Running head: UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 1
Unpacking the Effectivity of Summer Schools for Low-SES Students: An Exploratory Literature Review and Expert Consultation
F. Bontekoe University of Amsterdam
Student number: 11319755 Bachelor thesis
Bachelor Onderwijswetenschappen
Faculty of Social and Behavioural Sciences May, 2019
13445 words
Supervisor: Dhr. D. Jansen MSc
Abstract
According to research, students with a low socioeconomic status (SES) significantly lose knowledge and skills during summer vacation, which may lead to an achievement gap between students from high-SES and low-SES families. Summer schools could be part of the solution to this problem, as they focus on improving students’ academic competencies during summer. The present study aims to provide an overview of the characteristics of summer schools that could contribute to academic and non-academic outcomes of low-SES students, by reviewing the existing literature about summer schools. In order to gain further insight on summer schools in the Netherlands, three experts are interviewed about the effectivity and characteristics of Dutch summer schools. The results demonstrate that several input- and process characteristics are likely to contribute to summer school outcomes.
Input-characteristics that may contribute are maintaining a low student-teacher ratio, hiring high-quality teachers and providing a scripted and enriched curriculum. Several
process-characteristics that seem to be important are student motivation, starting preparations timely, clear communication and engagement of parents and community. This overview of
characteristics presents summer schools the tools to provide a more effective program for students with a low socioeconomic status. As the current study highlights the lack of Dutch research about summer schools in relation to SES, future research could examine
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 3
Table of Contents
Introduction ... 4
Research questions ... 11
Method ... 12
Overview of empirical results in research literature ... 13
Chapter 1. Characteristics of investigated summer schools ... 19
Chapter 2. Characteristics of effective summer schools that may contribute to the academic outcomes of low-SES students in primary and secondary education ... 27
Chapter 3. Characteristics of effective summer schools that may contribute to the non-academic outcomes of low-SES students in primary and secondary education ... 39
Chapter 4. Views of experts in the Netherlands about the effectivity and characteristics of summer schools for low-SES students ... 44
Discussion ... 54
Introduction
In a 2017 Ted-talk, Karim Abouelnaga, who grew up in a low-income neighbourhood himself, describes movingly how he discovered that loss of learning during the summer months is the main cause of the achievement gap between poor and rich children. Since students from low-income families receive little academic stimulation during summer
vacation, their achievements in academic knowledge and skills regress. To fight this learning gap, Abouelnaga developed an inspiring summer program for disadvantaged children to help them succeed in their academic career. Worldwide, an increasing number of students is learning in such out-of-school programs, which are increasingly receiving more scholarly attention (Akkerman & Bronkhorst, 2016).
The importance of schooling during the summer to prevent summer learning loss, is demonstrated by meta-analytic research conducted by Cooper, Nye, Charlton, Lindsay, and Greathouse (1996). They conclude that, during summer vacation, children in the United States forget parts of what they had learned the last school year, which often leads to losses in
knowledge and skills. Children can lose one to three months of schooling because of this ‘summer learning loss’ (Cooper et al., 1996), which may have consequences for their learning in subsequent grades. Research in 23.348 households in the United States shows that this learning gap seems to be bigger for children with a low socioeconomic status (SES) than for children with a higher SES (Gehernson, 2013). In the next section SES will be defined, followed by a presentation of the other concepts – summer school and effectivity – that are relevant in the present study.
SES and summer learning
Socioeconomic status is a complex construct, which can be defined as the relative position of an individual within the social class. This position is determined by a person’s
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 5
access to financial, social and cultural capital (Korous, Causadias, Bradley, & Luthar, 2018; Kraus, Piff, & Keltner, 2011). The concept is most commonly operationalised by three different components: income, educational attainment, and occupation (Diemer, Mistry, Wadsworth, Lopez & Reimers, 2013). Alexander, Entwisle and Olson (2007), for example, included the educational levels and job status of both parents, and students’ participation in meal subsidy programs at school to measure socioeconomic status. The researchers conducted anine-year longitudinal study about the relationship between socioeconomic status and summer learning and its long-term consequences. In their sample of 326 students, Alexander et al. (2007) found that the differences in summer learning during primary school years significantly contribute to the achievement gap between high-SES and low-SES students’ in the ninth year of the study. The study indicates that summer learning loss leads to increasing disadvantage for low-SES families across generations (Alexander et al., 2007). More recent research confirms these findings (Gehernson, 2013). Addressing the problem at the source, providing low-SES students with additional educational programs during summer could help reduce this educational achievement gap (Cooper et al., 1996; Lauer et al., 2006).
