%0 Thesis %1 sinha_estimating_1997 %A Sinha, Kimberly Wright %C Bowling Green, OH %D 1997 %K DEMOGRAPHY EDUCATION, FINANCE, Higher, SOCIOLOGY, %T Estimating first-time enrollment for the private, highly selective, national university: A market demography application of logistic regression-based price sensitivity analysis %X Applied demography does not offer a suitable method for predicting first-time enrollment at the private, highly-selective university. While college enrollment projection techniques appear in the literature, they have limited applicability to institutions where significant nondemograpkic factors, most notably student financial aid, affect enrollment. Research on enrollment response to financial aid comes from a variety of disciplines, but it is problematic in several ways. First, many studies are national in scope. Second, they vary regarding what is measured (e.g., access versus choice) and how it is measured (e.g., student price response coefficients versus delta-P statistics). Third and most importantly, the research does not effectively reveal the bottom line. For institutional enrollment managers, that bottom line is "Will we get the class, and at what cost?" This study illustrated the use of logistic regression-based financial aid sensitivity analysis as a method for estimating first-time enrollment and financial aid expenditures. Using admissions and financial aid data for a real university, a logistic regression model was fitted. Variables found to be significant predictors of enrollment were demonstrated need, predicted grade point average, total gift aid, financial aid applicant status, gender, early decision status, aid counter offer status, state residence status, and enrollment year. The model included several significant interaction terms. Standard logistic regression diagnostics gave evidence that the model provided a good fit with the data. The model was activated in a sensitivity analysis, where varying levels of institutional gift aid could be examined for their impact on enrollment and expenditures. When observed gift aid was used as input, the sensitivity analysis predicted a total enrollment of 317 students, compared to the observed enrollment of 327, a difference of \$-\$3.1\%. Similarly, the estimated total financial aid expenditure differed from the observed expenditure by \$-\$4.0\%. The adequacy of the simulation spreadsheet as a strategic planning tool was evaluated with respect to the case institution's financial aid objectives and guiding principles. Generally speaking, at need levels of \$8,000 or more, differences between planned and actual expenditures were small, and observed and expected yields were generally in line with one another.

@phdthesis{sinha_estimating_1997, abstract = {Applied demography does not offer a suitable method for predicting first-time enrollment at the private, highly-selective university. While college enrollment projection techniques appear in the literature, they have limited applicability to institutions where significant nondemograpkic factors, most notably student financial aid, affect enrollment. Research on enrollment response to financial aid comes from a variety of disciplines, but it is problematic in several ways. First, many studies are national in scope. Second, they vary regarding what is measured (e.g., access versus choice) and how it is measured (e.g., student price response coefficients versus delta-P statistics). Third and most importantly, the research does not effectively reveal the bottom line. For institutional enrollment managers, that bottom line is "Will we get the class, and at what cost?" This study illustrated the use of logistic regression-based financial aid sensitivity analysis as a method for estimating first-time enrollment and financial aid expenditures. Using admissions and financial aid data for a real university, a logistic regression model was fitted. Variables found to be significant predictors of enrollment were demonstrated need, predicted grade point average, total gift aid, financial aid applicant status, gender, early decision status, aid counter offer status, state residence status, and enrollment year. The model included several significant interaction terms. Standard logistic regression diagnostics gave evidence that the model provided a good fit with the data. The model was activated in a sensitivity analysis, where varying levels of institutional gift aid could be examined for their impact on enrollment and expenditures. When observed gift aid was used as input, the sensitivity analysis predicted a total enrollment of 317 students, compared to the observed enrollment of 327, a difference of \$-\$3.1\%. Similarly, the estimated total financial aid expenditure differed from the observed expenditure by \$-\$4.0\%. The adequacy of the simulation spreadsheet as a strategic planning tool was evaluated with respect to the case institution's financial aid objectives and guiding principles. Generally speaking, at need levels of \$8,000 or more, differences between planned and actual expenditures were small, and observed and expected yields were generally in line with one another.}, added-at = {2018-06-19T15:18:45.000+0200}, address = {Bowling Green, OH}, author = {Sinha, Kimberly Wright}, biburl = {https://www.bibsonomy.org/bibtex/2babef944254bf002214bd22a1421e94c/prophe}, interhash = {928198b24e70364344a012382349c04b}, intrahash = {babef944254bf002214bd22a1421e94c}, keywords = {DEMOGRAPHY EDUCATION, FINANCE, Higher, SOCIOLOGY,}, school = {Bowling Green State University}, shorttitle = {Estimating first-time enrollment for the private, highly selective, national university: {A} market demography application of logistic regression-based price sensitivity analysis}, timestamp = {2018-06-19T15:18:45.000+0200}, title = {Estimating first-time enrollment for the private, highly selective, national university: {A} market demography application of logistic regression-based price sensitivity analysis}, type = {{PhD} {Thesis}}, year = 1997 }