WEBVTT
1
00:00:01.120 --> 00:00:02.399
Welcome back everyone.
2
00:00:04.400 --> 00:00:09.039
Today, we're going to look at academic preparation and college dropout, evidence,
3
00:00:10.160 --> 00:00:12.599
counter arguments, and policy implications.
4
00:00:12.880 --> 00:00:14.160
So let's take a look and see.
5
00:00:13.960 --> 00:00:16.480
What's going on in the world of higher education policy
6
00:00:16.519 --> 00:00:20.879
and student success and inadequate academic preparation and poor early
7
00:00:20.960 --> 00:00:24.079
college performance are the predominant? Are they the predominant predictors
8
00:00:24.079 --> 00:00:25.000
of college dropout?
9
00:00:25.679 --> 00:00:25.879
And you?
10
00:00:26.039 --> 00:00:28.839
Studies suggest that learning about ones of academic ability and
11
00:00:28.839 --> 00:00:32.079
through grades plays a decisive role, estimating the dropout between
12
00:00:32.119 --> 00:00:35.439
the first and second years would decline by forty percent
13
00:00:35.560 --> 00:00:39.000
if student received no information about their great performance or
14
00:00:39.039 --> 00:00:43.159
academic ability. First thing we're going to do is outline
15
00:00:43.399 --> 00:00:46.399
the case supporting the perspective, drawing on the research and
16
00:00:46.439 --> 00:00:49.119
the data. We will then present counter arguments, also grounded
17
00:00:49.159 --> 00:00:53.880
in data and alternative predictors such as mental health, financial stress,
18
00:00:53.920 --> 00:00:54.840
social economics.
19
00:00:55.399 --> 00:00:56.119
And then at the end we.
20
00:00:56.119 --> 00:01:00.200
Will evaluate those studies and their limitations, and then we'll
21
00:01:00.240 --> 00:01:02.600
give you, hopefully a balanced conclusions. And at the end
22
00:01:03.359 --> 00:01:07.719
our analysis relies on peer reviewed research. Proponents of the
23
00:01:07.920 --> 00:01:11.359
view that academic preparation and early performance are strong predictors
24
00:01:11.400 --> 00:01:17.519
cite consistent patterns across multiple data sets Stein Brickner, published
25
00:01:17.519 --> 00:01:19.280
in the Journal of Labor Economics and based on a
26
00:01:19.359 --> 00:01:23.239
unique panel of study of low income students of Berea College.
27
00:01:24.280 --> 00:01:28.560
It's a private liberal arts institution providing full tuition subsidies,
28
00:01:28.680 --> 00:01:33.480
provides direct evidence using self reported expectations data collected before
29
00:01:33.480 --> 00:01:36.680
and after enrollment. The authors demonstrate that students enter college
30
00:01:36.719 --> 00:01:42.239
with overly optimistic beliefs about their academic ability, substantially discounting.
31
00:01:41.799 --> 00:01:43.120
The likelihood of poor grades.
32
00:01:43.519 --> 00:01:46.480
As they receive actual performance feedback, they update their beliefs
33
00:01:46.480 --> 00:01:50.599
in a manner broadly consistent what you would call Bayesian learning. Right,
34
00:01:51.040 --> 00:01:53.439
you learn something, and you change your belief for your theory.
35
00:01:54.760 --> 00:01:58.640
Simulations reveal that this learning process accounts for a substantial
36
00:01:58.680 --> 00:02:01.480
portion of the first to second attrition dropout rates, with
37
00:02:01.599 --> 00:02:05.079
fall forty percent of the absence of having that information.
38
00:02:05.840 --> 00:02:08.080
This finding aligns with broader evidence.