Summer schools
As highlighted, an increasing number of students learn in out-of-school contexts (Akkerman & Bronkhorst, 2016), resulting in supplementary educational interventions increasingly becoming the subject of scholarly attention. An example of such an intervention is a summer school, in which children receive supplementary academic instruction and practice during summer vacation (Park, Buchmann, Choi, & Merry, 2016). In general, summer schools are focused on improving academic and cognitive competencies of children (Faber, Timmerman & Kievitsbosch, 2014). A summer school is considered an out-of-school-time program, which means that the program takes place during hours when school is not in
session (Lauer et al., 2006). Out-of-school-time programs differ from the school context in varying degrees (Akkerman & Bronkhorst, 2016). For instance, out-of-school programs usually allow more time and space for students’ personal interest than school programs. However, in summer school there is often a high degree of continuity between the school context and the out-of-school context (Akkerman & Bronkhorst, 2016). Some summer programs are even organised by and in public schools besides their regular classes during the school year (Park et al., 2016). Summer schools are different from other out-of-school programs because they tend to be more focused on academic improvement (Lauer et al., 2006).
Summer schools can serve different purposes. For example, the goal of a summer program can be to remediate or prevent learning deficits, assist children with disabilities or help students meet requirements for passing on to the next grade. They can also focus on offering an enriched educational challenge to gifted students (Cooper, Charlton, Valentine, & Muhlenbruck, 2000). In all these goals, summer programs seem to be designed in such a way that they can meet the needs of a specific group of students, for instance students with a low socioeconomic status (Cooper et al., 2000).
Alongside the different purposes described above, there also exist different types of programs in which students can participate. As Figure 1 shows, such programs can be both private and public forms of supplementary education. Existing accounts, documented in Park et al. (2016), reveal that it is higher-SES families that mostly use private forms of
supplementary education. Hence, for the purpose of research delineation, this study will focus on public summer schools, which belong in the category of after-school academic programs (Park et al., 2016), as is also shown in the right half of Figure 1.
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 7
Figure 1. Overview of supplementary education. Adapted from: “Learning beyond the school walls: trends and implications.” By Park, H., Buchmann, C., Choi, J., & Merry, J. J. (2016). Annual Review of Sociology, 42, p. 235.
Effectivity of summer schools
Several meta-analytical studies have been conducted about the effectivity of summer schools. Cooper, Charlton, Valentine and Muhlenbruck (2000) reviewed 93 empirical studies about the effects of 89 summer schools in the United States. These summer schools were publicly funded or established by a public-private partnership. This meta-analysis examined different types of summer schools and concludes that both remedial summer programs and programs with different goals have a positive effect on students’ skills and knowledge. The effects seem to be bigger when summer schools provide individual or small group instruction. The study also demonstrates that the positive impact of summer school programs is smaller for students with a lower SES than for middle-class students. An explanation for this might be that summer schools in middle-class school districts have access to more social and economic resources.
In another meta-analytical study, Lauer et al. (2006) reviewed different forms of out-of-school-time math and reading programs for low-achieving students in the United States. They analysed 35 experimental and quasi-experimental studies. The results of this review demonstrate that summer schools have a positive impact on the achievement of low-performing students in mathematics and reading. No difference was found between the
effectivity of summer schools and other out-of-school programs (Lauer et al., 2006). More recently, Kim and Quinn (2014) conducted a meta-analysis of 41 classroom- and home-based summer reading interventions in the United States and Canada for primary school children from low-income families. According to this study, both classroom and home-based summer programs have a positive effect on different measures of reading achievement. The results indicate that classroom summer programs achieve more positive effects when the instruction is evidence-based. In contrast to Cooper et al. (2000), they conclude that low-income children benefit more from attending a summer program than children from mixed-income
backgrounds (Kim & Quinn, 2014). These different conclusions about the effectivity of summer schools for students with a low socioeconomic status, may be attributed to the aggregation that is inherent to meta-analysis, leaving practitioners and researchers with questions regarding underlying activities and processes of such programs.
Besides, as these three meta-analyses were conducted in North-America, conclusions may not be generalizable world-wide. Different contexts, along with different
operationalizations of both the socioeconomic and summer school construct, may fuel the investigation of a variety of effectivity dimensions. In the Netherlands, where research about summer schools is scarce, effectivity seems to be mostly defined in terms of preventing grade repetition. A lot of Dutch students have to repeat a grade in primary education, especially children with a lower SES (Driessen, 2013). Also, more than five percent of students in secondary education repeated a grade in 2015 (Bertling, Witvliet, Faber, & Timmerman, 2016). However, grade repetition is an inefficient, high-cost method for improving academic achievement, since students receive an extra year of instruction in all subjects, while their learning deficits are in only one or a few subjects (Van Vuuren & Van der Wiel, 2015). Therefore, it is important to prevent grade repetition. In the Netherlands, such prevention can
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 9
occur during the summer vacation, when summer schools provide students with opportunities to get back on track.
In order to reduce the number of students that repeat a grade, the Dutch council of secondary education (VO-raad) has provided support for spring- and summer schools since 2015, to help middle schools to establish their own summer programs or collaborate with external partners to organise such programs. Schools that apply for this support receive subsidies from the Ministry of Education. According to research on the effectiveness of these subsidised summer schools, they can be an effective way to reduce grade repetition, although no evidence has been found for long-term outcomes (Bertling et al., 2016). In conclusion, there are indications that summer schools can play a role in reducing both grade repetition and summer learning loss.
