View in EDS HTML Full Text PDF Full Text

Avid Participation in High School and Post-Secondary Success: An Evaluation and Cost Analysis

Saved in:
Bibliographic Details
Title: Avid Participation in High School and Post-Secondary Success: An Evaluation and Cost Analysis
Language: English
Authors: Todhunter-Reid, Abigail (ORCID 0000-0002-4677-8235), Burke, Amanda, Houchens, Paul, Howard, Mark
Source: Journal of Research on Educational Effectiveness. 2020 13(4):679-701.
Availability: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
Peer Reviewed: Y
Page Count: 23
Publication Date: 2020
Document Type: Journal Articles
Reports - Research
Education Level: High Schools
Secondary Education
Descriptors: Program Evaluation, Student Participation, High School Students, Cost Effectiveness, High School Graduates, Enrollment, Predictor Variables, College Preparation, Low Income Students, Minority Group Students
Geographic Terms: Florida
DOI: 10.1080/19345747.2020.1772421
ISSN: 1934-5747
Abstract: The purpose of this study is to evaluate the Advancement Via Individual Determination (AVID) program and to derive estimates of program costs. We used the Coarsened Exact Matching approach to match AVID students with non-AVID students on 40 baseline characteristics. After matching, we estimated group differences in high-school graduation and college enrollment. We used the ingredients method to estimate program costs and calculated cost-effectiveness ratios by the duration of participation. Findings indicate that students who complete at least one AVID elective have higher high school graduation and college enrollment rates than comparable non-AVID students. We discuss how AVID compares to other college outreach programs in terms of costs and effects.
Abstractor: As Provided
What Works Clearinghouse (WWC) Reviewed: Meets Evidence Standards with Reservations
WWC Study Page: https://ies.ed.gov/ncee/wwc/Study/90255
Entry Date: 2020
Accession Number: EJ1275427
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwGo90qJprSxCmzSTxV1HkxzAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDJAJD0S3QETzMf3LnwIBEICBm6H8C1MLMQvgdTfCfr6eahslBz8QqIFSARhsUfOvWJRjmPK0_RoaEmuI9bQ3ztCv8_ngL0By5KHOLhqhlq2URqVJFVGlP9sgveULrE8D9hWBm6bnd-ZdLPCV5Smv7w3V1iebPKXV0vAQKromu-x0gtWstIcSz4Pb0Mccpm6hc4_eTme4ED96F_hPVK_ODJXmEBTsYZKP1EcdeqUn
Text:
  Availability: 1
  Value: <anid>AN0146971614;[5ew9]01oct.20;2020Nov16.03:37;v2.2.500</anid> <title id="AN0146971614-1">AVID Participation in High School and Post-Secondary Success: An Evaluation and Cost Analysis </title> <p>The purpose of this study is to evaluate the Advancement Via Individual Determination (AVID) program and to derive estimates of program costs. We used the Coarsened Exact Matching approach to match AVID students with non-AVID students on 40 baseline characteristics. After matching, we estimated group differences in high-school graduation and college enrollment. We used the ingredients method to estimate program costs and calculated cost-effectiveness ratios by the duration of participation. Findings indicate that students who complete at least one AVID elective have higher high school graduation and college enrollment rates than comparable non-AVID students. We discuss how AVID compares to other college outreach programs in terms of costs and effects.</p> <p>Keywords: AVID; evaluation; cost-effectiveness; college enrollment; CEM</p> <hd id="AN0146971614-2">Introduction</hd> <p>College enrollment rates in the U.S. have changed little over the last near-decade and the gaps in access to post-secondary institutions are persistent. To address the inequities in post-secondary outcomes, college outreach programs such as Talent Search, College Bound, and Advancement Via Individual Determination (AVID) have been designed to support disadvantaged students in high school. The mechanisms of these programs vary. However, their common goal is to increase the college enrollment rates of underrepresented students.</p> <p>Some college outreach programs have been rigorously evaluated with mixed results. However, few researchers have addressed program costs and fewer have conducted comparative studies to examine which programs are the most efficient at producing the desired outcomes. The purpose of this study is to (a) evaluate the AVID program in one south Florida school district, (b) estimate program implementation costs, and (c) calculate cost-effectiveness (CE) ratios by duration of program participation. This way, the effects and costs of AVID can be compared to the effects and costs of alternative programs with similar goals.</p> <p>In 2017, 62% of Black, 60% of Latinx, and 71% of White high school completers enrolled in college, and the gaps in these post-secondary success rates have changed little since 2010 (U.S. Department of Education, National Center for Education Statistics, [<reflink idref="bib30" id="ref1">30</reflink>], Table 302.20). However, in 2010, 51% of low-income and 82% of high-income high-school completers enrolled in college, and this gap has changed notably in recent years. In 2016, 65% of low-income and 83% of high-income high-school completers enrolled in college (U.S. Department of Education, National Center for Education Statistics, [<reflink idref="bib29" id="ref2">29</reflink>], Table 302.30).</p> <p>College outreach may be one intervention that has contributed to the increase in college enrollment rates for low-income students. In 2006, the federal government allocated $824 million to fund TRIO programs, and these programs served nearly 900,000 students (U.S. Department of Education, [<reflink idref="bib27" id="ref3">27</reflink>]). TRIO programs include Talent Search and Upward Bound and constitute only a subset of the many college outreach programs adopted in the United States.</p> <p>College outreach programs aim to address the precipitating factors that lead to the under-representation of low-income and non-White students in college. Some of these factors include limited exposure to rigorous coursework in high school (Handwerk et al., [<reflink idref="bib10" id="ref4">10</reflink>]), low academic aspirations and expectations (Zhan & Sherraden, [<reflink idref="bib32" id="ref5">32</reflink>]), and limited guidance and resources in the home to help navigate the college selection and application process (Dumais & Ward, [<reflink idref="bib7" id="ref6">7</reflink>]). Therefore, most college outreach programs aim to expose participants to instructional rigor, social support, and post-secondary opportunities.</p> <p>Although college outreach programs are widely adopted, rigorous evaluations of these programs produce mixed results. Some results from quasi-experiments—where program participants are matched with non-participants on baseline characteristics—indicate that participants enroll in college at higher rates than their matched counterparts. Le et al. ([<reflink idref="bib17" id="ref7">17</reflink>]) estimated a 15.4 percentage point increase in college enrollment for students who participated in College Bound. Bowden and Belfield ([<reflink idref="bib3" id="ref8">3</reflink>]) estimated an 11.3 percentage point increase in college enrollment, on average across sites, for students who participated in Talent Search.</p> <p>Other results from rigorous studies are less clear. Mathematica Policy Research conducted a randomized controlled experiment to evaluate the Upward Bound program (Myers et al., [<reflink idref="bib22" id="ref9">22</reflink>]). The only effect of Upward Bound was found for students who had low educational expectations at baseline (i.e., a five percentage point difference in college enrollment rates). Domina ([<reflink idref="bib6" id="ref10">6</reflink>]) used a propensity score matching approach and data from the Educational Longitudinal Survey of 2002 to determine whether college outreach of any kind is associated with college enrollment. Domina ([<reflink idref="bib6" id="ref11">6</reflink>]) found that there was no significant difference in college enrollment rates between students who participated in college outreach and their matched counterparts.</p> <p>Given the wide adoption of college outreach programs and the limited evidence that they are widely effective, comparative research is needed to determine which programs are the most cost-effective. Along with estimates of program effects, there is considerable variation in the costs of implementation across programs. Omitting cost information from evaluations leaves educational leaders with limited evidence to inform their program adoption decisions. For example, without cost information, decision-makers will adopt programs with the largest effects when there are more cost-effective alternatives (Hollands et al., [<reflink idref="bib12" id="ref12">12</reflink>]).</p> <p>Although economic evaluations have been used in fields such as medicine and marketing for decades, uptake of this approach has been slow in the field of education (Levin & Belfield, [<reflink idref="bib18" id="ref13">18</reflink>]). The goal of this research is to contribute to the nascent and growing body of CE research in education by estimating CE ratios for one college outreach program: AVID. These CE ratios can be compared to the CE ratios of similar programs to determine which are the most efficient at increasing college enrollment rates for disadvantaged students.</p> <hd id="AN0146971614-3">The AVID Program</hd> <p>AVID is a college outreach program designed to engage and support low-income and ethnic minority students in secondary school and to prepare them for post-secondary life. Compared to other programs with similar goals such as Talent Search, which served nearly 310,000 students nation-wide in 2018 (U.S. Department of Education, [<reflink idref="bib28" id="ref14">28</reflink>]), AVID has a broad reach. In 2018, AVID served over 500,000 secondary students in the United States. 64% of these students were free or reduced-price lunch eligible and 78% were non-White. The School District of Palm Beach County is 1 of 36 school districts in the state of Florida that has adopted AVID. The program was first piloted in 2010, and by 2017, AVID was being used in twelve high schools in the District.</p> <p>The AVID high-school program is a set of 4-year-long electives that are taken in order, with each course designed to be more rigorous than the last. Students in the participating schools have the option to take the initial AVID elective during any year of high school or to take the AVID elective course series and participate for multiple years. The decision to persist in AVID is determined by the student and their teachers on a rolling basis.</p> <p>The AVID program is theorized to leverage several mechanisms that contribute to post-secondary success. The primary mechanisms are exposure to college options, instructional rigor, social support, and the development of effective study habits. In the AVID program, students discuss college often, classrooms are decorated with college paraphernalia, and students request and examine pamphlets and brochures from their dream institutions. The AVID instructional materials are designed to be rigorous, to engage students in debate, creative problem-solving, and in addressing real-world problems. Twice a week in the AVID tutorial, students work together to collaboratively solve the challenges they are facing in their English language arts and mathematics courses. And throughout the AVID elective, students are systematically trained on how to take notes, stay organized, and study for exams. Together these mechanisms are theorized to prepare students for advanced coursework and the rigor of college-level academics (Bendall et al., [<reflink idref="bib1" id="ref15">1</reflink>]).</p> <p>Student selection into the AVID elective is determined by the school and student characteristics. The standardized selection process for AVID participation as implemented nation-wide includes targeting students who have the potential to be first-generation college-goers (<ulink href="http://AVID.org">http://AVID.org</ulink>). These students are generally defined as mid-ranged achievers who originate from traditionally underserved populations. However, school districts define student potential and mid-range achievement using their own local metrics.</p> <p>In the District, during the years under evaluation, the AVID elective was only offered in Title I schools and targeted students who were low-income and non-White. AVID participants needed to demonstrate the potential for success in college. Therefore, selected participants scored near the proficiency cutoff point on the state-wide English language arts and mathematics assessments. That is, students who scored close to the proficiency cut off on the state assessments where deemed in need of intervention while also possessing the requisite skills to engage with rigorous course material.</p> <p>Students are nominated for AVID participation by themselves, their teachers, and their principals. In the District, teachers and guidance counselors ultimately make the decision of which students are enrolled in the AVID elective. As dictated by the standardized AVID program, parents and guardians of the selected students are invited to meet with the AVID elective teacher and must sign a contract which states their commitment to their child's education. If the parent or guardian does not sign the AVID contract, their child is not permitted to enroll in AVID.