39
00:02:08.080 --> 00:02:12.080
Systematic reviews confirmed that high school GPA and early college
40
00:02:12.080 --> 00:02:16.199
GPR among the most reliable predictors of persistence. For instance,
41
00:02:16.240 --> 00:02:20.400
analysis of large scale data indicate that students with stronger
42
00:02:20.439 --> 00:02:25.120
academic preparation and secondary school exhibit markedly higher completion rates,
43
00:02:25.159 --> 00:02:29.919
even after controlling for other variables. High school GPA outperformed
44
00:02:29.960 --> 00:02:34.000
standardized test scores such as the sat or ACT and
45
00:02:34.080 --> 00:02:40.000
forecasting graduation, underscoring the role of sustained academic readiness rather
46
00:02:40.080 --> 00:02:44.639
than a one time aptitude measure. Additional support emerges from
47
00:02:44.680 --> 00:02:49.159
predictive modeling studies employing machine learning on institutional data for
48
00:02:49.280 --> 00:02:52.960
early semester course performance and credit accumulation consistently rank as
49
00:02:52.960 --> 00:02:57.000
top features were identifying at risk students. These models achieved
50
00:02:57.039 --> 00:03:01.800
classic classification accuracies exceeding eighty five percent when academic metrics
51
00:03:01.840 --> 00:03:08.159
are prioritized, reenforcing the causal pathway from inadequate preparation manifested
52
00:03:08.240 --> 00:03:13.400
as mismatched expectations or skill gaps to withdraw. In essence,
53
00:03:13.439 --> 00:03:16.159
the evidence suggests that when students discover through grades that
54
00:03:16.199 --> 00:03:20.240
their academic ability is lower than anticipated, the resulting reassessment
55
00:03:20.240 --> 00:03:22.199
prompts a rational decision to.
56
00:03:22.240 --> 00:03:24.960
Exit, particularly in the first critical year.
57
00:03:26.000 --> 00:03:29.520
Nevertheless, a growing body of research challenges the primacy of
58
00:03:29.639 --> 00:03:33.479
academic factors, highlighting and said the influence of non academic variables.
59
00:03:34.120 --> 00:03:37.800
Mental health emerges as a particularly salient predictor. In a
60
00:03:37.919 --> 00:03:41.639
Danish court, study of students across educational levels found that
61
00:03:41.680 --> 00:03:45.360
poor mental health was associated with significantly elevated dropout risk
62
00:03:45.439 --> 00:03:49.319
and higher education, with males exhibiting a fivefold increase in odds.
63
00:03:50.120 --> 00:03:54.000
This association, that's all, it is, not cause, persisted after
64
00:03:54.000 --> 00:04:01.319
adjusting for social, economic confounders and academic indicators. Surveys reveal
65
00:04:01.360 --> 00:04:05.680
that emotional stress rank as the leading reasons students consider
66
00:04:05.759 --> 00:04:06.439
leaving college.
67
00:04:07.319 --> 00:04:09.680
A one recent Gallup analysis, fifty.
68
00:04:09.400 --> 00:04:11.719
Four percent of respond in sight of emotional stress, and
69
00:04:11.759 --> 00:04:14.800
forty three cited personal mental health issues as primary factors,
70
00:04:15.479 --> 00:04:17.199
outpacing academic concerns.
71
00:04:17.839 --> 00:04:19.399
Financial stress also adds on.
72
00:04:19.480 --> 00:04:22.199
Fifty nine percent of students reported considering dropping out due
73
00:04:22.199 --> 00:04:25.920
to funding uncertainty, with sixty one percent noting negative impacts
74
00:04:25.920 --> 00:04:31.199
on academic performance and eighty on mental health. Social economics
75
00:04:31.199 --> 00:04:36.360
asks further complicates the narrative. Reviews of university dropout indicate
76
00:04:36.399 --> 00:04:41.240
that economically disadvantaged students phase fifteen points higher attrition risk
77
00:04:41.839 --> 00:04:46.040
even when academic preparation is comparable. So you have psychosocial
78
00:04:46.079 --> 00:04:51.199
factors like motivation, social integration, family responsibilities frequently interact with
79
00:04:51.279 --> 00:04:55.759
and in some cases overshadow purely academic metrics. Give me
80
00:04:55.800 --> 00:04:59.480
an example, students experiencing financial hardship or caregiving demands may
81
00:04:59.560 --> 00:05:02.399
underperf mom academically as a consequence rather than as a
82
00:05:02.399 --> 00:05:06.199
cause of dropout risk. The findings imply that while poor
83
00:05:06.240 --> 00:05:09.079
early performance correlates with attrition, it may serve as a
84
00:05:09.079 --> 00:05:11.279
proxy for underlying stressors rather.