Study purpose
However, as mentioned above, there are still many questions concerning the characteristics and approaches of summer schools that relate to different outcomes of low-SES students. Therefore, a more descriptive review is needed. In order to structure the review, the present study draws on the input-process-output model (see figure 2) derived from
program theory, a theoretical approach to evaluation that describes how goals and outcomes of a program can be generated by the resources, activities and processes (Brousselle & Buregeya, 2018). The theory provides a framework to study which specific input- and
process-characteristics of a program contribute to the desired outcomes. Such a framework is often applied in school effectiveness research, where the school is seen as a system. In order to keep the system going, the school has certain inputs (e.g. resources, teachers), which are essential in the primary process and result in certain outcomes (Scheerens, 1990). As each system is likely to differ from other systems, in the present study, the model is purposefully
not defined a priori, but the input-process-output model is used as an exploratory guide to describe the input, process and outcomes of (effective) summer schools.
Several studies have been conducted that examine the effects of summer school for low-SES students, but this literature has not been reviewed before, using such a model. Therefore, this study will give an overview of the characteristics of summer schools that contribute to the academic and non-academic outcomes of low-SES students in primary and secondary education. To find an answer to this question, several sub-questions will be studied. First, the way summer schools are conceptualised in literature will be defined and the
characteristics of summer schools will be described. Second, the study will review literature on the characteristics of summer schools that contribute to the outcomes of low-SES students, with a specific distinction between academic and non-academic outcomes. Because little research on this subject has been conducted in the Netherlands, the views of experts about the effectivity and characteristics of Dutch summer schools will be described in the final chapter. The answer to this sub-question will be based on three interviews with Dutch experts who executed, supported or researched summer schools in the past few years.
Figure 2. Input-process-output model. Context
Input-characteristics
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 11
Research questions Main-question
What are the characteristics of effective summer schools that may contribute to the academic and non-academic outcomes of low-SES students in primary and secondary education? Sub-questions
1. What are the characteristics of investigated summer schools?
2. What are the characteristics of effective summer schools that may contribute to the academic outcomes of low-SES students in primary and secondary education?
3. What are the characteristics of effective summer schools that may contribute to the non-academic outcomes of low-SES students in primary and secondary education?
4. What are the views of experts in the Netherlands about the effectivity and characteristics of summer schools for low-SES students?
Method
For the literature review, a literature search has been conducted in three databases: ERIC, PsychInfo and WebofScience. The following search terms have been combined: 1. ((summer school*) OR (summer program*) OR (summer camp) OR (summer remedia*) OR (summer intervention)).ti,ab,id.
2. ((disadvantage*) OR (income) OR (ses) OR (socioeconomic status)).ti,ab,id.
3. ((middle school) OR (high school) OR (primary school) OR (elementary school) OR (grade school)).ti,ab,id.
As research results from the past two decades are most applicable to summer schools in the present time, the search has been limited to literature from the past sixteen years. Based on a screening of abstracts, 32 studies found in the databases were considered relevant. Inclusion criteria were that the study had to investigate the effectiveness of a summer school, feature students in primary or secondary education, be empirical and in English or Dutch. Based on these inclusion criteria, twenty of these studies have been included in the present literature review, primarily those featuring low-SES students. Three studies did not mention SES, but were considered relevant given their Dutch context.
To answer the fifth sub-question about the views of experts on the effectivity and characteristics of Dutch summer schools, three semi-structured interviews with experts have been conducted, based on the interview guide as shown in Appendix A. During the
interviews, experts were asked to react to the results of the literature review. These exploratory expert consultations were transcribed and coded for specific themes that were relevant for answering the sub-question (see table 3). Inductive and deductive coding were combined to develop a coding scheme that represents the content of the interviews.
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 13
Overview of empirical results in research literature
Table 1
Overview of the included literature Study
ID Author Country Focus of investigation Design + sample SES measure Input Process Outcomes Effective summer programs
1 Beach et
al., 2018 USA Effect of summer schools focused on sight-word efficiency and oral reading fluency Quasi-experimental – 32 participants from grades 2-3 Income Licensed teachers, teacher training and coaching, evidence-based, scripted curriculum, student-teacher ratio 3:1, 15 hours instruction - Academic: Steady reading skills in 2nd graders, significant improvement in reading skills on all measures but one for 3th graders
2 Bekman et
al., 2011 Turkey Effects of a pre-school intervention on literacy, numeracy and language skills Quasi-experimental – 185 participants, 6-year-old preschool students Economic disadvantage, defined by income, job level and occupation of parents 10-week program for 25 hours a week, licensed teachers, teacher training and coaching, student-teacher ratio 10:1 Parental involvement, through home visits and providing information Academic: Significant improvement in literacy skills, pre-numeracy skills, several measures of syntactic knowledge and story comprehension. Effect of parents’ educational level 3 Borman &
Dowling, 2006
USA Effects of summer program focused on reading, writing, and hands-on math Longitudinal, experimental – 686 participants from
Income Seven weeks of instruction, 3 weeks of teacher training, undergraduate Community involvement through fieldtrips, parental Academic: Significant longitudinal effect for reading and writing, when controlled for students’ attendance.
