ERIC EJ1015237: Transitions from High School to College

2013·0.15 MB·English

by  ERIC

Transitions from High School to College
Transitions from High School to College
Andrea Venezia and Laura Jaeger
Summary
The vast majority of high school students aspire to some kind of postsecondary education, yet
far too many of them enter college without the basic content knowledge, skills, or habits of mind
they need to succeed. Andrea Venezia and Laura Jaeger look at the state of college readiness
among high school students, the effectiveness of programs in place to help them transition to
college, and efforts to improve those transitions.
Students are unprepared for postsecondary coursework for many reasons, the authors write,
including differences between what high schools teach and what colleges expect, as well as large
disparities between the instruction offered by high schools with high concentrations of students
in poverty and that offered by high schools with more advantaged students. The authors also
note the importance of noncurricular variables, such as peer influences, parental expectations,
and conditions that encourage academic study.
Interventions to improve college readiness offer a variety of services, from academic prepara-
tion and information about college and financial aid, to psychosocial and behavioral supports,
to the development of habits of mind including organizational skills, anticipation, persistence,
and resiliency. The authors also discuss more systemic programs, such as Middle College High
Schools, and review efforts to allow high school students to take college classes (known as dual
enrollment). Evaluations of the effectiveness of these efforts are limited, but the authors report
that studies of precollege support programs generally show small impacts, while the more sys-
temic programs show mixed results. Dual-enrollment programs show promise, but the evalua-
tion designs may overstate the results.
The Common Core State Standards, a voluntary set of goals and expectations in English and
math adopted by most states, offer the potential to improve college and career readiness, the
authors write. But that potential will be realized, they add, only if the standards are supple-
mented with the necessary professional development to enable educators to help all students
meet academic college readiness standards, a focus on developing strong noncognitive knowl-
edge and skills for all students, and the information and supports to help students prepare and
select the most appropriate postsecondary institution.
www.futureofchildren.org
Andrea Venezia is a senior research associate for WestEd’s Regional Educational Laboratory West in San Francisco. Laura Jaeger is a
research associate at WestEd in San Francisco.
VOL. 23 / NO. 1 / SPRING 2013 117
Andrea Venezia and Laura Jaeger
As Sandy Baum, Charles Kurose, Early Assessment Program, and statewide
and Michael McPherson default curricula. Finally, it describes the
discuss in their article in this Common Core State Standards movement
issue, the postsecondary educa- and concludes with a discussion of both the
tion landscape in the United need for more comprehensive and systemic
States has changed dramatically over the past reforms and the challenges related to imple-
half-century.1 The aspirations and actions of menting them.
the vast majority of high school students have
shifted, with greater percentages of students Understanding the Problem
intending to complete some form of postsec- In recent years, roughly 3 million students
ondary education. For example, from 1980 to have been graduating from U.S high schools
2002, the share of tenth graders who aspired annually. According to the National Center
to earn at least a bachelor’s degree rose from for Educational Statistics, more than
41 percent to 80 percent, with the largest 2.9 million students graduated from U.S.
increase coming from low-income students.2 high schools in 2008, the last year for which
Unfortunately, far too many students enter data are available.3 A key question is, how
college without the basic content knowledge, many of these students are prepared for
skills, or habits of mind needed to perform college-level work?
college-level work successfully. As college-
going rates increase, the limitations of the College readiness is commonly understood
traditional and current structures, programs, as the level of preparation a student needs to
and practices designed to promote student enroll and succeed in a college program (cer-
success within both secondary and postsec- tificate, associate’s degree, or baccalaureate)
ondary education systems and institutions without requiring remediation.4 While there
become more visible. is no precise way of knowing how many high
school graduates meet this standard, the
This chapter discusses transitions from high largest nationally representative and continu-
school to college and some of the major ing assessment of what America’s students
efforts under way in states and schools to know and can do in various subject areas—
improve college preparation. It begins with the National Assessment of Educational
an overview of the problem, including esti- Progress (NAEP)—suggests that many
mates of the number of high school graduates students are likely falling short. The NAEP
who are not ready for college and the major determines students’ achievement level—
reasons why they are not. The chapter then basic, proficient, or advanced—based on
explores whether current conceptions of input from a broadly representative panel of
college readiness are adequate and also what teachers, education specialists, and members
it means for students to find the right col- of the general public. Students determined to
lege “fit.” Next, it reviews some of the major be proficient or advanced have demonstrated
interventions designed to improve college a competency over challenging subject
readiness, particularly among low-income matter that would be expected of entering
students: the federal TRIO programs, the college students, including subject-matter
Early College High School (ECHS) and knowledge, application of such knowledge
Middle College High School (MCHS) initia- to real-world situations, and analytical skills
tives, dual-enrollment programs, California’s appropriate to the subject matter. In 2009,
118 THE FUTURE OF CHILDREN
Transitions from High School to College
only 38 percent of twelfth-grade students rote memorization of algorithms, rather than
performed at or above the proficient level engaging students in problem-solving and
on NAEP’s reading assessment; even fewer, critical-writing exercises that develop both
26 percent, were at or above the proficient deeper knowledge of the content and the
level in mathematics.5 more general logical and analytical think-
ing skills valued at the postsecondary level.9
Other common assessments used to deter- Most public high schools offer at least one
mine college readiness are the ACT and Advance Placement (AP) or, less commonly,
SAT exams, which are typically adminis- one International Baccalaureate (IB) course.
