• No se han encontrado resultados

MARCO TEORICO

1.4 TENDENCIAS RECIENTES SOBRE EL LIDERAZGO

Table 3 presents a comprehensive list of summary statistics for the full analytic sample of challenge exam takers and broken out by SDC broad subject and key courses, separately. Column 1 describes the full, pooled analytic sample. Column 2 describes “Math” participants in College

Algebra, Pre-Calculus, and Statistics, while column 3 describes characteristics of just College Algebra

likely to have historically higher challenge exam pass rates relative to the schools that decide to no longer participate in Tennessee’s SDC program.

exam takers (who comprise 70.5 percent of all Math participants). In a similar fashion, column 4 provides figures for “CTE” exam participants in the courses Criminal Justice I, Health Information Technology, Agricultural Business, and Plant Science. Column 5 describes just Plant Science

participants (representing 76.0 percent of all CTE). Students may appear in more than one column if they took multiple SDC courses.

Variables of interest are grouped into four panels. Student-level demographic and

achievement variables include means for gender, race and ethnicity, disability and Ell status, as well as prior Algebra I EOC achievement.16 Means and standard deviations are presented for students’

standardized Algebra I scores. A student’s score is standardized (i.e., mean=0, sd=1) within the population of all students in the state of Tennessee (TN) who took the Algebra I EOC in the same year. Column 1 in the top panel indicates that, overall, SDC examinees in the analytic sample are mostly White (87.2 percent) 12th graders17 (71.0 percent), who are about equally likely to be male or

female. They also perform, on net, a little more than half a standard deviation above (0.56) the statewide mean on the Algebra I EOC exam.

The demographic profile of students in the analytic sample aligns in some ways with that of the broader set of all Tennessee high school students in the academic years covered (not shown). Still, along several dimensions, some stark differences emerge. The share of Tennessee high school students who are White in these years is about 69 percent, indicating that the analytic sample share is about 18 percentage points higher. In the opposite direction, about 24 percent of Tennessee high school students are Black, while in the analytic sample, the share is only 6.6 percent. The analytic

16 I do not have student-level data on eligibility for free or reduced-price meals – an indication of socioeconomic status – as I am legally prohibited from having it due to the manner in which TDOE interprets guidance from the federal Department of Agriculture.

17 For all SDC courses, grade levels 11 and 12 (and occasionally 10) are the grade levels suggested by TDOE as

appropriate times for students to take SDC courses. There are no 9th graders in the analytic sample because the earliest a 9th grade dual credit participant would be expected to graduate (if they took an SDC course in the first year, 2013-14) would be in the 2016-17 academic year.

sample is also higher achieving, on average, according to Algebra I standardized scores. The mean score for all Tennessee high schools is 0.260 (sd=0.992), which is less than half the value for the full analytic sample. Dual credit participants are not necessarily representative of all Tennessee high school students.

Moreover, the pooled sample masks variation between course subject areas. Math exam takers are more female (55.9 percent) and comprised of more non-White (15.8 percent) students than the full sample. Meanwhile, CTE exam takers are disproportionately more White (92.9 percent) and male (58.7 percent) than the Math examinees. Compared to Math, CTE courses also contain a relatively higher proportion of 10th graders (5.9 percent vs. 1.3 percent) and 11th graders (30.0

percent vs. 24.0 percent). Finally, and perhaps unsurprisingly, students who take SDC Math exams have the higher average Algebra I EOC performance, with students performing 0.753 standard deviations about the statewide mean (compared to 0.199 standard deviations above the mean for CTE).18 The second panel in Table 3 provides SDC course and exam information, illustrating further

variation in student experiences. CTE students more likely than Math students to take (18.2 percent vs. 11.2 percent) and pass (6.2 percent vs. 3.8 percent) other SDC exams.

