SARI Report 333, September 2004 Observations About Regents Scholarship Yield Rates: Change and Variance By Steve Chatman The proportion of new from high school students who accepted UC Davis Regents Scholarship offers increased sharply from 2003 to 2004. The increase was from about 8% in fall 2002 and 2003 to about 12% in 2004. This change was both statistically significant and important. Furthermore, it could not have been accurately anticipated from the existing data record. Non firstgeneration students were more likely to accept in 2004 than in the two previous years The analysis reported here will not attempt to explain why change occurred. Instead, these notes will be confined to observations about where the change occurred and what amount of change might be expected under normal circumstances. The results will attempt to achieve two goals. First, they will help to identify subpopulations where decision-making outcomes changed, and they will note changes in the group composition of students receiving scholarship offers. Second, the results will be used to establish a range within which normal variation might be expected in order to forecast yield for 2005. Information about change in decision making and in composition of scholarship offer recipients is shown in detail as Table 1. Therefore, this space will call attention to data patterns exhibiting significant change. The yield rate among the middle 50% of reported household incomes increased in 2004 The first changes noted were within the likelihood that the offered award would be accepted (yield). From fall 2002 to fall 2003, there was an increase in the proportion of award recipients who were first generation, but that change did not affect the overall yield in 2004. What did change in 2004 was the increase in likelihood that those who were NOT first generation would accept the offer. The acceptance rate among first-generation students did not change. Students from families in the financial mid-range of reported household incomes ($44,000 to $144,000) were more likely to accept a Regents Scholarship offer in fall 2004. The proportion of new freshmen accepting offers did not change in the top or bottom quartiles or for the group of students who did not report family income. To reiterate, the yield rate among the middle 50% of reported household incomes increased. Female award recipients were more likely to accept offers in fall 2004. Acceptance rates did not change among males. 1
Although the numbers were small, a statistical increase was noted for the likelihood that students from the Central Coastal region of California would accept offers. Only one of 20 Central Coast students accepted an offer in 2002 and 2003. Four of 10 accepted offers in 2004. There have also been a few significant changes in composition of students receiving offers. Over this three-year period, those offered scholarships were more likely to be first-generation college students, students with relatively low SAT Total scores (1380 and less), and students from families with incomes of $43,000 or less. If conditions affecting yield rates in fall 2004 continue, rates in fall 2005 will be in a range from 10% to 13.4% Those offered scholarships were less likely to be students with a parent who earned at least a bachelor s degree, to have top SAT scores (1500 and above), and to be from families with incomes above $144,000. There has also been a significant downward trend in the percentage of offer recipients from the San Francisco Bay area, and this is the single largest group of students. The second goal of this report is describe a range of normal variation within which change might be expected. To create a forecast for next year and to place a bound on that estimate requires assumptions about the extent to which conditions will change from this year to next year. We know that this year presented a significantly different yield rate. Can we now assume that environmental conditions and award characteristics will remain the same from this year to next year? The following estimates will assume that to be the case, but it is important that we recognize the inherent limitations of these assumptions. If in 2003 we had applied these same assumptions in order to forecast 2004, we would have predicted that the yield rate in fall 2004 would have been in the range of 7% to 10% with our best guess being 8.5%. Our forecast would have been too low. Recognizing that the circumstances of 2004 are significantly different than they were in 2003, our best guess for 2004 is that these new conditions will continue. If there is reason to believe that the 2005 award offers will be of less monetary value or prestige, then we can revert to the period from 2002 and 2003 as our base. Assuming a steady-state system no change in prestige, true or perceived value, economic conditions, or change in the process by which scholarships are awarded yield in fall 2005 will be in a range from 10% to 13.4% with a best guess being about 11.5%. 2