activities kindergarten
and 1st grade teachers, teacher-student ratio 8:1 involvement Effect of parents’ educational level on students’ reading. 4 Botorff,
2011 USA Comparing effects of a 4-week and a 6-week summer program on reading and mathematics, in 4 subgroups of students Experimental – 120 students from grade 6-8 Generic measure, SES operationalized by free-and reduced lunch prizes Duration of 4-6
weeks - Academic: Significant differences between four- and six-week
interventions for the mathematics outcome in the African American and special needs education subgroups. No effect for low-SES or white subgroup 5 Chaplin & Capizzano, 2006 USA Effects of a summer program on reading, academic self-perception, social behaviour and parental involvement Quasi-experimental – 835 participants, from grades 1-7 Income and college attendance of parents Scripted curriculum, duration of 25-30 instruction days Parental involvement, community involvement Academic: Significant effects on students’ reading Non-academic: Effects on the degree to which parents encouraged their children to read (parental involvement with child) 6 Eidahl,
2011 USA Effects of summer school on reading fluency, accuracy and comprehension Cross-sectional – 280 participants from grades 1-2 Generic measure, SES operationalized by free-and reduced lunch prizes Scripted curriculum, 15 instruction days - Academic: Significant effects on students’ reading fluency, accuracy, and comprehension. 7 Hamel,
2014 USA Effects of a summer program on career and college-going self efficacy Quasi-experimental – 52 students from grades 6-7 Income and education level of parents Six days of hands-on college-related activities, varying curriculum, 3-day - Non-academic:
Significant short term effects of the summer program for college-going and career
self-UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 15
staff training efficacy
8 Johnston,
2015 USA Effects of a summer reading program to prevent learning loss Quasi-experimental – 48 students from grades 2-7 Generic measure, SES operationalized by free and reduced lunch/breakfast prizes and scholarships to pay tuition Evidence-based, scripted curriculum, 3 weeks of instruction, 9 hours a week, student-teacher ratio 9:1 - Academic: Significant increase of students’ reading fluency and comprehension
9 Little et al.,
2018 USA Effects of summer school on high potential students’ math skills Quasi-experimental – 220 students from kindergarten to grade 2 Income Research-based practices in structured curriculum, 3-4 week program, professional development of teachers - Academic: Effect on mathematic achievement for both low- and normal income students.
10 Luftig,
2003 USA Comparing a 12 hour- public and a 27-hour private summer reading program Quasi-experimental – 36 students from grades 1-4
Income Public program: 10:1 student-teacher-ratio, licensed and experienced teachers Private program: One-on-one instruction, use of computer technology - Academic: Effect on students’ reading, significant F-value of 11.47. No difference between public and private program
11 Schachter,
2005 USA Evaluating impact of a summer reading
Experimental, longitudinal – 162 students
Income 7-week, 2 hours a day instruction, student-teacher
- Academic: Significant
short-term effects on reading comprehension
intervention from first
grade ratio 15:1, licensed teachers, scripted curriculum and decoding 12 Xu & De Arment, 2017
USA Short-term and long-term effects of summer literacy program Quasi-experimental longitudinal – 26 preschool participants
Income Duration of six weeks – 4 days a week, scripted curriculum, hands-on learning, literacy coaches, licensed teachers, professional development sessions Parental involvement in transition to kindergarten Academic: Significant effect of program on name writing, upper-case recognition and print/word awareness, no effect on rhyme awareness 13 Zvoch & Stevens, 2012
USA Effects of summer reading program and demographic indicators Quasi-experimental – 76 participants from kindergarten and first grade Income 5 weeks of instruction, 4 days a week, small class sizes
- Academic: Significant
effect on oral reading fluency, student ethnicity was the only significant demographic predictor. Income (operationalized by free-or reduced lunch prizes) was not a
significant predictor Less effective summer programs or inconclusive results
14 Bakle, 2010 USA Summer program focused on reading, language usage, and math
Longitudinal, quasi-experimental design – 850 participants from grade 2-5 Generic measure, SES defined by parental income, education and occupation, Student-teacher ratio 15:1, non-scripted curriculum, 60 hours instruction
- Academic: Only effective
for 2nd and 3th grade low-SES students in language/math, no effect or negative effect for other grade levels or SES-levels.