tered to high school juniors and seniors. In These courses are designed to be more rigor-
2012, only 25 percent of all ACT-tested high ous than a standard high school course and
school graduates met the College Readiness to foster the critical thinking skills expected
Benchmarks in all four subjects, meaning of college students. That said, the College
that they earned the minimum score needed Board, which administers the AP program,
to have a 50 percent chance of obtaining a reports that only 30 percent of 2011 public
“B” or higher in corresponding first-year high school graduates participated in AP
college courses. Fifty-two percent of gradu- courses and only 18.1 percent succeeded in
ates met the ACT’s reading benchmark, scoring 3 or higher (“qualified” to receive
46 percent met the mathematics benchmark, college credit or placement into advanced
and 67 percent met the English benchmark. courses) on at least one AP exam.10
Only 31 percent met the benchmark in sci-
ence.6 Looking at SAT data, among the high The decentralized nature of education in
school graduating class of 2012, only 43 per- the United States—in which states delegate
cent of all SAT takers met the SAT College authority to more than 15,000 local school
& Career Readiness Benchmark, which indi- districts to design and direct programs of
cates a 65 percent likelihood of obtaining a instruction—may partly explain the variation
“B-” average or higher during the first year in what high schools offer and how well they
of college.7 prepare students for college.11 In the 2010–11
academic year, more than 49 million stu-
The reasons why more high school graduates dents were enrolled in public elementary and
are not ready for college are complex and secondary schools.12 The key characteristics
highly dependent upon individual circum- of those schools show disparities by race and
stances. The factors are academic and non- ethnicity and by poverty level. For example,
academic; schools are able to control some of 60 percent of Asian/Pacific Islander and
them but not others, such as family variables just over half of white high school freshmen
and peer influences outside of school. On the attended schools in which the counselors
academic side, many studies over the past reported that the primary goal of the school
ten years have documented the disconnect guidance program was to help students pre-
between what high school teachers teach and pare for college. In contrast, only 44 percent
what postsecondary instructors expect with of black freshmen, 41 percent of Hispanic
regard to students’ preparation for first-year freshmen, and 29 percent of American
credit-bearing courses in college.8 High Indian/Alaskan Native freshmen attended
school courses, such as algebra, often teach such schools.13 White and Asian students are
content such as factoring equations by using more likely to attend low-poverty schools,
VOL. 23 / NO. 1 / SPRING 2013 119
Andrea Venezia and Laura Jaeger
while American Indian/Alaskan Native, Kallick coined the term “habits of mind” to
black, and Hispanic students are more likely describe a series of intelligent behaviors that
to attend high-poverty schools. In 2007–08, would help people be better problem solvers
approximately 91 percent of twelfth-graders and thus have more success in their lives.19
in low-poverty schools graduated with a David Conley refines the concept to describe
diploma, compared with 68 percent of the habits of mind necessary to succeed in
twelfth-graders in high-poverty schools college including critical thinking, an inquisi-
(based on eligibility for free or reduced- tive nature, a willingness to accept critical
price lunch).14 In that same year, 52 percent feedback, an openness to possible failure,
of high school graduates from low-poverty and the ability to cope with frustrating and
schools attended a four-year postsecondary ambiguous learning tasks.20
institution, compared with about 28 percent
of graduates from high-poverty schools.15 Are Current Measures of College
Unfortunately, current disparities could Readiness Adequate?
grow, given recent budget cuts to all levels With larger proportions of underserved stu-
of education—primary, secondary, and dent populations going to college, traditional
postsecondary—that are likely to affect low- indicators of academic preparation such as
income students the most.16 the SAT and ACT have come under fire.