</p> <hd id="AN0146971614-4">Past Research on AVID</hd> <p>Although researchers have conducted evaluations of AVID, the available evidence does not allow for robust inferences or causal claims. This is because most researchers have used descriptive statistics to tell their data stories and applied simple statistical tests like <emph>t</emph>-tests to determine if differences between AVID and non-AVID students are statistically significant. This is problematic because students who take AVID electives are, by design, different in many ways from students who do not take AVID electives. Evaluations that do not carefully account for preexisting differences between compared students are susceptible to selection bias. Therefore, it is difficult to determine whether any observed differences in outcomes between AVID and non-AVID students result from program participation.</p> <p>Most of the descriptive studies on AVID report positive relationships. For example, Huerta et al. ([<reflink idref="bib13" id="ref16">13</reflink>]) found that AVID students take more advanced courses than non-AVID students. Franklin ([<reflink idref="bib8" id="ref17">8</reflink>]) found that AVID students are more college-ready in terms of attendance, advanced course enrollment, and test scores than non-AVID students. And Black et al. ([<reflink idref="bib2" id="ref18">2</reflink>]) found that AVID students have more college plans and fewer absences than non-AVID students. However, these studies did not adjust for preexisting student characteristics.</p> <p>One recent quasi-experimental study took a rigorous approach to matching and examined the impact of a combined AVID and Teens of Promise (TOPS) program in one school district in Wisconsin (Kolbe et al., [<reflink idref="bib16" id="ref19">16</reflink>]). TOPS is an enhancement program designed to reinforce AVID goals through community mentoring and summer programming. Kolbe et al. ([<reflink idref="bib16" id="ref20">16</reflink>]) used a propensity score matching approach to create a comparison group and matched AVID/TOPS students with non-AVID/TOPS students on 18 baseline characteristics. These baseline characteristics included student demographics, parental education variables, and several measures of achievement. The authors found that AVID/TOPS students had higher rates of advanced course-taking, high school graduation, and college enrollment than students in the matched comparison group. The authors were unable to differentiate between the program effects of AVID and TOPS because all AVID participants were also involved with the TOPS program.</p> <p>Although these findings suggest that AVID is effective at improving an array of secondary and post-secondary outcomes, not all AVID studies uncovered positive relationships. One rigorous study—indeed the only AVID study that the What Works Clearinghouse deemed high-quality (What Works Clearinghouse, [<reflink idref="bib31" id="ref21">31</reflink>])—examined the effects of AVID course taking in high school on reading test scores. Students who took AVID electives were matched with students in the same schools who did not take AVID electives on ethnicity, gender, age, and baseline reading achievement. No discernable effects of AVID were detected. However, the study sample was small (<emph>n</emph> = 96), and program effects are difficult to detect in small samples.</p> <p>Findings from other correlational research indicate that there are no significant differences in outcomes between students who take AVID electives and students who do not (Connors, [<reflink idref="bib4" id="ref22">4</reflink>]; Dean, [<reflink idref="bib5" id="ref23">5</reflink>]; Peak, [<reflink idref="bib23" id="ref24">23</reflink>]; Smith et al., [<reflink idref="bib25" id="ref25">25</reflink>]). For example, Connors ([<reflink idref="bib4" id="ref26">4</reflink>]) examined the AVID program in two school districts in Florida. One hundred AVID students were exactly matched one-to-one with 100 non-AVID students who attended the same schools. The matching variables were gender, ethnicity, economic status, and baseline achievement. The authors found that AVID students and non-AVID students were not statistically different from each other on state measures of achievement in English language arts and mathematics.</p> <p>Some of the inconsistencies in AVID program effects may be due to variation in effects by the duration of program participation. It is a common finding that students who persist in AVID and complete multiple electives perform better than students who take only the initial AVID elective (Huerta et al., [<reflink idref="bib13" id="ref27">13</reflink>]; Kolbe et al., [<reflink idref="bib16" id="ref28">16</reflink>]; Mehan et al., [<reflink idref="bib20" id="ref29">20</reflink>]). For example, Huerta et al. ([<reflink idref="bib13" id="ref30">13</reflink>]) found that students who took AVID electives in both middle school and high school and for consecutive years outperformed students who took AVID electives only sporadically in high school. The long-term participants illustrated more college readiness in terms of test scores and enrollment in advanced coursework as seniors.</p> <p>Continued research is needed to provide a stronger evidence base on the effectiveness of AVID at increasing college enrollment rates. Past researchers have used small samples and limited methods. Those who made attempts at adjusting for baseline student characteristics (i.e., matched AVID students with non-AVID students) did so using a small number of covariates. This evaluation examines AVID in the 11th largest school district in the United States and matches AVID and non-AVID students on 40 baseline characteristics.</p> <hd id="AN0146971614-5">Economic Evaluations of College Outreach Programs</hd> <p>Economic evaluations have been used in fields such as medicine, transportation, and criminal justice for over 50 years. They provide information which allows practitioners to choose and implement the interventions that produce the greatest results for the least cost (Levin et al., [<reflink idref="bib19" id="ref31">19</reflink>]). Given the wide adoption of college outreach programs and the limited evidence that they are widely effective, comparative research is needed to determine which programs are the most cost-effective.</p> <p>Although little is known about the cost-effectiveness of college outreach programs, one comprehensive study examined the cost-effectiveness of drop-out prevention programs. Hollands et al. ([<reflink idref="bib12" id="ref32">12</reflink>]) used existing data to estimate CE ratios for a set of dropout prevention programs that the What Works Clearinghouse deemed to be effective. These programs include Talent Search, JOBSTART, National Guard Youth Challenge, and Job Corps. Of these programs, Talent Search produced the smallest CE ratio (Hollands et al., [<reflink idref="bib12" id="ref33">12</reflink>]).</p> <p>The Talent Search program provides academic, college, career, and financial counseling to average academic performers at school sites and is designed to increase high school graduation and college enrollment rates. The AVID program is comparable to the Talent Search program as both programs serve similar populations and share common goals. However, the other programs evaluated by Hollands et al. ([<reflink idref="bib12" id="ref34">12</reflink>]) are geared toward students who have already dropped out of high school. Therefore, they are not comparable to AVID.</p> <p>Kolbe et al. ([<reflink idref="bib16" id="ref35">16</reflink>]) conducted a thorough cost analysis of the AVID/TOPS program and estimated that it costs $4,440 per pupil per year to implement. Using a CE approach, the authors estimated that it costs $271,490 and $79,600 to produce each additional high school graduate and college enrollee, respectively. In a comprehensive CE analysis of Talent Search, Bowden and Belfield ([<reflink idref="bib3" id="ref36">3</reflink>]) found that each additional high school graduate produced by the program costs $43,440. Each additional college enrollee costs $33,380.</p> <p>The CE ratios for Talent Search are notably smaller than the CE ratios for AVID/TOPS which indicates that the former is more cost-effective than the latter. However, the latter cost estimates reflect the costs of implementing the AVID and the TOPS programs, simultaneously. Little is known about the cost-effectiveness of AVID as a standalone program.</p> <hd id="AN0146971614-6">The Current Study</hd> <p>The primary goal for this research is to examine the relationship of AVID elective course-taking with post-secondary outcomes after matching AVID students with non-AVID students on 40 baseline characteristics. A secondary goal of this research is to examine whether the relationship of AVID elective course-taking with post-secondary outcomes varies by duration of participation. Finally, this research is also designed to estimate AVID program costs and determine what level of participation (i.e., long-term vs. short-term) is the most cost-effective.</p> <p>In summary, this study is designed to answer the following questions:</p> <p></p> <ulist> <item> To what extent is AVID elective course-taking in high school related to (a) on-time high school graduation and (b) college enrollment the fall after high school graduation?</item> <p></p> <item> Do the effects of AVID elective course-taking vary by duration of participation?</item> <p></p> <item> What is the annual per pupil cost of AVID program implementation?</item> <p></p> <item> Is there a greater return on investment when students persist in AVID and take three to four electives than when students take only one or two electives?</item> </ulist> <p>Drawing on the AVID program literature, we hypothesized that AVID elective course-taking in high school is positively associated with on-time high school graduation and college enrollment. We also hypothesized that the effects will increase in magnitude the longer students are involved with AVID. The economic portion of the evaluation is exploratory. Therefore, we did not formulate a hypothesis regarding which mode of program participation is the most cost-effective.</p> <hd id="AN0146971614-7">Method</hd> <p></p> <hd id="AN0146971614-8">Description of Sample</hd> <p>We drew the study sample from a population of students enrolled in the School District of Palm Beach County, Florida—the 11th largest school district in the United States. The District serves ∼190,000 students annually, 59% of whom are free or reduced-price lunch eligible, and 70% are non-White. Three cohorts of students were included in this study: the 2015 graduation cohort, the 2016 graduation cohort, and the 2017 graduation cohort. We tracked students in these cohorts over six years from 8th-grade to one-year post-high-school graduation.</p> <p>The entire study sample consists of 44,813 students. However, only students who were enrolled in the District while in 8th grade and had all relevant baseline and outcome data were included in this study. The resulting sample for analysis consists of 29,485 students. The greatest loss of data came when students without baseline state test scores were dropped from the sample. This accounted for 67% of the missing data. Although this is a notable loss, students with missing baseline achievement data are not eligible for participation in AVID. Indeed, none of the students who had missing baseline achievement data were recorded as AVID course-takers. Therefore, dropping students in the comparison sample with missing achievement data improves the comparison for evaluation.</p> <p>Of the students in the study sample, 1,329 completed at least one AVID elective from 9th-grade to 12th-grade. Of the students enrolled in AVID, 52% were female, 53% were Black, 30% were Hispanic, 11% were White, and 79% were free and reduced-price lunch eligible. 75% of students enrolled in the AVID program over the study period were mid-range achievers at baseline. Mid-ranged achievers are those who scored close to the proficiency cutoff point on the Florida Standards Assessment. There are five achievement levels as defined by the state, and students scoring at a level three and higher are considered proficient. Therefore, the District defined mid-ranged achievers as having scored at an achievement level of two or three out of five on the Florida Standards Assessments. The approximate corresponding percentile ranges for these levels are 15th to 35th for level two students and 36th to 75th for level three students.</p> <hd id="AN0146971614-9">Measures</hd> <p>Unless otherwise noted, we drew all data from the District's Education Data Warehouse. All baseline data were drawn at the end of the year when students were exiting 8th-grade (pre-AVID exposure). Outcome data were drawn from the end of students' 12th-grade year and the fall following their high school graduation (post-AVID exposure).</p> <p>We used records of student course enrollment to identify <emph>AVID course-takers</emph>. Each of the four electives in the AVID course series is assigned a distinct course code. If a student had at least one record of AVID course enrollment, they were coded as AVID participants. An additional variable was generated to represent the total number of AVID electives each student took during high school.</p> <p>We used the Florida Department of Education diploma codes to identify <emph>high school graduates</emph>. For the purpose of this study, graduates include students who received any type of high school diploma, including a standard, performance-based, or special diploma. Graduates do not include students who received a certificate of completion. Students who transferred out of the District before graduation were excluded from the analysis.