85
00:05:11.120 --> 00:05:12.800
Than an independent driver.
86
00:05:15.160 --> 00:05:19.519
Interventions targeting only academic remediation without addressing mental health support
87
00:05:19.600 --> 00:05:25.839
or financial aid may therefore yield incomplete results. We also
88
00:05:25.879 --> 00:05:28.600
have to look at the methodological constraints of these claims.
89
00:05:29.800 --> 00:05:32.680
The Steinbrckner study we looked at earlier, while innovative in
90
00:05:32.720 --> 00:05:36.199
its use of real time data, is limited to a
91
00:05:36.240 --> 00:05:41.079
specific population low income students had a single tuition subsidized institution.
92
00:05:42.120 --> 00:05:47.120
Generalizability remains uncertain. Many predictive models rely on correlational designs,
93
00:05:47.720 --> 00:05:53.519
rendering causal inference very difficult. Unmeasured variables such as institutional
94
00:05:53.519 --> 00:05:56.800
support quality or external life events may confound observe relations
95
00:05:57.000 --> 00:06:01.480
between grades and dropout. Reported measures of expectations or mental
96
00:06:01.519 --> 00:06:06.199
health introduced potential response bias, and longitudinal studies often suffer
97
00:06:06.199 --> 00:06:11.480
from attrition themselves, potentially underestimating effects among the most vulnerable.
98
00:06:14.480 --> 00:06:18.680
So the other factors, rapid, technological and post pandemic shifts
99
00:06:18.680 --> 00:06:22.720
and higher education may render the pre twenty twenty data
100
00:06:23.000 --> 00:06:26.839
less predictive than today. These limitations underscore the need for cushion.
101
00:06:27.879 --> 00:06:32.240
In conclusion, the evidence substantiates that inadequate academic preparation and
102
00:06:32.439 --> 00:06:37.920
poor early college performance constitute strong empirically supportive predictors of dropout,
103
00:06:38.720 --> 00:06:42.920
particularly in the transition from its first to second year.
104
00:06:44.600 --> 00:06:47.959
Countervailing studies on mental health, financial stress, and social economic
105
00:06:48.000 --> 00:06:50.319
pressures demonstrate that these academic factors are.
106
00:06:50.240 --> 00:06:51.399
Not operating in isolation.
107
00:06:51.480 --> 00:06:56.879
They frequently interact and are exacerbated by non cognitive external influences.
108
00:06:57.439 --> 00:06:58.480
So what does that leave us?
109
00:06:58.920 --> 00:07:02.199
It looks like a policy REPS therefore requires, like usual,
110
00:07:02.360 --> 00:07:03.959
multifaceted interventions.
111
00:07:03.959 --> 00:07:05.040
It's not just one thing.
112
00:07:05.759 --> 00:07:09.240
The question now really is been timing to become is
113
00:07:09.240 --> 00:07:13.040
the pre college academic preparation. How much does it contribute
114
00:07:13.040 --> 00:07:17.000
to the dropout rate relative to mental health, financial issues,
115
00:07:17.079 --> 00:07:21.519
or socio economic factors. In other words, how big of
116
00:07:21.560 --> 00:07:23.480
the slices of the pie does each one of these have?
117
00:07:24.439 --> 00:07:27.360
And then now also the determines which interventions we should
118
00:07:27.360 --> 00:07:28.959
start off with first or focus more on.
119
00:07:30.040 --> 00:07:30.879
That's it for now.
120
00:07:32.399 --> 00:07:35.639
We encourage listeners to consult the reference studies for deeper analysis.