Percentages Over Time Table 1: Regents' Yield Has there been a change in yield over time? Accept 8% 8% 12% 56 67 99 Decline 92% 92% 88% 629 728 746 Yes; chi-square at.03 Has the change been associated with other factors known at time of offer? Composition Over Time First-Generation No 94% 88% 89% Accept 8% 8% 12% 49 55 88 Decline 92% 92% 88% 550 611 617 Yes 6% 12% 11% Accept 10% 7% 8% 4 6 7 Decline 90% 93% 92% 37 83 79 Race/Ethnicity African-American (Too Few) 1% 1% 1% 7 6 8 American-Indian (Too Few) 0% 1% 1% 0 4 7 Asian-American 44% 44% 44% Accept 4% 5% 8% 12 19 30 Decline 96% 95% 92% 291 330 345 Mexican-American/Latino(a) 3% 6% 5% Accept 9% 6% 7% 2 3 3 Decline 91% 94% 93% 21 46 41 White 40% 36% 38% Accept 15% 11% 18% 40 32 57 Decline 85% 89% 82% 235 251 263 Other 1% 1% 1% Accept 14% 13% 8% 1 1 1 Decline 86% 88% 92% 6 7 11 Unkown 10% 12% 9% Accept 0% 13% 8% 0 12 6 Decline 100% 88% 92% 70 84 73 Income Low Quartile ( < $43K) 14% 19% 17% Accept 10% 10% 15% 10 15 21 Decline 90% 90% 85% 87 138 120 2nd Quartile ($43K < x < 90K) 16% 16% 18% Accept 9% 7% 17% 10 9 26 Decline 91% 93% 83% 101 117 123 3rd Quartile ($90 < x < $144K) 16% 18% 16% Accept 12% 5% 16% 13 7 22 Decline 88% 95% 84% 99 135 115 Top Quartile ($144K<) 20% 17% 14% Accept 5% 9% 5% 7 12 6 Decline 95% 91% 95% 130 121 113 Missing 33% 30% 35% Accept 7% 10% 8% 16 24 24 Decline 93% 90% 92% 212 217 275 (continued) 3
(continued) Table 1: Regents' Yield Percentages Over Time Composition Over Time SAT Total Low Quartile (< 1380) 12% 33% 29% Accept 11% 13% 19% 9 34 47 Decline 89% 87% 81% 72 230 198 2nd Quartile (1380 < x < 1440) 24% 25% 25% Accept 13% 8% 13% 22 15 28 Decline 87% 92% 87% 141 179 181 3rd Quartile (1440 < x < 1490) 30% 20% 22% Accept 8% 8% 8% 16 13 14 Decline 92% 92% 92% 188 143 167 Top Quartile (> 1490) 34% 22% 24% Accept 4% 2% 4% 9 4 9 Decline 96% 98% 96% 223 173 195 HS GPA Low Quartile (< 4.21) 29% 21% 21% Accept 12% 10% 13% 24 17 24 Decline 88% 90% 87% 173 151 154 2nd Quartile (4.21 < x < 4.27) 21% 25% 25% Accept 7% 9% 8% 10 19 17 Decline 93% 91% 92% 131 182 192 3rd Quartile (4.27 < x < 4.32) 20% 24% 26% Accept 4% 8% 13% 5 15 29 Decline 96% 92% 87% 133 179 188 Top Quartile (> 4.32) 30% 29% 28% Accept 8% 6% 12% 17 15 28 Decline 92% 94% 88% 188 216 210 Academic Index* Low Quartile (< 9600) 22% 22% Accept 14% 16% 18 21 Decline 86% 84% 109 114 2nd Quartile (9600 < x < 9700) 24% 27% Accept 10% 16% 14 27 Decline 90% 84% 123 142 3rd Quartile (9700 < x < 9820) 27% 24% Accept 7% 12% 11 18 Decline 93% 88% 143 131 Top Quartile (> 9820) 27% 26% Accept 6% 9% 9 14 Decline 94% 91% 148 148 Sex Female 49% 51% 50% Accept 9% 8% 14% 30 31 59 Decline 91% 92% 86% 300 373 356 Male 51% 49% 50% Accept 7% 9% 9% 26 36 39 Decline 93% 91% 91% 324 355 381 (continued) 4
Percentages Over Time (continued) Table 1: Regents' Yield Composition Over Time Geographical Location in California Northern CA-Mountains (Too Few) 1% 0% 1% 6 4 5 Northern CA-Coast (Too Few) 0% 1% 0% 3 9 4 Northern CA-Sacramento Valley (Too Few) 1% 2% 0% 8 16 3 Local Area (Eld/Pla/Sac/Yol) 8% 10% 11% 17% 20% 22% 9 16 21 83% 80% 78% 45 65 76 SF Bay - Northern (Nap/Sol/Son) 2% 2% 3% 19% 40% 14% 3 8 4 81% 60% 86% 13 12 25 SF Bay (Ala/Con/Mar/SFr/SMa/StC) 52% 49% 45% 6% 5% 8% 20 19 30 94% 95% 92% 334 376 349 Central CA Coast 2% 1% 1% 11 9 10 9% 0% 40% 1 0 4 91% 100% 60% 10 9 6 San Juaquin - North (Mad/Mer/SJo/Sta) 3% 2% 4% 13% 11% 22% 3 2 8 87% 89% 78% 20 16 28 Central Valley - North (Mountains) (Too Few) 1% 0% 0% 4 1 4 San Juaquin - South 3% 5% 4% 11% 10% 30% 2 4 11 89% 90% 70% 16 37 26 Inyo Country (Too Few) 1% 0% 0% 7 2 4 Santa Barbara, Ventura, SLO 4% 2% 2% 13% 0% 15% 3 0 3 88% 100% 85% 21 13 17 Los Angeles 10% 11% 12% 5% 6% 3% 3 5 3 95% 94% 97% 62 82 96 Orange 4% 4% 5% 7% 6% 7% 2 2 3 93% 94% 93% 27 32 38 Riverside, San Bernardino 2% 2% 2% 15% 14% 26% 2 2 5 85% 86% 74% 11 12 14 San Diego, Imperial 6% 6% 6% 5% 2% 6% 2 1 3 95% 98% 94% 36 50 50 *Academic Performance Index Computation changed from fall 2002 to fall 2003. Source: SARI Report 333 5