</p> <p>We obtained <emph>college enrollment</emph> data from the National Student Clearinghouse (NSC). The NSC tracks student enrollment in post-secondary institutions during the fall after high-school graduation. Ninety-nine percent of post-secondary institutions participate in the NSC, and it is the most comprehensive source of college enrollment data available. Students who did not have enrollment data in the NSC were assumed to have not enrolled in a post-secondary institution the fall after high-school graduation. Students who had a record of enrollment the fall after high-school graduation were coded as enrolled.</p> <hd id="AN0146971614-10">Analytic Approach</hd> <p></p> <hd id="AN0146971614-11">The Program Evaluation</hd> <p>Students who take AVID electives are, by design, different from students who do not take AVID electives. Therefore, it is plausible that reported differences in outcomes between AVID students and non-AVID students result from preexisting student differences rather than the AVID intervention itself. For this study, we used an advanced matching approach, Coarsened Exact Matching (CEM), to create the comparison groups. CEM is a data pre-processing algorithm that combines the benefits of exact matching, stratification, and weighting. A set of strata are generated with different numbers of exactly matched control and treatment units, and weights are generated to adjust for the variation in stratum size. If the strata have at least one treatment and one control unit then the units in those strata are retained. Otherwise, the units are discarded. Therefore, there is no need to manually eliminate units outside of the extrapolation region (i.e., the region of overlap between treatment and control units) because the CEM approach does this automatically (Iacus et al., [<reflink idref="bib14" id="ref37">14</reflink>]).</p> <p>The CEM approach is preferable to the more commonly applied propensity score matching approach due to its efficiency and robustness to measurement error, among other attributes (Iacus et al., [<reflink idref="bib14" id="ref38">14</reflink>]; King et al., [<reflink idref="bib15" id="ref39">15</reflink>]; Stuart, [<reflink idref="bib26" id="ref40">26</reflink>]). When CEM is applied, students are exactly matched on dichotomous variables and proximally matched on continuous variables that have been temporarily coarsened (i.e., binned) into categories. Like other matching approaches, CEM closely matches each unit in the treatment group to one or more units in the control group. This creates two groups comprised of students who are similar in every measurable regard except whether they received the intervention. When this is accomplished, the two groups can be compared to determine if students who receive the intervention, in this case, AVID, have better outcomes than students who do not receive the intervention.</p> <p>For causal effects to be estimated using any matching approach, including CEM, the ignorability assumption must be met. This means that all the differences between the treatment and control groups must be accounted for by the observed covariates included in the matching procedure (Stuart, [<reflink idref="bib26" id="ref41">26</reflink>]). The primary selection factors for AVID involvement are students' achievement levels in English Language Arts and Mathematics (i.e., the program targets students who are mid-range achievers). The AVID program also targets ethnic minority students and students from economically disadvantaged family backgrounds.</p> <p>In this study, AVID and non-AVID students were exactly matched on the key characteristics known to be related to treatment assignment and proximally matched on additionally available covariates until the two groups were balanced on 40 baseline student characteristics. A comprehensive list of these characteristics can be found in Table 1.</p> <p>Table 1. Baseline group differences (means and proportions) between students who took at least one AVID elective and students who did not take an AVID elective before and after CE matching (FY15–FY17 cohorts).</p> <p> <ephtml> <table><thead><tr><td>Student characteristic</td><td>Pre-CE matching (<p id="ilm0011"><graphic href="uree_a_1772421_ilm0011.gif" content-type="Graph" /><math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub xmlns=""><mrow><mi>L</mi></mrow><mrow><mn>1</mn></mrow></msub></math></p> = 0.73)</td><td>Post-CE matching (<p id="ilm0012"><graphic href="uree_a_1772421_ilm0012.gif" content-type="Graph" /><math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub xmlns=""><mrow><mi>L</mi></mrow><mrow><mn>1</mn></mrow></msub></math></p> ∼ 0.00)</td></tr><tr><td>AVID (<italic>n</italic> = 1,329)</td><td>Non-AVID (<italic>n</italic> = 28,156)</td><td>Difference</td><td>AVID (<italic>n</italic> = 1,157)</td><td>Non-AVID (<italic>n</italic> = 9,240)</td><td>Difference</td></tr></thead><tbody valign="top"><tr><td>Female</td><td char=".">0.516</td><td char=".">0.498</td><td char=".">0.018</td><td char=".">0.520</td><td char=".">0.513</td><td char=".">0.007</td></tr><tr><td>White</td><td char=".">0.111</td><td char=".">0.382</td><td char=".">−0.270***</td><td char=".">0.115</td><td char=".">0.115</td><td char=".">0.000</td></tr><tr><td>Black</td><td char=".">0.534</td><td char=".">0.271</td><td char=".">0.263***</td><td char=".">0.551</td><td char=".">0.551</td><td char=".">0.000</td></tr><tr><td>Hispanic</td><td char=".">0.299</td><td char=".">0.282</td><td char=".">0.018</td><td char=".">0.283</td><td char=".">0.291</td><td char=".">−0.008</td></tr><tr><td>Other race</td><td char=".">0.055</td><td char=".">0.065</td><td char=".">−0.011</td><td char=".">0.052</td><td char=".">0.044</td><td char=".">0.008</td></tr><tr><td>ELL enrolled</td><td char=".">0.093</td><td char=".">0.055</td><td char=".">0.037***</td><td char=".">0.086</td><td char=".">0.080</td><td char=".">0.007</td></tr><tr><td>ELL follow-up</td><td char=".">0.284</td><td char=".">0.173</td><td char=".">0.111***</td><td char=".">0.266</td><td char=".">0.273</td><td char=".">−0.007</td></tr><tr><td>Non-ELL</td><td char=".">0.623</td><td char=".">0.771</td><td char=".">−0.148***</td><td char=".">0.647</td><td char=".">0.647</td><td char=".">0.000</td></tr><tr><td>Homeless</td><td char=".">0.011</td><td char=".">0.005</td><td char=".">0.005*</td><td char=".">0.010</td><td char=".">0.007</td><td char=".">0.003</td></tr><tr><td>Free/reduced lunch</td><td char=".">0.792</td><td char=".">0.517</td><td char=".">0.275***</td><td char=".">0.793</td><td char=".">0.793</td><td char=".">0.000</td></tr><tr><td>Primary lang. English</td><td char=".">0.552</td><td char=".">0.697</td><td char=".">−0.145***</td><td char=".">0.573</td><td char=".">0.583</td><td char=".">−0.010</td></tr><tr><td>Primary lang. Haitian-Creole</td><td char=".">0.207</td><td char=".">0.086</td><td char=".">0.121***</td><td char=".">0.212</td><td char=".">0.205</td><td char=".">0.007</td></tr><tr><td>Primary lang. Spanish</td><td char=".">0.215</td><td char=".">0.170</td><td char=".">0.046***</td><td char=".">0.201</td><td char=".">0.198</td><td char=".">0.004</td></tr><tr><td>Primary lang. Other</td><td char=".">0.026</td><td char=".">0.048</td><td char=".">−0.022***</td><td char=".">0.014</td><td char=".">0.014</td><td char=".">0.000</td></tr><tr><td>Born in the U.S.</td><td char=".">0.807</td><td char=".">0.845</td><td char=".">−0.038***</td><td char=".">0.834</td><td char=".">0.831</td><td char=".">0.003</td></tr><tr><td>Born in Mexico</td><td char=".">0.010</td><td char=".">0.013</td><td char=".">−0.003</td><td char=".">0.005</td><td char=".">0.005</td><td char=".">0.000</td></tr><tr><td>Born in Haiti</td><td char=".">0.105</td><td char=".">0.047</td><td char=".">0.057***</td><td char=".">0.105</td><td char=".">0.108</td><td char=".">−0.003</td></tr><tr><td>Born in other country</td><td char=".">0.078</td><td char=".">0.095</td><td char=".">−0.017*</td><td char=".">0.055</td><td char=".">0.055</td><td char=".">0.000</td></tr><tr><td>Gifted</td><td char=".">0.008</td><td char=".">0.026</td><td char=".">−0.018***</td><td char=".">0.008</td><td char=".">0.009</td><td char=".">−0.002</td></tr><tr><td>Non-ESE</td><td char=".">0.900</td><td char=".">0.883</td><td char=".">0.017∼</td><td char=".">0.910</td><td char=".">0.911</td><td char=".">−0.001</td></tr><tr><td>Learn/behavioral disability</td><td char=".">0.079</td><td char=".">0.082</td><td char=".">−0.003</td><td char=".">0.082</td><td char=".">0.080</td><td char=".">0.002</td></tr><tr><td>Other disability</td><td char=".">0.021</td><td char=".">0.035</td><td char=".">−0.014**</td><td char=".">0.008</td><td char=".">0.008</td><td char=".">0.000</td></tr><tr><td>Migrant</td><td char=".">0.033</td><td char=".">0.016</td><td char=".">0.017***</td><td char=".">0.017</td><td char=".">0.017</td><td char=".">0.000</td></tr><tr><td>ELA level 1</td><td char=".">0.092</td><td char=".">0.121</td><td char=".">−0.029**</td><td char=".">0.082</td><td char=".">0.082</td><td char=".">0.000</td></tr><tr><td>ELA level 2</td><td char=".">0.391</td><td char=".">0.254</td><td char=".">0.137***</td><td char=".">0.392</td><td char=".">0.392</td><td char=".">0.000</td></tr><tr><td>ELA level 3</td><td char=".">0.355</td><td char=".">0.303</td><td char=".">0.052***</td><td char=".">0.366</td><td char=".">0.366</td><td char=".">0.000</td></tr><tr><td>ELA level 4</td><td char=".">0.137</td><td char=".">0.212</td><td char=".">−0.076***</td><td char=".">0.139</td><td char=".">0.139</td><td char=".">0.000</td></tr><tr><td>ELA level 5</td><td char=".">0.026</td><td char=".">0.110</td><td char=".">−0.085***</td><td char=".">0.021</td><td char=".">0.021</td><td char=".">0.000</td></tr><tr><td>Math level 1</td><td char=".">0.106</td><td char=".">0.110</td><td char=".">−0.004</td><td char=".">0.105</td><td char=".">0.105</td><td char=".">0.000</td></tr><tr><td>Math level 2</td><td char=".">0.226</td><td char=".">0.158</td><td char=".">0.068***</td><td char=".">0.231</td><td char=".">0.231</td><td char=".">0.000</td></tr><tr><td>Math level 3</td><td char=".">0.441</td><td char=".">0.343</td><td char=".">0.098***</td><td char=".">0.459</td><td char=".">0.459</td><td char=".">0.000</td></tr><tr><td>Math level 4</td><td char=".">0.173</td><td char=".">0.226</td><td char=".">−0.053***</td><td char=".">0.162</td><td char=".">0.162</td><td char=".">0.000</td></tr><tr><td>Math level 5</td><td char=".">0.053</td><td char=".">0.163</td><td char=".">−0.110***</td><td char=".">0.042</td><td char=".">0.042</td><td char=".">0.000</td></tr><tr><td>ELA scale score</td><td char=".">235.44</td><td char=".">240.75</td><td char=".">−5.31***</td><td char=".">235.45</td><td char=".">235.15</td><td char=".">0.30</td></tr><tr><td>Math scale score</td><td char=".">244.20</td><td char=".">248.92</td><td char=".">−4.72***</td><td char=".">243.70</td><td char=".">243.59</td><td char=".">0.11</td></tr><tr><td>Advanced courses taken</td><td char=".">1.70</td><td char=".">2.11</td><td char=".">−0.41***</td><td char=".">1.67</td><td char=".">1.60</td><td char=".">0.07</td></tr><tr><td>Credits earned</td><td char=".">0.67</td><td char=".">0.90</td><td char=".">−0.23***</td><td char=".">0.65</td><td char=".">0.63</td><td char=".">0.02</td></tr><tr><td>Disciplinary referrals</td><td char=".">0.66</td><td char=".">0.64</td><td char=".">0.02</td><td char=".">0.63</td><td char=".">0.72</td><td char=".">−0.08</td></tr><tr><td>Out of school suspensions</td><td char=".">0.19</td><td char=".">0.18</td><td char=".">0.00</td><td char=".">0.18</td><td char=".">0.20</td><td char=".">−0.01</td></tr><tr><td>Days absent</td><td char=".">4.07</td><td char=".">5.34</td><td char=".">−1.27***</td><td char=".">3.97</td><td char=".">3.91</td><td char=".">0.05</td></tr></tbody></table> </ephtml> </p> <p>1 <emph>Note</emph>: Significance levels were derived from unconditional OLS and logistic regression models.</p> <p>2 ∼<emph>p</emph> < 0.10; *<emph>p</emph> < 0.05; **<emph>p</emph> < 0.01; ***<emph>p</emph> < 0.001.</p> <p>Most students in the study sample (66.0%) completed one to two AVID electives while the remaining students completed three to four AVID electives. As described by Stuart ([<reflink idref="bib26" id="ref42">26</reflink>]), when evaluating treatments with a level of dosage, separate matches can be made at multiple levels of exposure to determine which level is the most strongly related to the outcomes of interest. Generating multiple matched groups is also important because students who persist in AVID and take three to four electives in the course series may be different in baseline characteristics from students who take only one to two electives in the course series. For example, students who persist in AVID may be more behaviorally engaged and academically motivated than students who do not persist in AVID. Therefore, different comparison groups are generated for those students who persist in AVID and those who do not persist in AVID.</p> <p>To this end, three separate matches were made. Students who took any number of AVID electives, students who took one to two AVID electives, and students who took three to four AVID electives were matched with their own comparison groups comprised of students who did not take an AVID elective.</p> <p>Through the matching process, all samples were reduced to only include students with (a) exact matches on key variables (i.e., prior achievement levels, race and ethnicity, free and reduced lunch eligibility, disability status, English language learner status, and cohort membership) and (b) proximal matches on all other observed covariates. After matching, we made simple group comparisons to determine whether students who took AVID electives had more favorable post-secondary outcomes than similar students who did not take AVID electives. Significance levels for the effects were estimated using unconditional logistic regression models. Results from the full logistic regression models that include covariates can be found in the Supplemental Materials (Supplementary Appendix B, Table B1).