independent variable Coefficient Standard error t-statistic R2 D.W. statistic
Full-time          
Intercept term for 17-year-olds -17.24 0.434 -39.69 0.95 1.90*
Intercept term for 18-year-olds -14.74 0.412 -35.80    
Intercept term for 19-year-olds -14.65 0.409 -35.82    
Intercept term for 20-year-olds -14.98 0.410 -36.51    
Intercept term for 21-year-olds -15.12 0.411 -36.81    
Intercept term for 22-year-olds -15.70 0.415 -37.87    
Intercept term for 23-year-olds -16.00 0.412 -38.87    
Intercept term for 24-year-olds -16.47 0.420 -39.25    
Intercept term for 25- to 29-year-olds -17.09 0.409 -41.75    
Intercept term for 30- to 34-year-olds -17.78 0.413 -43.03    
Intercept term for 35- to 44-year-olds -18.08 0.416 -43.43    
Log of Hispanic per capita disposable income in current dollars 0.73 0.022 32.89    
Part-time          
Intercept term for 17-year-olds -15.03 0.486 -30.92 0.73 1.87*
Intercept term for 18-year-olds -12.95 0.473 -27.39    
Intercept term for 19-year-olds -12.62 0.468 -26.96    
Intercept term for 20-year-olds -12.89 0.475 -27.11    
Intercept term for 21-year-olds -12.73 0.476 -26.76    
Intercept term for 22-year-olds -13.06 0.476 -27.42    
Intercept term for 23-year-olds -12.88 0.470 -27.42    
Intercept term for 24-year-olds -13.15 0.474 -27.72    
Intercept term for 25- to 29-year-olds -13.45 0.464 -29.02    
Intercept term for 30- to 34-year-olds -13.82 0.463 -29.82    
Intercept term for 35- to 44-year-olds -13.66 0.462 -29.55    
Log of Hispanic per capita disposable income in current dollars 0.57 0.025 22.55    
* p <05.
R2 = Coefficient of determination.
D.W. statistic = Durbin-Watson statistic, a test for autocorrelation among regression residuals. For more details, see Johnston, J., and Dinardo, J. (1996). Econometric Methods. New York: McGraw-Hill.
NOTE: The regression method used to estimate the full-time and part-time equations was the pooled seemingly unrelated regression method. The time period used to estimate the equations is from 1980 to 2008. The number of observations is 319. For additional information, see Intriligator, M.D. (1978). Econometric Models, Techniques, & Applications. New Jersey: Prentice-Hall, Inc., pp. 165–173.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Enrollment in Degree-Granting Institutions by Race/Ethnicity Model, 1980–2008. (This table was prepared January 2010.)

 

Top

Table A-26. Estimated equations and model statistics for full-time and part-time college enrollment rates of Asian/Pacific Islander men

=
Independent variable Coefficient Standard error t-statistic R2 D.W. statistic
Full-time          
Intercept term for 17-year-olds -8.82 0.331 -14.87 0.94 1.92*
Intercept term for 18-year-olds -5.97 0.298 -10.11    
Intercept term for 19-year-olds -5.77 0.303 -9.69    
Intercept term for 20-year-olds -5.89 0.299 -9.94    
Intercept term for 21-year-olds -5.88 0.301 -9.87    
Intercept term for 22-year-olds -6.20 0.300 -10.48    
Intercept term for 23-year-olds -6.52 0.301 -10.88    
Intercept term for 24-year-olds -6.86 0.306 -11.46    
Intercept term for 25- to 29-year-olds -7.67 0.300 -13.19    
Intercept term for 30- to 34-year-olds -8.69 0.302 -14.98    
Intercept term for 35- to 44-year-olds -9.47 0.299 -16.47    
Log of Asian/Pacific Islander per capita disposable income          
in current dollars 0.29 0.015 19.20    
Part-time          
Intercept term for 17-year-olds -4.60 0.591 -7.78 0.71 1.85*
Intercept term for 18-year-olds -3.58 0.584 -6.12    
Intercept term for 19-year-olds -2.80 0.578 -4.84    
Intercept term for 20-year-olds -2.93 0.590 -4.96    
Intercept term for 21-year-olds -3.00 0.590 -5.09    
Intercept term for 22-year-olds -2.97 0.605 -4.91    
Intercept term for 23-year-olds -3.08 0.582 -5.29    
Intercept term for 24-year-olds -3.34 0.577 -5.80    
Intercept term for 25- to 29-year-olds -3.72 0.563 -6.60    
Intercept term for 30- to 34-year-olds -4.18 0.563 -7.42    
Intercept term for 35- to 44-year-olds -4.59 0.562 -8.18    
Log of Asian/Pacific Islander per capita disposable income          
in current dollars 0.07 0.029 2.43    
* p <05.
R2 = Coefficient of determination.
D.W. statistic = Durbin-Watson statistic, a test for autocorrelation among regression residuals. For more details, see Johnston, J., and Dinardo, J. (1996). Econometric Methods. New York: McGraw-Hill.
NOTE: The regression method used to estimate the full-time and part-time equations was the pooled seemingly unrelated regression method. The time period used to estimate the equations is from 1989 to 2008. The number of observations is 220. For additional information, see Intriligator, M.D. (1978). Econometric Models, Techniques, & Applications. New Jersey: Prentice-Hall, Inc., pp. 165–173. Race categories exclude persons of Hispanic ethnicity.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Enrollment in Degree-Granting Institutions by Race/Ethnicity Model, 1980–2008. (This table was prepared January 2010.)