Critics are concerned that wealthier students
As noted, nonacademic factors also affect have better opportunities to prepare for such
college readiness. Students’ families play tests, that the tests do not measure what is
an important role in setting expectations learned in the classroom, and that the tests
and creating conditions—from overseeing are not strong predictors of how students
completion of homework assignments to perform in college.21 In addition, the large
encouraging a variety of learning opportuni- numbers of students who plan to attend com-
ties outside of school—that make it more or munity college generally do not take the SAT
less likely that students will be prepared for or ACT because these tests are not required
college. Not surprisingly, research shows that for admission. Community colleges do use
students whose parents have gone to college standardized tests after matriculation, such
are more likely to attend college themselves.17 as the ACCUPLACER and COMPASS, to
Students are also influenced positively or determine if students need to take remedial
negatively by the people they encounter education in English language arts and math-
at school and in their community. Patricia ematics and then to place students in the
Gándara and Deborah Bial, for example, appropriate courses. As noted in the article in
state that many students face impediments this issue by Eric Bettinger, Angela Boatman,
such as limited cultural supports, commu- and Bridget Terry Long, these tests also have
nity resources, and peer supports, as well as been found to be poor predictors of how
racism, ineffective counseling, and limited students will perform academically.22
networking opportunities with people who
have succeeded in college.18 Finally, college Frustrations with the limitations of standard-
readiness can be influenced by noncogni- ized tests, together with new thinking and
tive skills that differ at the individual level research on what it means to be prepared
and may be related to both schooling and for college or a job right out of high school
family background. Arthur Costa and Bena (commonly referred to as “college and career
120 THE FUTURE OF CHILDREN
Transitions from High School to College
readiness”), have led to efforts to develop clear whether a single framework can sup-
new and more comprehensive measures. The port opportunities for students to be ready to
Educational Policy Improvement Center succeed at all postsecondary institutions and
(EPIC), Georgetown’s Center on Education within all workforce opportunities. Finally,
and the Workforce, the Association although these broadened definitions of
for Career and Technical Education, college readiness are intriguing, it is unclear
ConnectEd: the California Center for whether and how these notions may be incor-
College and Career, the Conference Board, porated into state educational policies or the
the National Association of State Directors assessment practices of typical high schools
of Career Technical Education Consortium, or school districts.
the Secretary’s Commission on Achieving
Necessary Skills, and Assessing and Teaching Finding the Right College Fit
21st Century Skills are among the groups A corollary to determining college readi-
and organizations that have developed new ness is the importance of helping students
college and career readiness standards.23 to find the right institutional fit, particularly
These standards include not only the English for students from low-income families or
language arts and mathematics necessary families that do not have experience with
for entering first-year college students to college. “Fit” includes aspects of a postsec-
take college-level credit-bearing courses but ondary institution such as its cost, location,
other competencies as well. For instance, size, student-faculty ratio, counseling and
some focus on twenty-first-century expecta- advising services, student body composition
tions. While these standards vary depending (for example, institutions that primarily serve
on the organization that developed them, students from a particular racial, ethnic,
they generally focus on quantitative STEM or religious background, or single-gender
(science, technology, engineering, and institutions), and areas of study offered or
mathematics) knowledge and skills; technical special areas of focus.25 Many traditionally
content (this area applies to preparation for underserved students often do not have the
career and technical education courses and option to matriculate farther away than the
includes a range of career-specific knowledge closest community college or broad-access
and skills); broad transferable skills (such as university because they need to stay close to
productive dispositions and behaviors); habits home to contain costs or help their family. In
of mind; and preparation for civic life (such addition, all students, but particularly stu-
as knowledge of the democratic process and dents who are traditionally underrepresented
civic engagement).24 These categories are not in college, often do not know enough both
mutually exclusive, and views differ about about themselves and their future goals and
what each category comprises and how much about postsecondary institutions to analyze
weight each component in a category should institutional fit.26
carry. In addition, there is no consensus
about whether college and career readiness An issue closely related to choice and fit is the
are different and, if so, how they differ. If tendency for some students to attend colleges
they are different, the concern is how schools that are less selective than those they are
can avoid curricular tracking by ethnic- qualified to attend. This phenomenon, known
ity and income levels. Moreover, if college as “undermatching,” refers to students who
and career readiness are different, it is not meet the admissions criteria for high-ranking
VOL. 23 / NO. 1 / SPRING 2013 121
Andrea Venezia and Laura Jaeger
colleges and universities based on test scores, in college immediately after high school,
rigorous course taking, and grades but who compared with 84 percent of those from the
instead go to less selective four-year col- most affluent families and 67 percent from
leges, two-year colleges, or no college at middle-income families. These inequalities
all.27 Available research findings suggest that have helped to drive the growth of precollege
undermatching is particularly a problem for outreach programs and large-scale interven-
students of color and from low-income fami- tions and reforms.