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 17 and social networks 15 Bowers & Schwartz, 2018
USA Effects of summer
reading program Quasi-experimental (no control group) – 686 participants from kindergarten and first grade Generic measure, operationalized by free-and reduced lunch prizes and neighbourhood 6-week program, 29 days of instruction, student-teacher ratio 3:1, non-licensed teachers, 1-day teacher training, scripted curriculum Student
motivation Academic: Significant effects for oral and written narratives, but not for reading fluency and comprehension. No learning loss.
16 McDaniel
et al., 2017 USA Effects of a summer program focused on oral reading fluency Quasi-experimental – 31 participants, ages 6-7 Income Experienced teachers, teacher training, 9-weeks of instruction, scripted curriculum - Academic: No significant
gain in reading fluency, no learning loss either.
Qualitative studies 17 Bekman et
al., 2016 Netherlands Characteristics, preparation and process of summer schools Qualitative, 79 different summer schools for 2121 middle and high school students SES is not mentioned explicitly 10 days of instruction on average, high quality lesson materials Clear communication, starting preparations timely Academic: 83% of participants passed on to the next grade after attending summer school, 79% passed the grade in the school year after that as well.
18 Faber, 2014 Netherla
nds Effects of summer school on preventing grade repetition Cross-sectional, qualitative design, 320 participants (15 programs) SES is not mentioned explicitly Student-teacher ratio 6:1, qualified, external teachers, structured curriculum Positive support from teachers, student motivation, small groups, parental Academic: 86% of participants passed on to the next grade after attending summer school
from middle and high school
involvement
19 Greenman,
2013 USA How experiential summer programs affect academic and non-academic outcomes Qualitative, case-study – 2 different summer schools, 120 students from grades 5-8 Income Duration of 6-weeks, 80 hours of instruction, student-teacher ratio 15:1, hands-on enriched activities, active learning Summer school feeling different to student than regular school, community engagement Academic: Effects on mathematics and literacy skills
Non-academic:
Development of social and emotional skills and 21st century skills 20 Haelermans
et al., 2017 Netherlands Process and effect of spring- and summer schools Qualitative, process- and impact assessment – 96 spring-schools, 133 summer schools, middle and high school SES is not mentioned explicitly - Communication towards parents and students, starting preparations timely Academic: 87% of participants passed on to the next grade after attending summer school. Students had a smaller chance of repeating a grade after participating in summer schools
Running head: UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 19
Chapter 1. Characteristics of investigated summer schools
This literature review will study the characteristics of effective summer schools that may contribute to the academic and non-academic outcomes of low-SES students in primary and secondary education. In order to find an answer to such question, it is important to outline the characteristics of the summer schools that are included in this review. Although the main focus of summer schools is improving students’ academic and cognitive competencies, programs can have several different purposes (Faber et al., 2014; Park et al., 2016). For example, summer programs can aim to remediate or prevent learning deficits, support disabled children, offer extra challenges to gifted students, or help students to pass on to the next grade (Cooper et al., 2000).
The present study is specifically focused on public summer schools for primary and secondary school students. However, upon further inspection, several included articles do not fully comply with this definition of summer schools, highlighting the conceptual ambiguity that exists in the research base. First of all, Luftig (2003) does not only study publicly funded summer schools but compares the effects of the same intervention provided by a private company or the local public school district. Second, two of the included articles involve summer schools for preschool children, who are not yet attending primary school (Bekman, Aksu-Koç, & Erguvanli-Taylan, 2011; Xu & De Arment, 2017). In spite of these different sample characteristics, the summer schools are thoroughly described in the studies, therefore making a valuable contribution to this review.
Given the conceptual ambiguity described above, it is important to find an elaborate example of a summer school fitting the conceptualisation outlined in the introduction, to distil characteristics from such an example. A suitable summer school for this purpose is a public summer program organised by BELL, a community-based organisation in the United States that offers supplementary education to low-income children (Chaplin & Capizzano, 2006).
Chaplin and Capizzano (2006) provide an elaborate description of the summer program and detailed information about both outcomes and input- and process characteristics.
The most prominent goal of this program was to improve academic math and literacy skills, but the program also has several non-academic goals, like social and emotional
development, building self-esteem and stimulating parent participation. Such summer schools took place in five different locations, only three of which were included in the study. The programs had a duration of five or six weeks, with classes offered for five days a week, eight hours a day. The summer school targeted academically challenged low-income students from minority groups, attending grades one to seven. Pre-designed learning methods were used for both math and reading and the children were divided into groups of fifteen, each group guided by a qualified teacher and a teaching assistant. Chaplin and Capizzano (2006) studied the effects of the BELL summer program using a quasi-experimental approach. Random assignment was used to assign 835 participants to the treatment or control group. Outcome measures that were studied were reading skills, academic self-concept, non-academic behaviour, and parental involvement in students’ lives. The SES-measures used in the research were family income and college attendance of the parents (Chaplin & Capizzano, 2006).