The third panel in Table 3 shows unconditional mean college enrollment figures within one year of a student’s expected high school graduation (spring) semester. Beyond basic enrollment, the table provides means for college choice based on characteristics of the institution at which a student enrolls. These choice outcomes fall along several dimensions including out-of-state versus in-state, private19 versus pubic, and Tennessee public four-year versus two-year versus Tennessee Technical

18 To put these standardized math scores into further context, the Math mean of 0.753 is in the 79th percentile in the distribution of all Tennessee high school students with a non-missing Algebra I SD score. The CTE mean of 0.199 is in the 52nd percentile. Math students scoring 0-10 points above the passing margin have a mean score of 0.91 which is in the 84th percentile of the score distribution, while CTE students scoring 0-10 points above have a mean score of 0.46 which is in the 68th percentile of the score distribution.

College. The preferred college enrollment variable (“Any college”) utilizes information from both NSC and THEC data sources. For comparison, I also show unconditional enrollment figures that just use THEC data (“Any college, just THEC”) because, for performance measures summarized in the panel below, I can only use THEC data.20

Over 78 percent of the full sample enrolls in some form of postsecondary education within the year following high school graduation, with approximately 92 percent (=.721/.784) of enrollees attending an in-state institution.21 Within the group of in-state enrollees, 89 percent enroll in public

institutions, of which 54 percent enroll in public four-year colleges. Like demographic variables, postsecondary enrollment outcome means exhibit wide variation based on the SDC course subgroup of interest. CTE students are less likely than Math students to enroll in any college within a year of graduation (64.7 percent vs. 85.6 percent), and of those students who do enroll in college, CTE students are more likely to enroll at in-state, public two-year institutions (53 percent) while Math students are more likely to enroll at in-state, public four-year colleges (49 percent).

The fourth and final panel in Table 3 summarizes means and standard deviations for college performance measures including credits attempted and earned, remedial credits earned, and

cumulative GPA. Because these measures can only be gathered for students who attend a Tennessee public institution in the THEC data, these summary statistics apply to a subset of the larger sample as noted by the italicized N’s in the row right above the panel. Compared to their Math examinee peers, CTE takers both attempt (25.7 vs. 30.7) and earn (21.9 vs. 28.3) fewer credits and earn more

20 Note that the percentage of students enrolled in “Any college (just THEC)” does not align with the percentage of students enrolled in an “In-State, public” which draws on information from both NSC and THEC data sources. This is because there 409 students who are flagged as first (or mostly) enrolling in an in-state, public institution in the NSC data but not in the THEC data. This discrepancy may arise because enrollment dates do not align in the two data sources for these students.

21 A very small percentage of students (3.0 percent, or 183 out of 6,027) appear to enroll in more than one institution type, which is why unconditional means along certain dimensions (out-of-state versus in-state, private versus public, etc.) do not add up perfectly in the First Year Postsecondary Enrollment panel.

remedial credit hours (4.93 vs. 0.89). Their first-year GPAs also appear to be slightly lower, though a higher proportion of CTE students who attend a THEC institution are likely to be missing a

recorded first-year GPA (9.0 percent compared to 2.6 percent for Math). I also track these same college enrollment and early performance measures within two years of expected high school graduation; though for brevity, these sample means are not included in the table.

The wide variation in demographic, college-going, and early performance measures between examinees in different courses underscores the importance of considering the effect of passing a challenge exam separately by broad subject area – Math and CTE – and where possible, by individual course. As a final illustration of the variation in these subject areas, Figure 1 shows the number of exam takers and the exam pass rate by Tennessee county for Math and CTE, with the darker shades representing higher values. The top two panels show exam participation counts. Although there is high overlap in the counties with schools that offer both Math and CTE SDC courses, a higher proportion of Math examinees are likely to be in located in more urban, densely- populated counties, while CTE examinees are scattered throughout more rural counties. The bottom two panels show the pass rates by county, revealing similar geographic variation in where Math and CTE students who successfully bank college credit are located. These geographic differences are likely to relate to a student’s likelihood of enrolling in college and the type of institution in which they choose to enroll. Moreover, these enrollment decisions are likely to spill over into a student’s early college performance as well.

Documento similar