</p> <p>The generalized equation for the matching algorithm is</p> <p>Graph</p> <p> <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>Y</mi></mrow><mrow><mi>i</mi></mrow></msub></math> </ephtml> =</p> <p>Graph</p> <p> <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>T</mi></mrow><mrow><mi>i</mi></mrow></msub></math> </ephtml> |</p> <p>Graph</p> <p> <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>X</mi></mrow><mrow><mi>i</mi></mrow></msub></math> </ephtml> where for each unit <emph>i</emph> (<emph>i = 1, ..., n</emph>) the treatment variable</p> <p>Graph</p> <p> <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>T</mi></mrow><mrow><mi>i</mi></mrow></msub></math> </ephtml> takes on the value of <emph>0</emph> if the unit is in the comparison group and <emph>1</emph> if the unit is in the AVID program group.</p> <p>Graph</p> <p> <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>X</mi></mrow><mrow><mi>i</mi></mrow></msub></math> </ephtml> represents a set of <emph>K</emph> covariates measured before AVID exposure, and</p> <p>Graph</p> <p> <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>Y</mi></mrow><mrow><mi>i</mi></mrow></msub></math> </ephtml> is the measured outcome. The treatment effect is then estimated as the simple difference in outcomes between the AVID group and the comparison group:</p> <p>Graph</p> <p> <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>T</mi><mi>E</mi></mrow><mrow><mi>i</mi></mrow></msub></math> </ephtml> =</p> <p>Graph</p> <p> <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>Y</mi></mrow><mrow><mi>i</mi></mrow></msub></math> </ephtml> <emph>(<reflink idref="bib1" id="ref43">1</reflink>) −</emph> </p> <p>Graph</p> <p> <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>Y</mi></mrow><mrow><mi>i</mi></mrow></msub></math> </ephtml> <emph>(0)</emph>.</p> <p>One important limitation of the matching procedure applied in this study is that not all selection factors are observed. A criterion for selection into AVID is the potential for success in college as identified by school personnel. Therefore, students selected for AVID participation possess a latent motivational trait that may be particularly salient for students who persist in AVID and take multiple electives in the course series. Moreover, parent involvement is an important selection factor because parents must sign an AVID contract which affirms their commitment to their students' education. If a parental figure is not available to sign the contract, their student is not allowed to enroll in the AVID program. Measures of student motivation and parental involvement are not housed in the District's Education Data Warehouse and were not included in the matching procedure. These unobserved characteristics may be partially captured by the measures of behavioral engagement included in the matching algorithm (i.e., attendance and discipline). However, selection bias cannot be ruled out as a threat to internal validity.</p> <p>An additional limitation of the matching procedure is that treatment and comparison students were not matched on school characteristics. This is because there was virtually no overlap in characteristics between schools implementing AVID and schools not implementing AVID. All schools implementing AVID were Title I schools and served large proportions of low-income and ethnic minority students. In contrast, non-AVID schools served large proportions of high-income and White students. At baseline, AVID schools served an average of 75% free and reduced-price lunch eligible students and 85% minority students. Non-AVID schools served an average of 26% free and reduced-price lunch eligible students and 42% minority students.</p> <p>Although the omission of school-level variables in the matching process is a limitation, it is not considered a severe threat to internal validity. This is because non-Title I schools serve more affluent student populations in safer neighborhoods and attract more highly qualified teachers. The positive neighborhood, peer, and teacher effects could all lead students in the comparison group to achieve greater post-secondary success than the students in the AVID groups. Therefore, the omission of school-level variables likely biases the results in the negative direction (i.e., making the AVID program appear less effective than it is). As a sensitivity check, we conducted additional analyses, with the same matching procedure used in the core analysis, that only included students in Title I schools.</p> <hd id="AN0146971614-12">The Cost Analysis</hd> <p>We designed the cost analysis to estimate the cost of AVID program participation per pupil per year. The annual cost of AVID implementation varies by year depending on the number of schools that implement the program and the number of electives offered at each school. Therefore, the annual cost of AVID implementation was estimated by taking the average annual cost across the six years under investigation. As recommended by leading education researchers, we applied the ingredients method to identify the cost components needed to implement the program elsewhere (Levin et al., [<reflink idref="bib19" id="ref44">19</reflink>]). These ingredients include personnel, training, facilities, and materials.</p> <p>We gathered data regarding the number of schools offering AVID electives, the number of electives offered, and the number of students served annually from the Education Data Warehouse and the AVID administrative site for the District. Through extensive interviews with AVID District staff and document review, budgetary and opportunity cost ingredients were identified, and local costs were estimated in 2018 dollars.</p> <hd id="AN0146971614-13">Results</hd> <p></p> <hd id="AN0146971614-14">Overall Effects of Any AVID Exposure</hd> <p>Table 1 presents the mean baseline differences between students who took at least one AVID elective from 9th grade to 12th grade and students who did not take an AVID elective, before and after matching. At baseline in 8th-grade, there are many significant differences between the students who go on to take an AVID elective and the students who do not. In the AVID group, there are larger proportions of Black, English language learner, free and reduced-price lunch eligible, and mid- to low-performing students than in the non-AVID group. Conversely, there are larger proportions of White, English speaking, gifted, and high-performing students in the non-AVID group than in the AVID group.</p> <p>These findings reinforce the need, described in the Methods section, to adjust for preexisting student differences before estimating the effects of the AVID program. To this end, we applied the CEM approach to create a comparison group comprised of students who are like the AVID students in every measurable way. Students were matched on 40 baseline characteristics, and the multivariate imbalance (as measured by the</p> <p>Graph</p> <p> <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>L</mi></mrow><mrow><mn>1</mn></mrow></msub></math> </ephtml> distance) between the AVID and non-AVID groups were eliminated. The approach retained 87% of AVID participants in the matched sample.</p> <p>Table 2 presents the differences in outcomes between students who completed at least one AVID elective from 9th-grade to 12th-grade and matched students who did not complete an AVID elective. The graduation rate of the AVID group (93.9%) is significantly higher than the graduation rate of the non-AVID group (90.5%). The college enrollment rate of the AVID group (59.9%) is also significantly higher than the college enrollment rate of the non-AVID group (56.4%). Results broken down by cohort year can be found in the Supplemental Materials (Supplementary Appendix A, Tables A1–A3).</p> <p>Table 2. Differences in post-secondary outcomes (post CE matching) between students who took at least one AVID elective and students who did not take an AVID elective (FY15–FY17 cohorts).</p> <p> <ephtml> <table><thead><tr><td>Outcome</td><td>AVID (<italic>n</italic> = 1,157)</td><td>Non-AVID (<italic>n</italic> = 9,240)</td><td>Difference</td></tr></thead><tbody valign="top"><tr><td>Graduated HS on time</td><td char=".">0.939</td><td char=".">0.905</td><td char=".">0.033***</td></tr><tr><td>Enrolled in college</td><td char=".">0.599</td><td char=".">0.564</td><td char=".">0.035*</td></tr></tbody></table> </ephtml> </p> <ulist> <item>3 <emph>Note</emph>: Significance levels were derived from unconditional logistic regression models.</item> <item>4 *<emph>p</emph> < 0.05; ***<emph>p</emph> < 0.001.</item> </ulist> <p>Table 3. Baseline group differences between students who took one to two AVID electives and students who took three to four AVID electives (FY15–FY17 cohorts).</p> <p> <ephtml> <table><thead><tr><td>Student characteristics</td><td>1–2 electives (<italic>n</italic> = 877)</td><td>3–4 electives (<italic>n</italic> = 452)</td><td>Difference</td></tr></thead><tbody valign="top"><tr><td>Female</td><td char=".">0.503</td><td char=".">0.542</td><td char=".">0.039</td></tr><tr><td>White</td><td char=".">0.107</td><td char=".">0.119</td><td char=".">0.012</td></tr><tr><td>Black</td><td char=".">0.535</td><td char=".">0.533</td><td char=".">−0.002</td></tr><tr><td>Hispanic</td><td char=".">0.314</td><td char=".">0.272</td><td char=".">−0.041</td></tr><tr><td>Other race</td><td char=".">0.044</td><td char=".">0.075</td><td char=".">0.031*</td></tr><tr><td>ELL enrolled</td><td char=".">0.096</td><td char=".">0.086</td><td char=".">−0.009</td></tr><tr><td>ELL exited</td><td char=".">0.303</td><td char=".">0.248</td><td char=".">−0.056*</td></tr><tr><td>Non-ELL</td><td char=".">0.601</td><td char=".">0.666</td><td char=".">0.065*</td></tr><tr><td>Homeless</td><td char=".">0.011</td><td char=".">0.009</td><td char=".">−0.003</td></tr><tr><td>Free/reduced lunch</td><td char=".">0.810</td><td char=".">0.759</td><td char=".">−0.051*</td></tr><tr><td>Primary lang. English</td><td char=".">0.531</td><td char=".">0.593</td><td char=".">0.062*</td></tr><tr><td>Primary lang. Haitian-Creole</td><td char=".">0.217</td><td char=".">0.188</td><td char=".">−0.029</td></tr><tr><td>Primary lang. Spanish</td><td char=".">0.226</td><td char=".">0.195</td><td char=".">−0.031</td></tr><tr><td>Primary lang. other</td><td char=".">0.026</td><td char=".">0.024</td><td char=".">−0.002</td></tr><tr><td>Born in the U.S.</td><td char=".">0.796</td><td char=".">0.830</td><td char=".">0.034</td></tr><tr><td>Born in Mexico</td><td char=".">0.011</td><td char=".">0.007</td><td char=".">−0.005</td></tr><tr><td>Born in Haiti</td><td char=".">0.111</td><td char=".">0.093</td><td char=".">−0.018</td></tr><tr><td>Born in other country</td><td char=".">0.082</td><td char=".">0.071</td><td char=".">−0.011</td></tr><tr><td>Gifted</td><td char=".">0.007</td><td char=".">0.009</td><td char=".">0.002</td></tr><tr><td>Non-ESE</td><td char=".">0.910</td><td char=".">0.881</td><td char=".">−0.029∼</td></tr><tr><td>Learn/behavioral disability</td><td char=".">0.072</td><td char=".">0.093</td><td char=".">0.021</td></tr><tr><td>Other disability</td><td char=".">0.018</td><td char=".">0.027</td><td char=".">0.008</td></tr><tr><td>Migrant</td><td char=".">0.040</td><td char=".">0.020</td><td char=".">−0.020∼</td></tr><tr><td>ELA level 1</td><td char=".">0.101</td><td char=".">0.073</td><td char=".">−0.028∼</td></tr><tr><td>ELA level 2</td><td char=".">0.389</td><td char=".">0.394</td><td char=".">0.005</td></tr><tr><td>ELA level 3</td><td char=".">0.353</td><td char=".">0.358</td><td char=".">0.005</td></tr><tr><td>ELA level 4</td><td char=".">0.131</td><td char=".">0.148</td><td char=".">0.017</td></tr><tr><td>ELA level 5</td><td char=".">0.025</td><td char=".">0.027</td><td char=".">0.001</td></tr><tr><td>Math level 1</td><td char=".">0.115</td><td char=".">0.088</td><td char=".">−0.027</td></tr><tr><td>Math level 2</td><td char=".">0.228</td><td char=".">0.223</td><td char=".">−0.005</td></tr><tr><td>Math level 3</td><td char=".">0.430</td><td char=".">0.462</td><td char=".">0.033</td></tr><tr><td>Math level 4</td><td char=".">0.170</td><td char=".">0.179</td><td char=".">0.009</td></tr><tr><td>Math level 5</td><td char=".">0.057</td><td char=".">0.046</td><td char=".">−0.011</td></tr><tr><td>ELA scale score</td><td char=".">235.00</td><td char=".">236.28</td><td char=".">1.282</td></tr><tr><td>Math scale score</td><td char=".">243.85</td><td char=".">244.88</td><td char=".">1.031</td></tr><tr><td>Number of advanced courses taken</td><td char=".">1.68</td><td char=".">1.73</td><td char=".">0.042</td></tr><tr><td>Number of credits earned</td><td char=".">0.65</td><td char=".">0.72</td><td char=".">0.075</td></tr><tr><td>Number of disciplinary referrals</td><td char=".">0.72</td><td char=".">0.55</td><td char=".">−0.171∼</td></tr><tr><td>Number of suspensions</td><td char=".">0.22</td><td char=".">0.13</td><td char=".">−0.081*</td></tr><tr><td>Number of days absent</td><td char=".">4.40</td><td char=".">3.41</td><td char=".">−0.991**</td></tr></tbody></table> </ephtml> </p> <ulist> <item>5 <emph>Note</emph>: Significance levels were derived from unconditional OLS and logistic regression models.</item> <item>6 ∼<emph>p</emph> < 0.10; *<emph>p</emph> < 0.05; **<emph>p</emph> < 0.01.</item> </ulist> <p>As a sensitivity check, we conducted additional analyses using only students who attend Title I schools (Supplementary Appendix B, Table B2). The effect estimates of the AVID program are notably larger in this analysis particularly in the case of college enrollment. Students in the AVID group graduated from high school at a rate that was 4.3 percentage points higher than students in the non-AVID matched group. The difference between the compared groups for the college enrollment outcome was 7.5 percentage points.