121
00:07:37.000 --> 00:07:39.319
And hope you enjoyed the deep dive
1
00:00:01.120 --> 00:00:02.399
Welcome back everyone.
2
00:00:04.400 --> 00:00:09.039
Today, we're going to look at academic preparation and college dropout, evidence,
3
00:00:10.160 --> 00:00:12.599
counter arguments, and policy implications.
4
00:00:12.880 --> 00:00:14.160
So let's take a look and see.
5
00:00:13.960 --> 00:00:16.480
What's going on in the world of higher education policy
6
00:00:16.519 --> 00:00:20.879
and student success and inadequate academic preparation and poor early
7
00:00:20.960 --> 00:00:24.079
college performance are the predominant? Are they the predominant predictors
8
00:00:24.079 --> 00:00:25.000
of college dropout?
9
00:00:25.679 --> 00:00:25.879
And you?
10
00:00:26.039 --> 00:00:28.839
Studies suggest that learning about ones of academic ability and
11
00:00:28.839 --> 00:00:32.079
through grades plays a decisive role, estimating the dropout between
12
00:00:32.119 --> 00:00:35.439
the first and second years would decline by forty percent
13
00:00:35.560 --> 00:00:39.000
if student received no information about their great performance or
14
00:00:39.039 --> 00:00:43.159
academic ability. First thing we're going to do is outline
15
00:00:43.399 --> 00:00:46.399
the case supporting the perspective, drawing on the research and
16
00:00:46.439 --> 00:00:49.119
the data. We will then present counter arguments, also grounded
17
00:00:49.159 --> 00:00:53.880
in data and alternative predictors such as mental health, financial stress,
18
00:00:53.920 --> 00:00:54.840
social economics.
19
00:00:55.399 --> 00:00:56.119
And then at the end we.
20
00:00:56.119 --> 00:01:00.200
Will evaluate those studies and their limitations, and then we'll
21
00:01:00.240 --> 00:01:02.600
give you, hopefully a balanced conclusions. And at the end
22
00:01:03.359 --> 00:01:07.719
our analysis relies on peer reviewed research. Proponents of the
23
00:01:07.920 --> 00:01:11.359
view that academic preparation and early performance are strong predictors
24
00:01:11.400 --> 00:01:17.519
cite consistent patterns across multiple data sets Stein Brickner, published
25
00:01:17.519 --> 00:01:19.280
in the Journal of Labor Economics and based on a
26
00:01:19.359 --> 00:01:23.239
unique panel of study of low income students of Berea College.
27
00:01:24.280 --> 00:01:28.560
It's a private liberal arts institution providing full tuition subsidies,
28
00:01:28.680 --> 00:01:33.480
provides direct evidence using self reported expectations data collected before
29
00:01:33.480 --> 00:01:36.680
and after enrollment. The authors demonstrate that students enter college
30
00:01:36.719 --> 00:01:42.239
with overly optimistic beliefs about their academic ability, substantially discounting.
31
00:01:41.799 --> 00:01:43.120
The likelihood of poor grades.
32
00:01:43.519 --> 00:01:46.480
As they receive actual performance feedback, they update their beliefs
33
00:01:46.480 --> 00:01:50.599
in a manner broadly consistent what you would call Bayesian learning. Right,
34
00:01:51.040 --> 00:01:53.439
you learn something, and you change your belief for your theory.
35
00:01:54.760 --> 00:01:58.640
Simulations reveal that this learning process accounts for a substantial
36
00:01:58.680 --> 00:02:01.480
portion of the first to second attrition dropout rates, with
37
00:02:01.599 --> 00:02:05.079
fall forty percent of the absence of having that information.
38
00:02:05.840 --> 00:02:08.080
This finding aligns with broader evidence.