 

Top

Table A-27. Estimated equations and model statistics for full-time and part-time college enrollment rates of Asian/Pacific Islander women

Independent variable Coefficient Standard error t-statistic R2 D.W. statistic
Full-time          
Intercept term for 17-year-olds -13.81 0.622 -22.18 0.96 1.97*
Intercept term for 18-year-olds -11.43 0.613 -18.66    
Intercept term for 19-year-olds -10.83 0.618 -17.51    
Intercept term for 20-year-olds -11.13 0.619 -17.98    
Intercept term for 21-year-olds -11.24 0.615 -18.28    
Intercept term for 22-year-olds -11.81 0.615 -19.21    
Intercept term for 23-year-olds -12.20 0.614 -19.88    
Intercept term for 24-year-olds -12.73 0.625 -20.37    
Intercept term for 25- to 29-year-olds -13.61 0.611 -22.29    
Intercept term for 30- to 34-year-olds -14.91 0.615 -24.22    
Intercept term for 35- to 44-year-olds -15.40 0.617 -24.96    
Log of Asian/Pacific Islander per capita disposable income          
in current dollars 0.59 0.032 18.67    
Part-time          
Intercept term for 17-year-olds -13.96 0.518 -26.94 0.86 1.95*
Intercept term for 18-year-olds -12.15 0.510 -23.82    
Intercept term for 19-year-olds -11.25 0.539 -20.88    
Intercept term for 20-year-olds -11.66 0.523 -22.31    
Intercept term for 21-year-olds -11.22 0.520 -21.59    
Intercept term for 22-year-olds -11.47 0.500 -22.94    
Intercept term for 23-year-olds -11.83 0.509 -23.25    
Intercept term for 24-year-olds -12.05 0.526 -22.91    
Intercept term for 25- to 29-year-olds -12.54 0.496 -25.30    
Intercept term for 30- to 34-year-olds -13.22 0.498 -26.53    
Intercept term for 35- to 44-year-olds -13.05 0.493 -26.49    
Log of Asian/Pacific Islander per capita disposable income          
in current dollars 0.53 0.025 20.88    
* p <05.
R2 = Coefficient of determination.
D.W. statistic = Durbin-Watson statistic, a test for autocorrelation among regression residuals. For more details, see Johnston, J., and Dinardo, J. (1996). Econometric Methods. New York: McGraw-Hill.
NOTE: The regression method used to estimate the full-time and part-time equations was the pooled seemingly unrelated regression method. The time period used to estimate the equations is from 1989 to 2008. The number of observations is 220. For additional information, see Intriligator, M.D. (1978). Econometric Models, Techniques, & Applications. New Jersey: Prentice-Hall, Inc., pp. 165–173. Race categories exclude persons of Hispanic ethnicity.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Enrollment in Degree-Granting Institutions by Race/Ethnicity Model, 1980–2008. (This table was prepared January 2010.)

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.