</p> <p>The fact that the estimates from this sensitivity check are larger than those from the core analyses suggests that they are positively biased by student selection factors. Students in AVID schools who are selected for participation in the elective have latent characteristics the program seeks (i.e., potential). Eligible students in the same schools who are not selected for participation may be less motivated academically nor interested in college. However, in the core analysis, AVID students are matched with students in the entire non-AVID population. This population is comprised of students who attend non-Title I schools and who may be more inclined to pursue college given their family background characteristics.</p> <hd id="AN0146971614-15">Variation in Effects by Duration of Participation</hd> <p>Table 3 presents differences in baseline characteristics between students who took one to two AVID electives and students who took three to four AVID electives. At baseline, students who persisted in AVID and took three to four electives had fewer disciplinary referrals, suspensions, and absences in 8th grade than students who took only one to two AVID electives. Moreover, fewer students who persisted in AVID were free or reduced-price lunch eligible and had a primary language other than English at baseline.</p> <p>We matched both AVID student groups with their own comparison groups comprised of students who did not take AVID electives. Tables 4 and 5 present the differences between the AVID student groups and the non-AVID comparison groups before and after matching. After the matching procedures, the samples are reduced to include only students with close matches on baseline student characteristics. Most AVID students were retained in the matched samples (i.e., 83% of the one to two elective groups and 80% of the three to four elective group).</p> <p>Table 4. Baseline group differences (means and proportions) between students who took one to two AVID electives and students who did not take an AVID elective before and after CE matching (FY15–FY17 cohorts).</p> <p> <ephtml> <table><thead><tr><td>Student characteristic</td><td>Pre-CE matching (<p id="ilm0013"><graphic href="uree_a_1772421_ilm0013.gif" content-type="Graph" /><math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub xmlns=""><mrow><mi>L</mi></mrow><mrow><mn>1</mn></mrow></msub></math></p> = 0.76)</td><td>Post-CE matching (<p id="ilm0014"><graphic href="uree_a_1772421_ilm0014.gif" content-type="Graph" /><math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub xmlns=""><mrow><mi>L</mi></mrow><mrow><mn>1</mn></mrow></msub></math></p> ∼ 0.00)</td></tr><tr><td>AVID (<italic>n</italic> = 877)</td><td>Non-AVID (<italic>n</italic> = 28,156)</td><td>Difference</td><td>AVID (<italic>n</italic> = 729)</td><td>Non-AVID (<italic>n</italic> = 8,657)</td><td>Difference</td></tr></thead><tbody valign="top"><tr><td>Female</td><td char=".">0.503</td><td char=".">0.498</td><td char=".">0.005</td><td char=".">0.512</td><td char=".">0.531</td><td char=".">−0.020</td></tr><tr><td>White</td><td char=".">0.107</td><td char=".">0.382</td><td char=".">−0.275***</td><td char=".">0.115</td><td char=".">0.115</td><td char=".">0.000</td></tr><tr><td>Black</td><td char=".">0.535</td><td char=".">0.271</td><td char=".">0.264***</td><td char=".">0.558</td><td char=".">0.558</td><td char=".">0.000</td></tr><tr><td>Hispanic</td><td char=".">0.314</td><td char=".">0.282</td><td char=".">0.032*</td><td char=".">0.300</td><td char=".">0.300</td><td char=".">0.000</td></tr><tr><td>Other race</td><td char=".">0.044</td><td char=".">0.065</td><td char=".">−0.021*</td><td char=".">0.026</td><td char=".">0.026</td><td char=".">0.000</td></tr><tr><td>ELL enrolled</td><td char=".">0.096</td><td char=".">0.055</td><td char=".">0.040***</td><td char=".">0.097</td><td char=".">0.096</td><td char=".">0.002</td></tr><tr><td>ELL exited</td><td char=".">0.303</td><td char=".">0.173</td><td char=".">0.130***</td><td char=".">0.285</td><td char=".">0.287</td><td char=".">−0.002</td></tr><tr><td>Non-ELL</td><td char=".">0.601</td><td char=".">0.771</td><td char=".">−0.170***</td><td char=".">0.617</td><td char=".">0.617</td><td char=".">0.000</td></tr><tr><td>Homeless</td><td char=".">0.011</td><td char=".">0.005</td><td char=".">0.006*</td><td char=".">0.001</td><td char=".">0.001</td><td char=".">0.000</td></tr><tr><td>Free/reduced lunch</td><td char=".">0.810</td><td char=".">0.517</td><td char=".">0.293***</td><td char=".">0.807</td><td char=".">0.807</td><td char=".">0.000</td></tr><tr><td>Primary lang. English</td><td char=".">0.531</td><td char=".">0.697</td><td char=".">−0.166***</td><td char=".">0.547</td><td char=".">0.568</td><td char=".">−0.021</td></tr><tr><td>Primary lang. Haitian-Creole</td><td char=".">0.217</td><td char=".">0.086</td><td char=".">0.131***</td><td char=".">0.224</td><td char=".">0.210</td><td char=".">0.014</td></tr><tr><td>Primary lang. Spanish</td><td char=".">0.226</td><td char=".">0.170</td><td char=".">0.056***</td><td char=".">0.219</td><td char=".">0.212</td><td char=".">0.007</td></tr><tr><td>Primary lang. other</td><td char=".">0.026</td><td char=".">0.048</td><td char=".">−0.021**</td><td char=".">0.010</td><td char=".">0.010</td><td char=".">0.000</td></tr><tr><td>Born in the U.S.</td><td char=".">0.796</td><td char=".">0.845</td><td char=".">−0.049***</td><td char=".">0.813</td><td char=".">0.801</td><td char=".">0.012</td></tr><tr><td>Born in Mexico</td><td char=".">0.011</td><td char=".">0.013</td><td char=".">−0.001</td><td char=".">0.012</td><td char=".">0.020</td><td char=".">−0.008</td></tr><tr><td>Born in Haiti</td><td char=".">0.111</td><td char=".">0.047</td><td char=".">0.063***</td><td char=".">0.114</td><td char=".">0.118</td><td char=".">−0.004</td></tr><tr><td>Born in other country</td><td char=".">0.082</td><td char=".">0.095</td><td char=".">−0.013</td><td char=".">0.060</td><td char=".">0.060</td><td char=".">0.000</td></tr><tr><td>Gifted</td><td char=".">0.007</td><td char=".">0.026</td><td char=".">−0.019**</td><td char=".">0.007</td><td char=".">0.009</td><td char=".">−0.002</td></tr><tr><td>Non-ESE</td><td char=".">0.910</td><td char=".">0.883</td><td char=".">0.027*</td><td char=".">0.945</td><td char=".">0.945</td><td char=".">0.000</td></tr><tr><td>Learn/behavioral disability</td><td char=".">0.072</td><td char=".">0.082</td><td char=".">−0.010</td><td char=".">0.048</td><td char=".">0.047</td><td char=".">0.001</td></tr><tr><td>Other disability</td><td char=".">0.018</td><td char=".">0.035</td><td char=".">−0.017**</td><td char=".">0.007</td><td char=".">0.007</td><td char=".">0.000</td></tr><tr><td>Migrant</td><td char=".">0.040</td><td char=".">0.016</td><td char=".">0.024***</td><td char=".">0.016</td><td char=".">0.016</td><td char=".">0.000</td></tr><tr><td>ELA level 1</td><td char=".">0.101</td><td char=".">0.121</td><td char=".">−0.019∼</td><td char=".">0.096</td><td char=".">0.096</td><td char=".">0.000</td></tr><tr><td>ELA level 2</td><td char=".">0.389</td><td char=".">0.254</td><td char=".">0.135***</td><td char=".">0.384</td><td char=".">0.384</td><td char=".">0.000</td></tr><tr><td>ELA level 3</td><td char=".">0.353</td><td char=".">0.303</td><td char=".">0.051**</td><td char=".">0.366</td><td char=".">0.366</td><td char=".">0.000</td></tr><tr><td>ELA level 4</td><td char=".">0.131</td><td char=".">0.212</td><td char=".">−0.081***</td><td char=".">0.132</td><td char=".">0.132</td><td char=".">0.000</td></tr><tr><td>ELA level 5</td><td char=".">0.025</td><td char=".">0.110</td><td char=".">−0.085***</td><td char=".">0.022</td><td char=".">0.022</td><td char=".">0.000</td></tr><tr><td>Math level 1</td><td char=".">0.115</td><td char=".">0.110</td><td char=".">0.005</td><td char=".">0.114</td><td char=".">0.114</td><td char=".">0.000</td></tr><tr><td>Math level 2</td><td char=".">0.228</td><td char=".">0.158</td><td char=".">0.070***</td><td char=".">0.225</td><td char=".">0.225</td><td char=".">0.000</td></tr><tr><td>Math level 3</td><td char=".">0.430</td><td char=".">0.343</td><td char=".">0.087***</td><td char=".">0.450</td><td char=".">0.450</td><td char=".">0.000</td></tr><tr><td>Math level 4</td><td char=".">0.170</td><td char=".">0.226</td><td char=".">−0.056***</td><td char=".">0.166</td><td char=".">0.166</td><td char=".">0.000</td></tr><tr><td>Math level 5</td><td char=".">0.057</td><td char=".">0.163</td><td char=".">−0.106***</td><td char=".">0.045</td><td char=".">0.045</td><td char=".">0.000</td></tr><tr><td>ELA scale score</td><td char=".">235.00</td><td char=".">240.75</td><td char=".">−5.75***</td><td char=".">235.19</td><td char=".">234.60</td><td char=".">0.60</td></tr><tr><td>Math scale score</td><td char=".">243.85</td><td char=".">248.92</td><td char=".">−5.07***</td><td char=".">243.55</td><td char=".">243.46</td><td char=".">0.08</td></tr><tr><td>Advanced courses taken</td><td char=".">1.68</td><td char=".">2.11</td><td char=".">−0.43***</td><td char=".">1.69</td><td char=".">1.66</td><td char=".">0.03</td></tr><tr><td>Credits earned</td><td char=".">0.65</td><td char=".">0.90</td><td char=".">−0.26***</td><td char=".">0.65</td><td char=".">0.61</td><td char=".">0.04</td></tr><tr><td>Disciplinary referrals</td><td char=".">0.72</td><td char=".">0.64</td><td char=".">0.08</td><td char=".">0.67</td><td char=".">0.73</td><td char=".">−0.06</td></tr><tr><td>Out of school suspensions</td><td char=".">0.22</td><td char=".">0.18</td><td char=".">0.03</td><td char=".">0.20</td><td char=".">0.19</td><td char=".">0.01</td></tr><tr><td>Days absent</td><td char=".">4.40</td><td char=".">5.34</td><td char=".">−0.94***</td><td char=".">4.44</td><td char=".">4.57</td><td char=".">−0.13</td></tr></tbody></table> </ephtml> </p> <ulist> <item>7 <emph>Note</emph>: Significance levels were derived from unconditional OLS and logistic regression models.</item> <item>8 ∼<emph>p</emph> < 0.10; *<emph>p</emph> < 0.05; **<emph>p</emph> < 0.01; ***<emph>p</emph> < 0.001.</item> </ulist> <p>Table 5. Baseline group differences (means and proportions) between students who took three to four AVID electives and students who did not take an AVID elective before and after CE matching (FY15–FY17 cohorts).</p> <p> <ephtml> <table><thead><tr><td>Student characteristic</td><td>Pre-CE matching (<p id="ilm0015"><graphic href="uree_a_1772421_ilm0015.gif" content-type="Graph" /><math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub xmlns=""><mrow><mi>L</mi></mrow><mrow><mn>1</mn></mrow></msub></math></p> = 0.91)</td><td>Post-CE matching (<p id="ilm0016"><graphic href="uree_a_1772421_ilm0016.gif" content-type="Graph" /><math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub xmlns=""><mrow><mi>L</mi></mrow><mrow><mn>1</mn></mrow></msub></math></p> ∼ 0.00)</td></tr><tr><td>AVID (<italic>n</italic> = 452)</td><td>Non-AVID (<italic>n</italic> = 28,156)</td><td>Difference</td><td>AVID (<italic>n</italic> = 360)</td><td>Non-AVID (<italic>n</italic> = 2,687)</td><td>Difference</td></tr></thead><tbody valign="top"><tr><td>Female</td><td char=".">0.542</td><td char=".">0.498</td><td char=".">0.044∼</td><td char=".">0.542</td><td char=".">0.542</td><td char=".">0.000</td></tr><tr><td>White</td><td char=".">0.119</td><td char=".">0.382</td><td char=".">−0.262***</td><td char=".">0.119</td><td char=".">0.119</td><td char=".">0.000</td></tr><tr><td>Black</td><td char=".">0.533</td><td char=".">0.271</td><td char=".">0.262***</td><td char=".">0.550</td><td char=".">0.550</td><td char=".">0.000</td></tr><tr><td>Hispanic</td><td char=".">0.272</td><td char=".">0.282</td><td char=".">−0.009</td><td char=".">0.275</td><td char=".">0.275</td><td char=".">0.000</td></tr><tr><td>Other race</td><td char=".">0.075</td><td char=".">0.065</td><td char=".">0.010</td><td char=".">0.056</td><td char=".">0.056</td><td char=".">0.000</td></tr><tr><td>ELL enrolled</td><td char=".">0.086</td><td char=".">0.055</td><td char=".">0.031**</td><td char=".">0.067</td><td char=".">0.067</td><td char=".">0.000</td></tr><tr><td>ELL exited</td><td char=".">0.248</td><td char=".">0.173</td><td char=".">0.074***</td><td char=".">0.208</td><td char=".">0.208</td><td char=".">0.000</td></tr><tr><td>Non-ELL</td><td char=".">0.666</td><td char=".">0.771</td><td char=".">−0.105***</td><td char=".">0.725</td><td char=".">0.725</td><td char=".">0.000</td></tr><tr><td>Homeless</td><td char=".">0.009</td><td char=".">0.005</td><td char=".">0.004</td><td char=".">0.011</td><td char=".">0.006</td><td char=".">0.005</td></tr><tr><td>Free/reduced lunch</td><td char=".">0.759</td><td char=".">0.517</td><td char=".">0.242***</td><td char=".">0.789</td><td char=".">0.789</td><td char=".">0.000</td></tr><tr><td>Primary lang. English</td><td char=".">0.593</td><td char=".">0.697</td><td char=".">−0.104***</td><td char=".">0.642</td><td char=".">0.662</td><td char=".">−0.021</td></tr><tr><td>Primary lang. Haitian-Creole</td><td char=".">0.188</td><td char=".">0.086</td><td char=".">0.102***</td><td char=".">0.167</td><td char=".">0.146</td><td char=".">0.020</td></tr><tr><td>Primary lang. Spanish</td><td char=".">0.195</td><td char=".">0.170</td><td char=".">0.025</td><td char=".">0.183</td><td char=".">0.183</td><td char=".">0.001</td></tr><tr><td>Primary lang. other</td><td char=".">0.024</td><td char=".">0.048</td><td char=".">−0.023*</td><td char=".">0.008</td><td char=".">0.008</td><td char=".">0.000</td></tr><tr><td>Born in the U.S.