lies. A descriptive study that used case studies
to examine how social class and high school Interventions Designed to Boost
guidance operations interact to influence high College Readiness
school students’ educational aspirations found A variety of programs are now available to
that female students, African American stu- help boost the college readiness of today’s
dents, and students from low-socioeconomic high school students. Current interventions
backgrounds are most likely to undermatch.28 and reform efforts use a range of strate-
Analyses of longitudinal data suggest that gies to attempt to address a wide variety of
students who undermatch are significantly student needs regarding college readiness.
less likely to graduate. In their study of sixty- Strategies range from academic preparation
eight public colleges and universities, includ- to psychosocial and behavioral supports and
ing twenty-one flagship institutions in four the development of appropriate habits of
states, William Bowen and colleagues found mind (such as organization, anticipation, per-
that students who attended the most selective sistence, and resiliency). While each inter-
colleges for which they were academically vention tends to focus on a distinct group of
qualified were more likely to graduate than students and to emphasize different aspects
were similar students who undermatched of college readiness, there is considerable
by enrolling in colleges for which they were overlap in the strategies these efforts use in
overqualified.29 helping students have access to, be prepared
for, and succeed in postsecondary school-
These findings, together with the well- ing. In this section, we discuss some of the
documented pattern of students from better-known programs; their strategies are
middle- and upper-income families attend- summarized in table 1.
ing four-year institutions, while low-income
students are concentrated in two-year Federal TRIO Programs
community colleges, reflect major weak- Since 1965, an estimated 2 million students
nesses in the college-choice process for have graduated from college with the spe-
many students, especially minority and cial assistance and support of federal TRIO
low-income students.30 The inequalities in programs, such as Upward Bound and Talent
college-going and success rates by ethnic- Search, which provide outreach and student
ity and income groups are stark. As of 2010, services to individuals from low-income
60.5 percent of the college population was backgrounds, those with disabilities, and
white non-Hispanic students, compared with those who are first-generation college-going to
14.5 percent black students and 13.0 percent help them successfully navigate their educa-
Hispanic students.31 Moreover, in 2009, only tional pathways from middle school through
55 percent of high school graduates from post-baccalaureate programs.32 Upward
the lowest family income quintile enrolled Bound academic preparation provides
122 THE FUTURE OF CHILDREN
Transitions from High School to College
Table 1. Strategies Used by Selected College Readiness Interventions and Reforms
Areas of student need
Better academic Increased Greater Better information Better alignment Development of
preparation psychosocial exposure about college and between appropriate habits
and behavioral to college financial aid high school and of mind
support college
Intervention reform
assessment and
strategy
curricula
TRIO
Upward Bound (cid:23) (cid:23) (cid:23) (cid:23)
Talent Search (cid:23) (cid:23) (cid:23)
GEAR UP (cid:23) (cid:23) (cid:23) (cid:23) (cid:23)
Early College High School and
Middle College (cid:23) (cid:23) (cid:23) (cid:23) (cid:23) (cid:23)
High School
Dual Enrollment (cid:23) (cid:23) (cid:23)
Early Assessment Program (cid:23) (cid:23)
Default curricula (cid:23) (cid:23)
Source: Authors.
participants with instruction in mathematics, Gaining Early Awareness and Readiness
laboratory sciences, composition, literature, for Undergraduate Programs (GEAR UP),
and foreign languages. It also offers academic a federal program established by Congress
and social support through tutoring, coun- as part of the 1998 reauthorization of the
seling, mentoring, cultural enrichment, and Higher Education Act, provides six-year
grants to states and to partnerships (among
work-study programs, and provides education
local elementary and secondary schools, insti-
to improve the financial and economic literacy
tutions of higher education, and community
of students. Talent Search provides students
organizations) to serve cohorts of students
and their parents with information about col-
attending high poverty schools beginning no
lege admissions requirements, scholarships,
later than the seventh grade and following
and financial aid. It also provides social sup-
them through high school. In contrast to pro-
port through counseling and helping students
grams such as Upward Bound that focus on
understand their educational options. Upward
academic preparation, GEAR UP programs
Bound and Talent Search both include
take a more systemic approach by provid-
services designed for disconnected student
ing college scholarships, academic support
groups, such as students who drop out of
services and counseling, and college-related
high school, students who have limited
information. They also attempt to work with
English proficiencies, students from groups
the parents and families of the students.