Based on the definition of summer schools provided in the introduction and the example described above, in the current literature review, the following criteria are adopted: duration, context, teachers, academic outcomes, non-academic outcomes, method of the study, participants, and SES-measure used in the study.
Duration
First of all, the time span of summer schools differs across the studies. The duration of the included summer schools varies from one week (e.g., Hamel, 2014) to ten weeks (e.g.,
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 21
Bekman et al., 2011), with an average length of five weeks in the 20 studies. The intensity of summer schools, meaning the amount of hours students spend in the program each day, is different for each study, although not every study provides information about intensity. For instance, the least intensive summer program provides only one hour of instruction each day (Beach, McIntyre, Phillippakos, Mraz, Pilonieta, & Vintinner, 2018), while one of the programs is in session for eight hours a day (Chaplin & Capizzano, 2006). As table 1
suggests, duration and intensity, which both contribute to the total amount of instruction hours in a summer program, differ for each summer school.
Context
The duration of summer schools partly depends on the country and context the program takes place in. Most of the included studies (k = 16) are conducted in the United States, where schools are out of session for eight to twelve weeks during summer vacation. This is in contrast with three of the included studies, which have been conducted in the Netherlands. Dutch summer recess takes approximately six weeks (Rijksoverheid, n.d.). This difference possibly explains why summer learning loss seems to be an issue in the United States, while no evidence of the problem has been found for the Dutch context (Cooper et al., 1996; Driessen, 2013). As a result, most Dutch summer schools set different goals than summer programs in the US, for instance by defining effectivity in terms of preventing grade repetition (Faber et al., 2014). Program objectives and content may also be different for the Turkish summer school, which is included in the review, as it takes place in a completely different, non-western context (Bekman et al., 2011). Hence, the characteristics of summer schools that contribute to student outcomes may differ for each context, as cultural,
Teachers
An important characteristic to explore is the quality of the teachers. Not every study provides information about (aspects related to) teachers, but the literature indicates a lot of heterogeneity among summer schools in terms of teachers. Some summer schools recruited professional, licensed teachers, while other programs work with college-students and volunteers as instructors. Several programs have a team composed of both professional teachers and non-professional instructors (e.g. McDaniel, Mcleod, & Carter, 2017). Other schools work with a combination of a teacher and teacher-assistant for each class (Bekman et al., 2011). Besides the quality of teachers, table 1 suggests that student-teacher ratios also differ among summer schools. For instance, the summer school studied by McDaniel et al. (2017) has six teachers for a group of 31 students, meaning a one to five ratio. The biggest class size found is one teacher for a group of fifteen students (Bakle, 2010). This indicates that teacher-related characteristics of summer schools vary among the studies.
Measuring outcomes
Except for one, all of the included summer schools have academic aims, which complies with the definition of summer schools described in the introduction. In fact, only three of the 20 included studies researched the non-academic outcomes of summer schools. The most significant part (k = 12) of the summer programs have multiple objectives, while some of them focus on one academic skill (k = 8), as can be seen in table 1. In order to teach the academic and non-academic knowledge and skills, many studies (k = 11) rapport the use of a scripted or structured curriculum.
Academic outcomes. Different academic outcomes are measured in the studies, in order to analyse the effectivity of summer programs. The majority (k = 12) of summer
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 23
schools aims to improve students’ reading and literacy skills, as described in table 1. Other academic outcomes studied are numeracy and mathematics, language skills, and writing. Academic outcomes are mostly measured by oral or written tests. For instance, Little, Adelson, Kearney, Cash, and O’Brien (2018) used the ‘Measures of Academic Progress – Mathematics’, a computer-adaptive test to measure mathematics achievement. Johnston, Riley, Ryan and Kelly-Vance (2015) measured oral reading fluency, using oral tests
administered by trained researchers. Students were assessed in the fall before summer school and in the spring after summer school. Three studies, in the Dutch context, researched
summer schools with the primary goal of helping students pass on to the next grade. These summer schools provided additional instruction in many different academic subjects.
Non-academic outcomes. Given that summer schools’ main focus is on academic knowledge and competencies, only three of the included studies explored the effects of summer schools on non-academic outcomes. First of all, Chaplin and Capizzano (2006) studied the effect of a summer school on students’ social-emotional development and self-esteem, and on the extent to which parents participated in their children’s lives. The researchers used a self-evaluation instrument to measure self-esteem, while they measured non-academic behaviour of the child through the ‘Social Skills Rating System’ filled in by parents. Second, Greenman (2013) researched the effect of summer school on social and emotional growth and 21st-century skills such as problem-solving and critical thinking. This qualitative study used interviews and non-participating observation to measure non-academic outcomes. Finally, Hamel (2014) researched a non-academic summer school with career and college-going self-efficacy as the major goal. They used self-evaluation instruments like the ‘Middle School Self-Efficacy Scale’ and the ‘College-Going Self-Efficacy scale’ to measure these non-academic outcomes. In conclusion, academic and non-academic goals and
Method
As Table 1 shows, the studies included in this review used different research designs. The majority (k = 15) used an experimental or quasi-experimental design, while the other studies were cross-sectional. Three of the articles describe qualitative research, using a case study design or a process-evaluation. Longitudinal research was conducted in four of the studies. Sample sizes ranged from 30 to 850 participants.