</td><td char=".">0.830</td><td char=".">0.845</td><td char=".">−0.015</td><td char=".">0.864</td><td char=".">0.864</td><td char=".">0.000</td></tr><tr><td>Born in Mexico</td><td char=".">0.007</td><td char=".">0.013</td><td char=".">−0.006</td><td char=".">0.006</td><td char=".">0.017</td><td char=".">−0.012</td></tr><tr><td>Born in Haiti</td><td char=".">0.093</td><td char=".">0.047</td><td char=".">0.046***</td><td char=".">0.081</td><td char=".">0.074</td><td char=".">0.006</td></tr><tr><td>Born in other country</td><td char=".">0.071</td><td char=".">0.095</td><td char=".">−0.024∼</td><td char=".">0.050</td><td char=".">0.045</td><td char=".">0.005</td></tr><tr><td>Gifted</td><td char=".">0.009</td><td char=".">0.026</td><td char=".">−0.017*</td><td char=".">0.000</td><td char=".">0.000</td><td char=".">0.000</td></tr><tr><td>Non-ESE</td><td char=".">0.881</td><td char=".">0.883</td><td char=".">−0.003</td><td char=".">0.919</td><td char=".">0.919</td><td char=".">0.000</td></tr><tr><td>Learn/behavioral disability</td><td char=".">0.093</td><td char=".">0.082</td><td char=".">0.011</td><td char=".">0.058</td><td char=".">0.049</td><td char=".">0.010</td></tr><tr><td>Other disability</td><td char=".">0.027</td><td char=".">0.035</td><td char=".">−0.008</td><td char=".">0.022</td><td char=".">0.032</td><td char=".">−0.010</td></tr><tr><td>Migrant</td><td char=".">0.020</td><td char=".">0.016</td><td char=".">0.004</td><td char=".">0.025</td><td char=".">0.023</td><td char=".">0.002</td></tr><tr><td>ELA level 1</td><td char=".">0.073</td><td char=".">0.121</td><td char=".">−0.048**</td><td char=".">0.064</td><td char=".">0.064</td><td char=".">0.000</td></tr><tr><td>ELA level 2</td><td char=".">0.394</td><td char=".">0.254</td><td char=".">0.140***</td><td char=".">0.394</td><td char=".">0.394</td><td char=".">0.000</td></tr><tr><td>ELA level 3</td><td char=".">0.358</td><td char=".">0.303</td><td char=".">0.056*</td><td char=".">0.375</td><td char=".">0.375</td><td char=".">0.000</td></tr><tr><td>ELA level 4</td><td char=".">0.148</td><td char=".">0.212</td><td char=".">−0.064**</td><td char=".">0.144</td><td char=".">0.144</td><td char=".">0.000</td></tr><tr><td>ELA level 5</td><td char=".">0.027</td><td char=".">0.110</td><td char=".">−0.084***</td><td char=".">0.022</td><td char=".">0.022</td><td char=".">0.000</td></tr><tr><td>Math level 1</td><td char=".">0.088</td><td char=".">0.110</td><td char=".">−0.022</td><td char=".">0.083</td><td char=".">0.083</td><td char=".">0.000</td></tr><tr><td>Math level 2</td><td char=".">0.223</td><td char=".">0.158</td><td char=".">0.065***</td><td char=".">0.225</td><td char=".">0.225</td><td char=".">0.000</td></tr><tr><td>Math level 3</td><td char=".">0.462</td><td char=".">0.343</td><td char=".">0.120***</td><td char=".">0.489</td><td char=".">0.489</td><td char=".">0.000</td></tr><tr><td>Math level 4</td><td char=".">0.179</td><td char=".">0.226</td><td char=".">−0.047*</td><td char=".">0.161</td><td char=".">0.161</td><td char=".">0.000</td></tr><tr><td>Math level 5</td><td char=".">0.046</td><td char=".">0.163</td><td char=".">−0.117***</td><td char=".">0.042</td><td char=".">0.042</td><td char=".">0.000</td></tr><tr><td>ELA scale score</td><td char=".">236.28</td><td char=".">240.75</td><td char=".">−4.47***</td><td char=".">236.18</td><td char=".">236.20</td><td char=".">−0.02</td></tr><tr><td>Math scale score</td><td char=".">244.88</td><td char=".">248.92</td><td char=".">−4.04***</td><td char=".">244.56</td><td char=".">244.48</td><td char=".">0.09</td></tr><tr><td>Advanced courses taken</td><td char=".">1.73</td><td char=".">2.11</td><td char=".">−0.39**</td><td char=".">1.66</td><td char=".">1.74</td><td char=".">−0.08</td></tr><tr><td>Credits earned</td><td char=".">0.72</td><td char=".">0.90</td><td char=".">−0.18**</td><td char=".">0.71</td><td char=".">0.68</td><td char=".">0.03</td></tr><tr><td>Disciplinary referrals</td><td char=".">0.55</td><td char=".">0.64</td><td char=".">−0.09</td><td char=".">0.63</td><td char=".">0.66</td><td char=".">−0.03</td></tr><tr><td>Out of school suspensions</td><td char=".">0.13</td><td char=".">0.18</td><td char=".">−0.05</td><td char=".">0.14</td><td char=".">0.19</td><td char=".">−0.05</td></tr><tr><td>Days absent</td><td char=".">3.41</td><td char=".">5.34</td><td char=".">−1.93***</td><td char=".">3.43</td><td char=".">3.57</td><td char=".">−0.14</td></tr></tbody></table> </ephtml> </p> <ulist> <item>9 <emph>Note</emph>: Significance levels were derived from unconditional OLS and logistic regression models.</item> <item>10 ∼<emph>p</emph> < 0.10; *<emph>p</emph> < 0.05; **<emph>p</emph> < 0.01; ***<emph>p</emph> < 0.001.</item> </ulist> <p>Table 6 presents differences in effects by the number of AVID electives taken. To be clear, students who took one to two AVID electives were not matched with students who took three to four electives. All matches were made to non-AVID students to estimate program effects. The graduation rate of the one to two AVID elective groups (92.7%) is significantly higher than the graduation rate of their matched non-AVID group (89.1%). The college enrollment rate of the one to two AVID elective groups (56.7%) is not significantly different from the enrollment rate of their matched non-AVID group (56.4%).</p> <p>Table 6. Differences in post-secondary outcomes (post CE matching) for students who took AVID electives vs. students who did not take AVID electives by the number of electives taken (FY15–FY17 cohorts).</p> <p> <ephtml> <table><thead><tr><td>Outcome</td><td>1–2 AVID electives taken</td><td>3–4 AVID electives taken</td></tr><tr><td>AVID (<italic>n</italic> = 729)</td><td>Non-AVID (<italic>n</italic> = 8,657)</td><td>Difference</td><td>AVID (<italic>n</italic> = 360)</td><td>Non-AVID (<italic>n</italic> = 2,687)</td><td>Difference</td></tr></thead><tbody valign="top"><tr><td>Graduated HS on time</td><td char=".">0.927</td><td char=".">0.891</td><td char=".">0.037**</td><td char=".">0.961</td><td char=".">0.921</td><td char=".">0.040**</td></tr><tr><td>Enrolled in college</td><td char=".">0.567</td><td char=".">0.555</td><td char=".">0.011</td><td char=".">0.686</td><td char=".">0.587</td><td char=".">0.099***</td></tr></tbody></table> </ephtml> </p> <ulist> <item>11 <emph>Note</emph>: Significance levels were derived from unconditional logistic regression models.</item> <item>12 **<emph>p</emph> < 0.01; ***<emph>p</emph> < 0.001.</item> </ulist> <p>The graduation rate of the three to four AVID elective group (96.1%) is significantly higher than the graduation rate of their matched non-AVID group (92.1%). The college enrollment rate of the three to four AVID elective group (68.6%) is also significantly and notably higher than the enrollment rate of their matched non-AVID group (58.7%).</p> <hd id="AN0146971614-16">Program Cost per Pupil per Year</hd> <p>In estimating the annual cost of AVID program implementation, we identified three nested cost categories: District overhead, per school costs, and elective costs.</p> <p>District overhead includes the salaries of the personnel who manage the implementation of the AVID program and provide training and support to school staff (Supplementary Appendix C, Table C1). Four administrative personnel were identified as AVID related District-level staff. Two of the AVID administrators worked full-time on AVID programming and training and their entire annual salary and benefits packages were counted in the cost of AVID. The remaining two AVID administrators played support roles for multiple programs and dedicated approximately 5% of their work to AVID programming. Therefore, only 5% of their annual salary and benefits packages were counted in the cost of AVID. Estimates of the average salary and benefits packages for administrators were derived from interviews with the District's Human Recourses Department.</p> <p>The per school costs include AVID Summer Institute training for school staff, mobile whiteboards used in AVID elective classrooms, and the AVID Library Package and annual membership fees (Supplememtary Appendix C, Table C2). Average school costs were derived from budgetary documents housed by the District's School Board.</p> <p>The per elective costs include the salaries of certified teachers who teach the AVID electives, the hourly wages of college-level tutors hired to facilitate small group instruction during the AVID tutorial (which occurs twice a week), general classroom supplies such as paper, and the use of classroom facilities (Supplementary Appendix C, Table C3). Only 20% of AVID teacher salaries were included in the per elective costs because the average teacher course load is five courses per day. The hourly tutor wage of fifteen dollars is set by the District, and the cost of facilities was estimated by using the hourly rental rate for a classroom as determined by document review.</p> <p>We calculated the annual cost to implement the AVID program in the District per student by first multiplying (a) the per school costs with the number of schools implementing AVID and (b) the per elective costs with the total number of electives offered in the District. The total school costs, total elective costs, and the District overhead costs were then summed and divided by the total number of students who took the AVID elective. These per-pupil annual costs were estimated separately for each year under the study period, and an average was calculated across years (Supplementary Appendix C, Table C4).</p> <p>In sum, we estimated the average annual cost to implement the AVID program in the District to be $1,230 per pupil.</p> <hd id="AN0146971614-17">Program Cost per Student Positively Impacted by AVID</hd> <p>In total, we estimated that the AVID program graduated an additional 41 students from high school and enrolled an additional 44 students in college from 2015 to 2017 (Table 7). The counts of students impacted are calculated by multiplying the proportional difference in outcomes between the matched groups by the size of the AVID samples. These estimates were derived from the data in Table 6.</p> <p>Table 7. Number of additional students who had a positive outcome that can be attributed to AVID participation (FY15–FY17 cohorts).</p> <p> <ephtml> <table><thead><tr><td>Outcome</td><td>One or two AVID electives</td><td>three or four AVID electives</td><td>Total no. of students impacted by AVID</td></tr></thead><tbody valign="top"><tr><td>Graduated HS on time</td><td char=".">27</td><td char=".">14</td><td char=".">41</td></tr><tr><td>Enrolled in college</td><td char=".">8</td><td char=".">36</td><td char=".">44</td></tr></tbody></table> </ephtml> </p> <p>13 <emph>Note</emph>: The counts of students impacted are derived from the estimates in Table 6.</p> <p>Short-term involvement with AVID (i.e., one or two electives are taken) produced the majority of the new high school graduates (65.9%). Long-term involvement with AVID (i.e., three or four electives taken) produced the majority of the new college enrollees (81.8%).</p> <p>To calculate the cost per additional positive outcome, the cost per student is multiplied by the number of students in the served sample and divided by the number of students who were positively impacted as measured by the program evaluation. Therefore, for students who take one or two AVID electives, the cost for each additional high school graduate is:</p> <p></p> <ulist> <item> $49,815 = (729 students served) ($1,230 per year) (1.5 years)/27 additional high school graduates</item> </ulist> <p>For students who take three or four AVID electives, the cost for each additional high school graduate is:</p> <p></p> <ulist> <item> $110,700 = (360 students) ($1,230 per year) (3.5 years)/14 additional high school graduates</item> </ulist> <p>For students who take three or four AVID electives, the cost for each additional college enrollee is:</p> <p></p> <ulist> <item> $43,050 = (360 students) ($1,230 per year) (3.5 years)/36 additional college enrollees</item> </ulist> <hd id="AN0146971614-18">Discussion</hd> <p>This study examined the effects of AVID program participation on high-school graduation and college enrollment and estimated the cost of program implementation. We used the CEM data pre-processing algorithm to generate comparison groups for the evaluation. We applied the ingredients method to estimate program implementation costs and calculated CE ratios by the duration of participation.</p> <p>Findings from the evaluation indicate that AVID students graduate from high school and enroll in college at higher rates than non-AVID students in the matched comparison group. The relationship of AVID elective course-taking with high school graduation is similar regardless of how long students are involved with AVID (i.e., a 3.7 percentage point difference when students take one to two AVID electives and a 4.0 percentage point difference when students take three to four AVID electives). However, only long-term participation in AVID is related to college enrollment (i.e., a 9.9 percentage point difference when students take three to four AVID electives).</p> <p>The cost analysis revealed that, on average, the cost to implement the AVID program in the District is $1,230 per student per year. The cost per additional high school graduate is $49,815 if a student took one to two AVID electives and $110,700 if a student took three to four AVID electives. The cost per additional college enrollee is $43,050 and this effect was only evident for students who took three to four AVID electives.