that are traditionally underrepresented in
postsecondary education, students with dis- Funding for these programs, however, is
abilities, homeless students, and students who inadequate to reach all the students in need
are in foster care or are aging out of the foster of them. In 2011, 951 Upward Bound pro-
care system. grams served more than 64,000 students
VOL. 23 / NO. 1 / SPRING 2013 123
Andrea Venezia and Laura Jaeger
nationwide, and more than 300,000 stu- school and college credit. Typically, high
dents in grades six through twelve across the school and college faculty work together to
nation were involved with Talent Search.33 ensure that curricula and instruction within
In 2010, Congress appropriated more than the high school align well with credit-bearing
$300 million for GEAR UP, which served college-level coursework. Some ECHSs and
748,000 students through 42 state grants MCHSs work with feeder middle schools
and 169 partnership grants.34 Despite the to begin this “scaffolding,” or alignment, of
large number of students being served by curricula and instruction in earlier grades.
these programs, not all eligible students are Another difference from most large com-
being reached. According to the Council prehensive high schools is that ECHSs and
for Opportunity in Education, 11 million MCHSs try to provide students with a full
students are eligible for and need access to range of support services, including advisory
services through TRIO programs, but federal classes, college counseling, peer support, psy-
funding is sufficient to serve less than 7 per- chosocial and behavioral supports, and career
cent of those eligible students.35 experience opportunities for all students.38
Middle College and Early College High Dual-enrollment programs also provide
Schools; Dual Enrollment opportunities for high school students to
The most comprehensive of all the efforts take college-level classes and earn both
discussed here are Middle College High high school and college credit but with-
Schools (MCHSs) and Early College High out the additional supports of the MCHC
Schools (ECHSs). These are small schools and ECHS models.39 Historically, dual-
(the average size is around 250 students) enrollment programs have been offered in
that serve students historically underrepre- highly resourced high schools with large
sented in college populations and that aim to percentages of students who matriculate
coordinate student services, decrease repeti- into college. Increasingly, however, dual
tion in curriculum, make college attainable, enrollment is being offered in high schools
and eliminate the need for remediation. The serving high-need populations. Students
first MCHS opened in 1974 at LaGuardia do not pay for the dual-enrollment courses,
Community College in New York; there are so they can accumulate free college credit
now 40 MCHSs across the United States.36 in high school and potentially shorten the
The ECHS Initiative, which builds off the time it takes to complete a degree once they
MCHS model and is supported by the Bill matriculate, therefore accelerating their
and Melinda Gates Foundation, includes progression from high school to and through
approximately 270 schools serving more than college. A critical issue is that the standards
75,000 students in 28 states.37 Both models for dual-enrollment courses must remain
attempt to create strong college-going cul- college level.