Participants
Although the characteristics of participants vary for each study, the majority of summer schools in the included literature only targeted towards elementary school students, particularly from lower grades. Table 1 shows that twelve of the studies had participants in either grades 1 or 2, or both. Two studies included only preschool participants. Eight summer schools included participants in middle school, grades 6-8. The included Dutch studies (k = 3) are the only ones including high school students in their sample. In these summer schools, participants from the second-last year of high school were overrepresented. In the
Netherlands, the education system differentiates students into different levels from the age of twelve. Participants in the Dutch studies had all different levels, but most of them attended the second-highest level of school (havo).
In half of the included studies, students who participated were predominantly African-American or Hispanic-African-American. In three studies, participants from special needs education were included. For example, Greenman (2013) researched two different summer schools. In the first one, 70% of the student population lived in poverty, 46% was Hispanic or African American and 17% enrolled in special education. In the second summer school Greenman (2013) studied, 94% lived in poverty, 88% was Hispanic or African American and 24% were
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 25
students with special needs.
SES measure
Most participants in the included studies were from families with a low socioeconomic status. This construct has been operationalised using different measures.. Several studies (k = 5) speak of SES without further defining the concept, but operationalize it by qualification for free and reduced lunch prizes. One study defines SES by the combination of family income, levels of parental education and occupation, and the extent of social networks (Bakle, 2010). Generally, operationalisations of SES include one of these factors. Most research (k = 8) focuses only on the variable family-income, which is in some cases operationalised by qualification for free- and reduced lunch prices as well. Requirements for free and reduced lunch prizes are subject to state and federal regulations and guidelines. Three studies focus on other measures for SES. For example, Bekman et al. (2011) use the concept of economic disadvantage, which they operationalise by income, educational level and occupation of the parents. In another study, participants are low-income, first-generation students from ethnic minorities (Hamel, 2014). In three of the included studies, which are all conducted in the Netherlands, SES is not explicitly mentioned. The focus on SES can, however, take different forms, for instance when the summer school aimed to recruit low-SES students, resulting in low-SES being a starting condition in the sample. Other studies factored SES in as a control variable, while the summer school did not specifically target low-SES students.
The description of these characteristics reveals a few shortcomings of the existing literature on summer schools. First, limited research has been found about summer schools for high school students. Second, the majority of studies is conducted in the United States, which provides a limited scope of the international literature about summer schools. Finally, some
researchers provide very little information about the design and content of the summer schools they studied. In several cases, there is little information about the summer school teachers, the intensity of the program, or the goals of the researched summer schools. The lack of information on the characteristics described may hamper our understanding of the characteristics that contribute to the academic and non-academic outcomes of students, which are described in Chapter 2 and Chapter 3, respectively.
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 27
Chapter 2. Characteristics of effective summer schools that may contribute to the academic outcomes of low-SES students in primary and secondary education
The characteristics of summer schools described in the previous chapter might each influence student-outcomes in their own way. Considering that many summer programs are aimed at improving academic competencies of specific groups of students, it is relevant to find out which characteristics may contribute to the low-SES students’ academic outcomes.
Duration
As can be seen in table 1, most included summer schools provide information about the program duration, which varies from one to ten weeks. Two studies specifically focused on researching the effects of duration on students’ academic outcomes. First of all, Botorff (2011) compared a four-week summer school program with three hours of instruction each day to a six-week program in the next year with 4,5 hours of instruction each day. Both summer schools aimed to improve reading comprehension, mathematics and other academic skills. This quasi-experimental study selected 4 subgroups of 30 participants from grades six to eight. The subgroups were African-American students, white students, students with special needs, and students that qualified for and received free or reduced lunch, thereby indicating low-income background. Results showed statistically significant differences between the four- and six-week interventions for the mathematics outcome in the African American and special needs education subgroups (Botorff, 2011). No significant effect was found in the sample consisting of other subgroups or for the other outcomes. These results suggest that the duration of summer schools may contribute to academic outcomes for certain target populations, but not explicitly for low-income students.
Using a similar empirical design, Luftig (2003) compared two summer reading programs targeted at students attending grades one to four. As described in table 1, results
show significant differences between the control group and both intervention groups. The first summer school was a two-week, public, school-based program, where students received instruction for six hours a week. The second program, provided by a private learning centre, was in session for three weeks, nine hours a week. The 36 participants in both summer schools were mostly African American students from low-income families. The results indicate that the summer schools both had an impact on students’ reading skills, but no difference was found between the public and the private intervention. As the mean amount of instruction hours differed significantly between both summer schools, this points out that more instruction time does not always contribute to better results (Luftig, 2003). From these two studies, no conclusions can be drawn concerning the influence of program-duration on low-SES students’ academic outcomes. Possibly, duration only influences the effectivity of summer school in specific situations, for certain subgroups of students. Hence, other characteristics ought to be considered as well.