</p> <p>Compared to other programs designed to improve post-secondary outcomes for low income and ethnic minority students, the AVID program is moderately cost-effective. Kolbe et al. ([<reflink idref="bib16" id="ref45">16</reflink>]) found that the cost for each additional high school graduate produced by the AVID/TOPS program is $271,490, and the cost for each additional college enrollee is $79,600. Bowden and Belfield ([<reflink idref="bib3" id="ref46">3</reflink>]) found that the cost for each additional high school graduate produced by the Talent Search program is $43,440, and the cost for each additional college enrollee is $33,380. Therefore, the Talent Search program is more cost-effective than the standalone AVID program and the AVID/TOPS program, and the standalone AVID program is more cost-effective than the AVID/TOPS program.</p> <p>The per-pupil cost to implement the Talent Search program ($3,290) is smaller than the cost to implement the AVID program ($4,310 for students who take three to four electives). Therefore, the small CE ratios for the Talent Search program are due to large effect sizes (8.7 percentage points for high school graduation and 11.3 percentage points for college enrollment) and relatively low costs.</p> <p>The effect estimates of the AVID/TOPS program (6.3 percentage points for high school graduation and 21.3 percentage points for college enrollment) are larger in magnitude than the effects of the AVID standalone program. However, the per-pupil cost to implement the AVID/TOPS program ($16,990) is much larger than the cost to implement the standalone AVID program. Therefore, the large CE ratios for the AVID/TOPS program are the result of high program implementation costs.</p> <hd id="AN0146971614-19">Limitations</hd> <p>For this study, we made a strong attempt to balance methodological rigor with the time constraints of applied school district research. Although this applied study builds on past evaluations of AVID, continued academic research is needed to estimate the true causal effects and costs of the program.</p> <hd id="AN0146971614-20">The Program Evaluation</hd> <p>We used the CEM approach to match AVID and non-AVID students on 40 baseline student characteristics prior to examining differences in outcomes. However, there are unobserved selection factors that were omitted from the matching process. For example, to enroll in the AVID program, the parents and guardians of the participating students must meet with school staff and sign an AVID contract to affirm their commitment to the goals of the program. If a parent or guardian is not available for this level of participation, the student is not eligible to take the AVID elective. Therefore, AVID students may have higher levels of parental involvement than non-AVID students.</p> <p>Moreover, students selected to participate in AVID tend to demonstrate college-going aspirations and maybe more academically motivated than otherwise similar students who are not selected to participate. The motivational qualities of AVID participants may have been partially captured by the covariates included in the matching algorithm (i.e., attendance, discipline, and achievement). However, these covariates are merely proxies for important motivational factors (Green et al., [<reflink idref="bib9" id="ref47">9</reflink>]; Hirschfield & Gasper, [<reflink idref="bib11" id="ref48">11</reflink>]; Railsback, [<reflink idref="bib24" id="ref49">24</reflink>]).</p> <p>It is certainly plausible that motivated students with high levels of parental involvement are more likely to graduate from high school and enroll in college than similar students with low levels of motivation and parental involvement, regardless of whether they take an AVID elective. Therefore, the estimates of program effectiveness reported here may be positively biased and overstated due to these omissions.</p> <p>An additional limitation of the evaluation is that we did not examine variation in effects by the school site. It is unknown whether program effects are homogeneous across school sites or if the effects of AVID vary as a function of school characteristics. Moreover, incorporating AVID into the school curriculum might influence school culture. The program may have spillover effects that impact students who attend AVID schools and do not take AVID electives. If this is the case, the effects of the AVID program that we report may be underestimated.</p> <hd id="AN0146971614-21">The Cost Analysis</hd> <p>Cost analyses are often absent in education evaluation research (Levin & Belfield, [<reflink idref="bib18" id="ref50">18</reflink>]). For this study, we applied the highly recommended ingredients method to estimate program costs. However, we relied primarily on budgetary data to estimate the costs of the training, membership fees, and materials that were directly related to AVID implementation. The use of budgetary data can be misleading because it only captures expenditures and does not include opportunity costs. Additionally, we used local prices for ingredients. The use of local prices makes it difficult to generalize findings to other regions of the U.S. where AVID is implemented (Levin et al., [<reflink idref="bib19" id="ref51">19</reflink>]). And finally, we estimated program costs retrospectively. Data from the most recent year of implementation were used to estimate the cost of implementation spanning a six-year historical period. Therefore, changes in implementation over time may be related to changes in costs that are not captured in this cost analysis.</p> <p>Due to these limitations, the preliminary cost estimates reported here cannot be directly compared to the cost estimates of similar interventions in the published literature (see Hollands et al., [<reflink idref="bib12" id="ref52">12</reflink>] for example). We hope that there is enough information in the Supplemental Materials, Supplementary Appendix C, Tables C1–C4 for future researchers to expand upon this cost analysis and extrapolate more complete and global cost estimates.</p> <p>However, even if it were possible to estimate the true costs of program implementation, CE analysis is not the ideal approach for comparing programs that leverage several mechanisms and impact multiple outcomes (Levin et al., [<reflink idref="bib19" id="ref53">19</reflink>]). In this study, separate CE ratios were estimated for each outcome. However, a given student may have experienced both positive outcomes as a result of AVID participation. Indeed, students may have experienced additional positive outcomes such as increased behavioral engagement. One way to address this issue is to conduct a cost-benefit analysis that would include any long-term benefit of exposure to AVID such as reductions in crime and incarceration and higher future earnings. However, a cost-benefit analysis was beyond the scope of this study.</p> <hd id="AN0146971614-22">Implications for Practice</hd> <p>The finding that the relationship of AVID elective course-taking with high school graduation is similar regardless of how long students are involved with AVID has implications for practice. How this information is used depends on what school districts aim to accomplish with the AVID program. For example, if the goal is to increase high school graduation rates, expanding opportunities for short-term involvement with AVID may be the most cost-effective approach. If the goal is to increase college enrollment, long-term involvement in AVID may be necessary to impact this outcome. If the latter is the case, additional research is needed to identify the factors that contribute to AVID program dropout so that program implementers can provide targeted support to students who are at risk of leaving the program.</p> <p>Another way to cost-effectively increase participation in AVID may be to improve the student selection processes. In the School District of Palm Beach County, students deemed to have the potential for academic success are referred to the AVID program by principals and teachers. However, college outreach programs may be "superfluous for the educationally oriented students who voluntarily participate" (Domina, [<reflink idref="bib6" id="ref54">6</reflink>], p. 147). To improve this selection process, researchers in the District are developing predictive models to calculate students' probabilities of high school graduation and college enrollment. The resulting predictions can help teachers and administrators identify possibly overlooked students who would benefit from involvement with AVID.</p> <hd id="AN0146971614-23">Implications for Research</hd> <p>Districts aiming to track the effectiveness of the AVID program—and other programs like it—can improve on this research design by flagging all students who initially self-select to participate in the program regardless of whether they actually participate. This way, it would be possible to limit the comparison sample to students who not only meet the criteria for inclusion but who also take initiative to participate. When resources are limited, as they often are in public education, not all students who are eligible and wish to participate in a program are allowed admittance. If the students who are not allowed admittance are tracked (e.g., put on a waiting list) they can be matched with students in the program group for an impact evaluation. Such an approach can inherently adjust for unobserved selection factors and improve causal estimates of program effects (Murnane & Willett, [<reflink idref="bib21" id="ref55">21</reflink>]).</p> <p>More work is also needed to understand the mechanisms that drive the association of AVID elective course taking and post-secondary success. For example, future research could examine the extent to which instructional rigor mediates the association of AVID course-taking and college enrollment to determine whether rigor is the primary mechanism that drives the relationship. Future research could also examine whether program effects vary as a function of student demographics, the grade level the elective was taken, and by school site. A better understanding of which mechanisms are the most influential, under what conditions, and for whom would allow practitioners to streamline the program and potentially shrink the CE ratios presented here.</p> <hd id="AN0146971614-24">Conclusions</hd> <p>Findings from this research suggest that the AVID program, as implemented in one large urban school district, is effective at increasing high school graduation and college enrollment rates. Students who take one to two AVID electives experience the same benefits as students who take three to four AVID electives in terms of high school graduation (i.e., a 3.7 percentage point difference versus a 4.0 percentage point difference, respectively). However, students who take three to four AVID electives experience far greater benefits than students who take one to two AVID electives in terms of college enrollment (i.e., a 9.9 percentage point difference versus a 1.1 percentage point difference, respectively).</p> <p>The cost analysis revealed that, on average, the cost to implement the AVID program in the District is $1,230 per student per year. This finding suggests that the AVID program is more expensive to implement than the Talent Search program and less expensive to implement than the AVID-TOPS program. These findings contribute to a nascent but growing body of needed evidence regarding the cost-effectiveness of alternative programs designed to improve post-secondary outcomes for disadvantaged students. Although limited, this information can help decision-makers select the programs that will produce the greatest return on investment.</p> <ref id="AN0146971614-25"> <title> Footnotes </title> <blist> <bibl id="bib1" idref="ref15" type="bt">1</bibl> <bibtext> Supplemental data for this article can be accessed online at <ulink href="http://dx.doi.org/10.1080/19345747.2020.1772421">http://dx.doi.org/10.1080/19345747.2020.1772421</ulink></bibtext> </blist> </ref> <ref id="AN0146971614-26"> <title> References </title> <blist> <bibtext> Bendall, P., Bollhoefer, A., & Koilpillai, V. (2015). AVID critical thinking and engagement: A schoolwide approach. AVID Press.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref18" type="bt">2</bibl> <bibtext> Black, A. C., Little, C. A., McCoach, D. B., Purcell, J. H., & Siegle, D. (2008). Advancement Via Individual Determination: Method selection in conclusions about program effectiveness. The Journal of Educational Research, 102 (2), 111 – 124. https://doi.org/10.3200/JOER.102.2.111-124</bibtext> </blist> <blist> <bibl id="bib3" idref="ref8" type="bt">3</bibl> <bibtext> Bowden, A. B., & Belfield, C. (2015). Evaluating the talent search TRIO program: A benefit-cost analysis and cost-effectiveness analysis. Journal of Benefit-Cost Analysis, 6 (3), 572 – 602. https://doi.org/ 10.1017/bca.2015.48</bibtext> </blist> <blist> <bibl id="bib4" idref="ref22" type="bt">4</bibl> <bibtext> Connors, L. (2010). Achievement for Advancement Via Individual Determination (avid) students and non-avid students in select central Florida high schools in 2007–2009: A comparative study [Doctoral dissertation, University of Central Florida]. University of Central Florida Online Repository. <ulink href="http://stars.library.ucf.edu/etd/4319">http://stars.library.ucf.edu/etd/4319</ulink></bibtext> </blist> <blist> <bibl id="bib5" idref="ref23" type="bt">5</bibl> <bibtext> Dean, J. C. (2009). The relationship between Advancement Via Individual Determination (AVID) and middle school student academic achievement and stakeholders' perceptions: A southern California case study [Unpublished doctoral dissertation]. Walden University.