tures throughout each school and to partner
with colleges to provide dual-enrollment The U.S. Department of Education reports
opportunities, college visits, and other con- that as of 2005, 98 percent of community
nections with postsecondary education. Dual- colleges and 77 percent of public four-year
enrollment courses are college-level courses, colleges were participating in dual-enrollment
taught either in high schools or colleges, for programs.40 Most of these programs serve a
which high school students receive both high relatively small number of students at specific
124 THE FUTURE OF CHILDREN
Transitions from High School to College
school sites, however. Thirty-eight states have Another state-level reform effort receiving
policies that allow for dual enrollment, but attention is California’s Early Assessment
some states will not allow both high schools Program (EAP), a collaborative effort that
and colleges to receive funding for the same started in 2004 among the state board
course. Postsecondary faculty members teach of education, the California Department
some dual-enrollment courses, while others of Education, and the California State
are taught by high school teachers who have University system. The EAP provides an
completed training at the postsecondary insti- assessment of college readiness in English
tution that is providing the college credit. The and mathematics for one system of higher
education in California (the state universi-
kinds of courses offered through dual enroll-
ties) to help students prepare for placement
ment also vary a great deal. Some institutions
exams before they enroll in college and
provide access to any course requested by
thus avoid the need for remediation once
the participating high schools students, while
they reach college. The EAP uses students’
others limit course options based on available
scores on California’s eleventh-grade assess-
sections and other factors. A growing number
ment as indicators of students’ readiness for
of dual-enrollment courses are in career and
college-level work in the state university and
technical education.41
community college systems. Incoming high
school seniors receive notification in August
State-Level Reforms
before their senior year about their level of
More recently, various state-level reforms
readiness and the courses they can take to
have emerged that address specific areas of
improve their academic preparation. Students
college readiness through key leverage points
who score high enough on the EAP (or on the
within a state system. A growing number of
SAT or ACT) are exempt from taking postsec-
these programs focus on students’ academic
ondary placement tests and can go right into
preparation and better alignment between
college-level courses.43
high schools and colleges in the curricula
and assessment tools they use. One example
Evidence on Effectiveness
is the implementation (typically statewide) of
Although they employ a range of strate-
default curricula, which attempt to eliminate
gies, these programs all share the same aim:
tracking in which some high school students
to increase the rates at which participants
complete a college preparatory curricu-
complete high school and enroll in and gradu-
lum while others complete a set of courses
ate from college. Rigorous evidence regard-
that does not prepare them well to succeed
ing the effectiveness of these postsecondary
in education or training past high school.
readiness reforms is relatively small, however.
Instead, these states are requiring all high As a result, we focus on results from those
school students to enroll in coursework that studies that used the most rigorous methods
aligns with postsecondary entrance require- available. To begin, we discuss two TRIO
ments. By 2015, at least twenty-one states and programs focused on connecting high school
the District of Columbia will have default students from low-income and first generation
curriculum requirements in place; these typi- college-going families to college—Upward
cally call for four years of English and math- Bound and Talent Search; we then present
ematics and at least three years of science or findings on GEAR UP. We also summarize
social science, or both.42 research on MCHCs and ECHSs to provide
VOL. 23 / NO. 1 / SPRING 2013 125
Andrea Venezia and Laura Jaeger
information and evidence about systemic included about 5,000 Talent Search partici-
approaches, but evaluations of systemic pants along with a comparison sample of
reform efforts tend to be less rigorous, both more than 70,000 students created through
because a control or comparison group is propensity score matching report positive
not easy to construct and because method- effects on high school completion and college
ologically sound evaluations are often unaf- enrollment. In both cases, Talent Search
fordable for small-scale precollege outreach participants completed high school at a
programs.44 significantly higher rate (86 percent in Texas;
85 percent in Florida) than did comparison
From a methodological perspective, experi- group students (77 percent in Texas; 70 per-
mental design is particularly useful when cent in Florida).48
addressing evaluation questions about the
effectiveness of programs or other interven-
tions, because it provides the strongest data
Evaluations of Upward
possible about whether observed outcomes
Bound, Talent Search, and
are the result of a given program or inno-
vation. Experimental designs include the
GEAR UP have yielded mixed
random assignment of students either to a
findings on the programs’
treatment group, which receives the inter-
vention, or a control group, which does not.
impact on the high school
Any variation in outcomes may be attributed
courses participants take—
to the intervention.45 When it is not feasible
to assign participants randomly to treat-
the number one predictor
ment and control groups, researchers may
of college readiness. Results
use quasi-experimental designs, including
regression discontinuity, difference-in-
on longer-run outcomes for
difference, interrupted time series, and
Talent Search have been more
propensity score matching. Regression
discontinuity is differentiated from the
positive, however.
other quasi-experimental designs because
researchers maintain control over the treat-
ment; participants are assigned to a program
or comparison group on the basis of a cutoff In contrast, findings for postsecondary
score on a preprogram measure.46 enrollment and completion were more mixed
for Upward Bound participants. Mathematica
Evaluations of Upward Bound, Talent Policy Research conducted a randomized
Search, and GEAR UP have yielded mixed assignment study with a nationally repre-
findings on the programs’ impact on the sentative sample of sixty-seven Upward
high school courses participants take—the Bound projects hosted by two- and four-year
number one predictor of college readiness.47 colleges and universities. Researchers found
Results on longer-term outcomes for Talent that the program had no detectable effect on
Search have been more positive, however. the rate of overall postsecondary enrollment,
For example, two quasi-experimental studies the type or selectivity of the postsecond-
of Talent Search in Texas and Florida that ary institution attended, or the likelihood
126 THE FUTURE OF CHILDREN