39
00:02:08.080 --> 00:02:12.080
Systematic reviews confirmed that high school GPA and early college
40
00:02:12.080 --> 00:02:16.199
GPR among the most reliable predictors of persistence. For instance,
41
00:02:16.240 --> 00:02:20.400
analysis of large scale data indicate that students with stronger
42
00:02:20.439 --> 00:02:25.120
academic preparation and secondary school exhibit markedly higher completion rates,
43
00:02:25.159 --> 00:02:29.919
even after controlling for other variables. High school GPA outperformed
44
00:02:29.960 --> 00:02:34.000
standardized test scores such as the sat or ACT and
45
00:02:34.080 --> 00:02:40.000
forecasting graduation, underscoring the role of sustained academic readiness rather
46
00:02:40.080 --> 00:02:44.639
than a one time aptitude measure. Additional support emerges from
47
00:02:44.680 --> 00:02:49.159
predictive modeling studies employing machine learning on institutional data for
48
00:02:49.280 --> 00:02:52.960
early semester course performance and credit accumulation consistently rank as
49
00:02:52.960 --> 00:02:57.000
top features were identifying at risk students. These models achieved
50
00:02:57.039 --> 00:03:01.800
classic classification accuracies exceeding eighty five percent when academic metrics
51
00:03:01.840 --> 00:03:08.159
are prioritized, reenforcing the causal pathway from inadequate preparation manifested
52
00:03:08.240 --> 00:03:13.400
as mismatched expectations or skill gaps to withdraw. In essence,
53
00:03:13.439 --> 00:03:16.159
the evidence suggests that when students discover through grades that
54
00:03:16.199 --> 00:03:20.240
their academic ability is lower than anticipated, the resulting reassessment
55
00:03:20.240 --> 00:03:22.199
prompts a rational decision to.
56
00:03:22.240 --> 00:03:24.960
Exit, particularly in the first critical year.
57
00:03:26.000 --> 00:03:29.520
Nevertheless, a growing body of research challenges the primacy of
58
00:03:29.639 --> 00:03:33.479
academic factors, highlighting and said the influence of non academic variables.
59
00:03:34.120 --> 00:03:37.800
Mental health emerges as a particularly salient predictor. In a
60
00:03:37.919 --> 00:03:41.639
Danish court, study of students across educational levels found that
61
00:03:41.680 --> 00:03:45.360
poor mental health was associated with significantly elevated dropout risk
62
00:03:45.439 --> 00:03:49.319
and higher education, with males exhibiting a fivefold increase in odds.
63
00:03:50.120 --> 00:03:54.000
This association, that's all, it is, not cause, persisted after
64
00:03:54.000 --> 00:04:01.319
adjusting for social, economic confounders and academic indicators. Surveys reveal
65
00:04:01.360 --> 00:04:05.680
that emotional stress rank as the leading reasons students consider
66
00:04:05.759 --> 00:04:06.439
leaving college.
67
00:04:07.319 --> 00:04:09.680
A one recent Gallup analysis, fifty.
68
00:04:09.400 --> 00:04:11.719
Four percent of respond in sight of emotional stress, and
69
00:04:11.759 --> 00:04:14.800
forty three cited personal mental health issues as primary factors,
70
00:04:15.479 --> 00:04:17.199
outpacing academic concerns.
71
00:04:17.839 --> 00:04:19.399
Financial stress also adds on.
72
00:04:19.480 --> 00:04:22.199
Fifty nine percent of students reported considering dropping out due
73
00:04:22.199 --> 00:04:25.920
to funding uncertainty, with sixty one percent noting negative impacts
74
00:04:25.920 --> 00:04:31.199
on academic performance and eighty on mental health. Social economics
75
00:04:31.199 --> 00:04:36.360
asks further complicates the narrative. Reviews of university dropout indicate
76
00:04:36.399 --> 00:04:41.240
that economically disadvantaged students phase fifteen points higher attrition risk
77
00:04:41.839 --> 00:04:46.040
even when academic preparation is comparable. So you have psychosocial
78
00:04:46.079 --> 00:04:51.199
factors like motivation, social integration, family responsibilities frequently interact with
79
00:04:51.279 --> 00:04:55.759
and in some cases overshadow purely academic metrics. Give me
80
00:04:55.800 --> 00:04:59.480
an example, students experiencing financial hardship or caregiving demands may
81
00:04:59.560 --> 00:05:02.399
underperf mom academically as a consequence rather than as a
82
00:05:02.399 --> 00:05:06.199
cause of dropout risk. The findings imply that while poor
83
00:05:06.240 --> 00:05:09.079
early performance correlates with attrition, it may serve as a
84
00:05:09.079 --> 00:05:11.279
proxy for underlying stressors rather.
