X hits on this document

PDF document

The Characteristics Associated with Student - page 37 / 94

301 views

0 shares

0 downloads

0 comments

37 / 94

Number of Withdrawals

In reality, this variable conveys little more information than the Withdrawal Indicator variable. For both variables, the borrowers who did not withdraw from the university – the missing category -- have a 4.0 percent default rate; the borrowers who withdraw have much higher default rates. While the table below shows that some categories of borrowers have extremely high default rates, the pattern is not consistent. In some cases, a higher number of withdrawals is associated with a higher default rate; but in other cases, it is related to a lower default rate. This lack of a consistent pattern tends to undermine this variable as a predictor of default risk. Nevertheless, the number of withdrawals that a borrower experiences has a fairly weak but statistically significant relationship to default.

10,441

96.0

439

4.0

10,880

1,390

91.8

124

8.2

1,514

232

92.4

19

7.6

251

73

89.0

9

11.0

82

21

75.0

7

25.0

28

19

90.5

2

9.5

21

12,176

95.3

600

4.7

12,776

Number of Withdrawals

0 1 2 3 4 5 or more All Undergraduates

Total

% of

% of

row

N

row

N

No

Default

Yes

Admission Code

With a default rate of 3.4 percent, only borrowers who were “Accepted by Academics”, based upon having high class ranks and high SAT scores, have a rate that is below the average for Texas A&M. In contrast, borrowers who were only accepted after a closer review of their college applications and high school transcripts have a considerably higher default rate of 5.4 percent. Borrowers accepted on a provisional basis, because they did not meet the coursework requirements and the academic credentials for admission “by Academics” or “by Review”, have a much higher 8.8 percent default rate. The relationship between Admission Code and default is statistically significant. This finding reinforces the idea that the academic preparedness of students influences whether they default on student loans.

Borrowers who have missing admission codes belong disproportionately to the earlier cohorts, which have higher default rates than later cohorts. This might explain why the missing category in the table has a higher than average default rate.

34

Document info
Document views301
Page views301
Page last viewedWed Dec 07 13:10:56 UTC 2016
Pages94
Paragraphs7456
Words29247

Comments