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref10" type="bt">6</bibl> <bibtext> Domina, T. (2009). What works in college outreach: Assessing targeted and schoolwide interventions for disadvantaged students. Educational Evaluation and Policy Analysis, 31 (2), 127 – 152. https://doi.org/10.3102/0162373709333887</bibtext> </blist> <blist> <bibl id="bib7" idref="ref6" type="bt">7</bibl> <bibtext> Dumais, S. A., & Ward, A. (2010). Cultural capital and first-generation college success. Poetics, 38 (3), 245 – 265. https://doi.org/10.1016/j.poetic.2009.11.011</bibtext> </blist> <blist> <bibl id="bib8" idref="ref17" type="bt">8</bibl> <bibtext> Franklin, S. L. G. (2011). An analysis of the relationship of the AVID program on the educational outcomes attributed to college ready graduates [Unpublished doctoral dissertation]. Tarleton State University.</bibtext> </blist> <blist> <bibl id="bib9" idref="ref47" type="bt">9</bibl> <bibtext> Green, J., Liem, G. A. D., Martin, A. J., Colmar, S., Marsh, H. W., & McInerney, D. (2012). Academic motivation, self-concept, engagement, and performance in high school: Key processes from a longitudinal perspective. Journal of Adolescence, 35 (5), 1111 – 1122. https://doi.org/10.1016/j.adolescence.2012.02.016</bibtext> </blist> <blist> <bibtext> Handwerk, P., Tognatta, N., Coley, R. J., & Gitomer, D. H. (2008). Access to success: Patterns of advanced placement participation in US high schools (Policy Information Report). Educational Testing Service.</bibtext> </blist> <blist> <bibtext> Hirschfield, P. J., & Gasper, J. (2011). The relationship between school engagement and delinquency in late childhood and early adolescence. Journal of Youth and Adolescence, 40 (1), 3 – 22. https://doi.org/10.1007/s10964-010-9579-5</bibtext> </blist> <blist> <bibtext> Hollands, F., Bowden, A. B., Belfield, C., Levin, H. M., Cheng, H., Shand, R., Pan, Y., & Hanisch-Cerda, B. (2014). Cost-effectiveness analysis in practice: Interventions to improve high school completion. Educational Evaluation and Policy Analysis, 36 (3), 307 – 326. https://doi.org/10.3102/0162373713511850</bibtext> </blist> <blist> <bibtext> Huerta, J. J., Watt, K. M., & Butcher, J. T. (2013). Examining Advancement Via Individual Determination (AVID) and its impact on middle school rigor and student preparedness. American Secondary Education, 2, 41.</bibtext> </blist> <blist> <bibtext> Iacus, S. M., King, G., & Porro, G. (2012). Causal inference without balance checking: Coarsened exact matching. Political Analysis, 20 (1), 1 – 24. https://doi.org/10.1093/pan/mpr013</bibtext> </blist> <blist> <bibtext> King, G., Nielsen, R., Coberley, C., Pope, J. E., Wells, A. (2011). Comparative effectiveness of matching methods for causal inference. Harvard University. https://gking.harvard.edu/files/psparadox.pdf</bibtext> </blist> <blist> <bibtext> Kolbe, T., Kinsley, P., Feldman, R. C., & Goldrick-Rab, S. (2018). From the (academic) middle to the top: An evaluation of the AVID/TOPS college access program. Journal of Education for Students Placed at Risk, 23 (4), 304 – 335. https://doi.org/10.1080/10824669.2018.1530114</bibtext> </blist> <blist> <bibtext> Le, V. N., Mariano, L. T., & Faxon-Mills, S. (2016). Can college outreach programs improve college readiness? The case of the college bound, St. Louis program. Research in Higher Education, 57 (3), 261 – 287. https://doi.org/10.1007/s11162-015-9385-8</bibtext> </blist> <blist> <bibtext> Levin, H. M., & Belfield, C. (2015). Guiding the development and use of cost-effectiveness analysis in education. Journal of Research on Educational Effectiveness, 8 (3), 400 – 419. https://doi.org/10.1080/19345747.2014.915604</bibtext> </blist> <blist> <bibtext> Levin, H. M., McEwan, P. J., Belfield, C., Bowden, A. B., & Shand, R. (2017). Economic evaluation in education: Cost-effectiveness and benefit-cost analysis. SAGE.</bibtext> </blist> <blist> <bibtext> Mehan, H., Hubbard, L., Villanueva, I., Lintz, A., & Okamoto, D. (1996). Constructing school success: The consequences of untracking low achieving students. Cambridge University Press.</bibtext> </blist> <blist> <bibtext> Murnane, R. J., & Willett, J. B. (2010). Methods matter: Improving causal inference in educational and social science research. Oxford University Press.</bibtext> </blist> <blist> <bibtext> Myers, D., Olsen, R., Seftor, N., Young, J., Tuttle, C. (2004). The impacts of regular upward bound: Results from the third follow-up data Collection (MPR Reference No. 8464-600). Mathematica Policy Research, Inc.</bibtext> </blist> <blist> <bibtext> Peak, D. J. (2010). A correlational study of the Advancement Via Individual Determination (AVID) program with middle school student achievement in mathematics [Unpublished doctoral dissertation]. University of Denver.</bibtext> </blist> <blist> <bibtext> Railsback, J. (2004). Increasing student attendance: Strategies from research and practice. Northwest Regional Educational Laboratory. https://files.eric.ed.gov/fulltext/ED484551.pdf</bibtext> </blist> <blist> <bibtext> Smith, J. B., Elder, E. C., & Stevens, K. (2014). Evaluation of a college readiness program: Advancement Via Individual Determination (AVID). Review of Higher Education & Self-Learning, 7 (25), 23 – 60.</bibtext> </blist> <blist> <bibtext> Stuart, E. A. (2010). Matching methods for causal inference: A review and a look forward. Statistical Science: A Review Journal of the Institute of Mathematical Statistics, 25 (1), 1 – 21. https://doi.org/10.1214/09-STS313</bibtext> </blist> <blist> <bibtext> U.S. Department of Education. (2007). Federal TRIO programs-recent funding. Office of Post-Secondary Education. https://www2.ed.gov/about/offices/list/ope/trio/recentfunding.html</bibtext> </blist> <blist> <bibtext> U.S. Department of Education. (2019). Programs. Talent search program. Office of Post-Secondary Education. https://www2.ed.gov/programs/triotalent/funding.html</bibtext> </blist> <blist> <bibtext> U.S. Department of Education, National Center for Education Statistics. (2017). Table 302.30 Percentage of recent high school completers enrolled in college, by income level: 1975 through 2016. In Digest of Education Statistics. U.S. Department of Education. https://nces.ed.gov/programs/digest/d17/tables/dt17_302.30.asp?current=yes</bibtext> </blist> <blist> <bibtext> U.S. Department of Education, National Center for Education Statistics. (2018). Table 302.20 Percentage of recent high school completers enrolled in college, by race/ethnicity: 1960 through 2017. In Digest of Education Statistics. U.S. Department of Education. https://nces.ed.gov/programs/digest/d18/tables/dt18_302.20.asp?current=yes</bibtext> </blist> <blist> <bibtext> What Works Clearinghouse. (2010). AVID (Advancement Via Individual Determination) (WWC intervention Report). U.S. Department of Education.</bibtext> </blist> <blist> <bibtext> Zhan, M., & Sherraden, M. (2011). Assets and liabilities, educational expectations, and children's college degree attainment. Children and Youth Services Review, 33 (6), 846 – 854. https://doi.org/10.1016/j.childyouth.2010.12.006</bibtext> </blist> </ref> <aug> <p>By Abigail Todhunter-Reid; Amanda Burke; Paul Houchens and Mark Howard</p> <p>Reported by Author; Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib30" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib29" firstref="ref2"></nolink> <nolink nlid="nl3" bibid="bib27" firstref="ref3"></nolink> <nolink nlid="nl4" bibid="bib10" firstref="ref4"></nolink> <nolink nlid="nl5" bibid="bib32" firstref="ref5"></nolink> <nolink nlid="nl6" bibid="bib17" firstref="ref7"></nolink> <nolink nlid="nl7" bibid="bib22" firstref="ref9"></nolink> <nolink nlid="nl8" bibid="bib12" firstref="ref12"></nolink> <nolink nlid="nl9" bibid="bib18" firstref="ref13"></nolink> <nolink nlid="nl10" bibid="bib28" firstref="ref14"></nolink> <nolink nlid="nl11" bibid="bib13" firstref="ref16"></nolink> <nolink nlid="nl12" bibid="bib16" firstref="ref19"></nolink> <nolink nlid="nl13" bibid="bib31" firstref="ref21"></nolink> <nolink nlid="nl14" bibid="bib23" firstref="ref24"></nolink> <nolink nlid="nl15" bibid="bib25" firstref="ref25"></nolink> <nolink nlid="nl16" bibid="bib20" firstref="ref29"></nolink> <nolink nlid="nl17" bibid="bib19" firstref="ref31"></nolink> <nolink nlid="nl18" bibid="bib14" firstref="ref37"></nolink> <nolink nlid="nl19" bibid="bib15" firstref="ref39"></nolink> <nolink nlid="nl20" bibid="bib26" firstref="ref40"></nolink> <nolink nlid="nl21" bibid="bib11" firstref="ref48"></nolink> <nolink nlid="nl22" bibid="bib24" firstref="ref49"></nolink> <nolink nlid="nl23" bibid="bib21" firstref="ref55"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1275427
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Avid Participation in High School and Post-Secondary Success: An Evaluation and Cost Analysis
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Todhunter-Reid%2C+Abigail%22">Todhunter-Reid, Abigail</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0002-4677-8235">0000-0002-4677-8235</externalLink>)<br /><searchLink fieldCode="AR" term="%22Burke%2C+Amanda%22">Burke, Amanda</searchLink><br /><searchLink fieldCode="AR" term="%22Houchens%2C+Paul%22">Houchens, Paul</searchLink><br /><searchLink fieldCode="AR" term="%22Howard%2C+Mark%22">Howard, Mark</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Journal+of+Research+on+Educational+Effectiveness%22"><i>Journal of Research on Educational Effectiveness</i></searchLink>. 2020 13(4):679-701.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 23
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2020
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22High+Schools%22">High Schools</searchLink><br /><searchLink fieldCode="EL" term="%22Secondary+Education%22">Secondary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Program+Evaluation%22">Program Evaluation</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Participation%22">Student Participation</searchLink><br /><searchLink fieldCode="DE" term="%22High+School+Students%22">High School Students</searchLink><br /><searchLink fieldCode="DE" term="%22Cost+Effectiveness%22">Cost Effectiveness</searchLink><br /><searchLink fieldCode="DE" term="%22High+School+Graduates%22">High School Graduates</searchLink><br /><searchLink fieldCode="DE" term="%22Enrollment%22">Enrollment</searchLink><br /><searchLink fieldCode="DE" term="%22Predictor+Variables%22">Predictor Variables</searchLink><br /><searchLink fieldCode="DE" term="%22College+Preparation%22">College Preparation</searchLink><br /><searchLink fieldCode="DE" term="%22Low+Income+Students%22">Low Income Students</searchLink><br /><searchLink fieldCode="DE" term="%22Minority+Group+Students%22">Minority Group Students</searchLink>
– Name: Subject
  Label: Geographic Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Florida%22">Florida</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1080/19345747.2020.1772421
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 1934-5747
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: The purpose of this study is to evaluate the Advancement Via Individual Determination (AVID) program and to derive estimates of program costs. We used the Coarsened Exact Matching approach to match AVID students with non-AVID students on 40 baseline characteristics. After matching, we estimated group differences in high-school graduation and college enrollment. We used the ingredients method to estimate program costs and calculated cost-effectiveness ratios by the duration of participation. Findings indicate that students who complete at least one AVID elective have higher high school graduation and college enrollment rates than comparable non-AVID students. We discuss how AVID compares to other college outreach programs in terms of costs and effects.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: CodeSource
  Label: What Works Clearinghouse (WWC) Reviewed
  Group: SrcInfo
  Data: Meets Evidence Standards with Reservations
– Name: URL
  Label: WWC Study Page
  Group: URL
  Data: <link linkTarget="URL" linkTerm="https://ies.ed.gov/ncee/wwc/Study/90255" linkWindow="_blank">https://ies.ed.gov/ncee/wwc/Study/90255</link>
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2020
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1275427
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1275427
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1080/19345747.2020.1772421
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 23
        StartPage: 679
    Subjects:
      – SubjectFull: Program Evaluation
        Type: general
      – SubjectFull: Student Participation
        Type: general
      – SubjectFull: High School Students
        Type: general
      – SubjectFull: Cost Effectiveness
        Type: general
      – SubjectFull: High School Graduates
        Type: general
      – SubjectFull: Enrollment
        Type: general
      – SubjectFull: Predictor Variables
        Type: general
      – SubjectFull: College Preparation
        Type: general
      – SubjectFull: Low Income Students
        Type: general
      – SubjectFull: Minority Group Students
        Type: general
      – SubjectFull: Florida
        Type: general
    Titles:
      – TitleFull: Avid Participation in High School and Post-Secondary Success: An Evaluation and Cost Analysis
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Todhunter-Reid, Abigail
      – PersonEntity:
          Name:
            NameFull: Burke, Amanda
      – PersonEntity:
          Name:
            NameFull: Houchens, Paul
      – PersonEntity:
          Name:
            NameFull: Howard, Mark
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Type: published
              Y: 2020
          Identifiers:
            – Type: issn-print
              Value: 1934-5747
          Numbering:
            – Type: volume
              Value: 13
            – Type: issue
              Value: 4
          Titles:
            – TitleFull: Journal of Research on Educational Effectiveness
              Type: main
ResultId 1