85
00:05:11.120 --> 00:05:12.800
Than an independent driver.
86
00:05:15.160 --> 00:05:19.519
Interventions targeting only academic remediation without addressing mental health support
87
00:05:19.600 --> 00:05:25.839
or financial aid may therefore yield incomplete results. We also
88
00:05:25.879 --> 00:05:28.600
have to look at the methodological constraints of these claims.
89
00:05:29.800 --> 00:05:32.680
The Steinbrckner study we looked at earlier, while innovative in
90
00:05:32.720 --> 00:05:36.199
its use of real time data, is limited to a
91
00:05:36.240 --> 00:05:41.079
specific population low income students had a single tuition subsidized institution.
92
00:05:42.120 --> 00:05:47.120
Generalizability remains uncertain. Many predictive models rely on correlational designs,
93
00:05:47.720 --> 00:05:53.519
rendering causal inference very difficult. Unmeasured variables such as institutional
94
00:05:53.519 --> 00:05:56.800
support quality or external life events may confound observe relations
95
00:05:57.000 --> 00:06:01.480
between grades and dropout. Reported measures of expectations or mental
96
00:06:01.519 --> 00:06:06.199
health introduced potential response bias, and longitudinal studies often suffer
97
00:06:06.199 --> 00:06:11.480
from attrition themselves, potentially underestimating effects among the most vulnerable.
98
00:06:14.480 --> 00:06:18.680
So the other factors, rapid, technological and post pandemic shifts
99
00:06:18.680 --> 00:06:22.720
and higher education may render the pre twenty twenty data
100
00:06:23.000 --> 00:06:26.839
less predictive than today. These limitations underscore the need for cushion.
101
00:06:27.879 --> 00:06:32.240
In conclusion, the evidence substantiates that inadequate academic preparation and
102
00:06:32.439 --> 00:06:37.920
poor early college performance constitute strong empirically supportive predictors of dropout,
103
00:06:38.720 --> 00:06:42.920
particularly in the transition from its first to second year.
104
00:06:44.600 --> 00:06:47.959
Countervailing studies on mental health, financial stress, and social economic
105
00:06:48.000 --> 00:06:50.319
pressures demonstrate that these academic factors are.
106
00:06:50.240 --> 00:06:51.399
Not operating in isolation.
107
00:06:51.480 --> 00:06:56.879
They frequently interact and are exacerbated by non cognitive external influences.
108
00:06:57.439 --> 00:06:58.480
So what does that leave us?
109
00:06:58.920 --> 00:07:02.199
It looks like a policy REPS therefore requires, like usual,
110
00:07:02.360 --> 00:07:03.959
multifaceted interventions.
111
00:07:03.959 --> 00:07:05.040
It's not just one thing.
112
00:07:05.759 --> 00:07:09.240
The question now really is been timing to become is
113
00:07:09.240 --> 00:07:13.040
the pre college academic preparation. How much does it contribute
114
00:07:13.040 --> 00:07:17.000
to the dropout rate relative to mental health, financial issues,
115
00:07:17.079 --> 00:07:21.519
or socio economic factors. In other words, how big of
116
00:07:21.560 --> 00:07:23.480
the slices of the pie does each one of these have?
117
00:07:24.439 --> 00:07:27.360
And then now also the determines which interventions we should
118
00:07:27.360 --> 00:07:28.959
start off with first or focus more on.
119
00:07:30.040 --> 00:07:30.879
That's it for now.
120
00:07:32.399 --> 00:07:35.639
We encourage listeners to consult the reference studies for deeper analysis.
121
00:07:37.000 --> 00:07:39.319
And hope you enjoyed the deep dive