Teachers
The quality of teachers in summer programs varies widely. However, quality is mostly based on training, certificates, and teaching experience, as shown in table 1. For instance, McDaniel, McLeod, and Carter (2017) studied a nine-week summer program focused on oral reading fluency, in a sample of 31 six- and seven-year-old students. The results demonstrate no significant gain in oral reading fluency, but no summer learning loss either. Although the program seems less effective, this study highlights the importance of teachers who are trained and experienced in education. Two of the six teachers in this summer school were licensed and had teaching experience in regular schooling. The other four had experience leading recreational activities and field trips with low-income students. All teachers received training to provide reading instruction, prior to the summer program. McDaniel et al. (2017)
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 29
emphasize that both recruiting experienced teachers and providing them with training to work with reading interventions, might help prevent summer reading loss. Important to note is that McDaniel et al did not measure any improvement of students’ reading skills after the summer program, which raises the question whether high-quality teachers can help promote positive results in other summer programs.
Table 1 shows that six effective summer schools hired experienced or licensed teachers, and five of them provided teacher training prior to the program. Besides, in many programs (k = 6), classes consisted of less than ten students. Beach, McIntyre, Phillippakos, Mraz, Pilonieta, and Vintinner (2018), for example, found moderate effects of a fifteen-day summer school on sight-word efficiency and oral reading fluency for third-grade students. Training and coaching teachers and using evidence-based instructional strategies were important features for the effectivity of this program. The ten teachers were all licensed and most of them were experienced. Teachers received training to implement the intervention and they were coached during the first week of the summer school. There was relatively more room for individual and small group instruction since the student-teacher ratio was
approximately three to one. In this way, the study stresses the importance of experienced and trained teachers in literacy-focused summer schools and the benefit of a low student-teacher ratio.
Similar to Beach et al. (2018), Schachter (2005) found significant short-term effects of a seven-week summer school for low-income first-grade students on both reading
comprehension and decoding. A class in this summer school consisted of fifteen students, guided by one licensed teacher. Although a class of fifteen students is relatively big for a summer program, the student-teacher ratio is still low compared to regular schools. An example of a summer program with a noticeably low student-teacher ratio is provided by Bowers and Schwartz (2018), who examined a summer literacy program with one teacher for
each three students. The results seem inconclusive, since they indicate significant effects for oral and written narratives, but not for measures of reading fluency and comprehension. However, in this six-week summer program for first to fifth-grade students, each group of eight students was instructed by a non-licensed teacher, supported by two graduate students in speech-language pathology. Hence, there were many possibilities for individual instruction and practice, but teachers were inexperienced and not licensed. Possibly, these two
characteristics, student-teacher ratio and quality of teachers, cooperate in stimulating students’ academic achievement.
Concluding from table 1 and the examples described above, hiring high-quality, experienced teachers and training them to work with the intervention may possibly contribute to the academic outcomes of a summer school. Keeping the classes small, with a low student-teacher ratio, might also have a positive effect on academic outcomes.
Curriculum
Besides the quality of the teachers, the curriculum of summer schools may also influence the effectivity. Table 1 demonstrates that six of the effective summer programs that were included in current review work with a scripted curriculum. For instance, Xu and De Arment (2017) studied a summer school for 26 pre-school students, aimed to improve literacy skills. This program ran for six weeks, with four full school days every week. The researchers found significant differences between program participants and control group students for name writing, upper-case recognition and print, and word awareness. No significant effect was found for rhyme awareness. The program was specifically designed to promote the transition from pre-school to kindergarten for students from low-SES backgrounds, for which a scripted curriculum was used, with hands-on learning activities and lesson plan templates, to improve consistency during the program. Literacy coaches were present to support teachers in
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 31
implementing this curriculum. Licensed preschool and kindergarten teachers came together twice in professional development sessions, before providing instruction in the summer program. As this summer school uses a scripted curriculum and seems to promote students’ academic literacy outcomes, the study provides first indications that scripted curriculum is a relevant characteristic for summer schools.
Another study emphasising scripted curriculum is the research of Beach et al. (2018), who found significant effects of a summer program on sight word efficiency and oral reading fluency. The summer school used an evidence-based scripted curriculum that provided explicit instruction in early reading skills and intensive guidance for students. In comparable research, Johnston et al. (2015), who studied a three-week summer reading program for students in second through seventh grade, highlight scripted curriculum as an important summer school characteristic. The results of this research indicate that the program effectively increased participants’ reading fluency and comprehension. The summer school used a
scripted curriculum, that included several evidence-based reading activities, like repeated reading, one-on-one reading, and research based comprehension strategies. Based on the results from table 1 and the examples described above (Beach et al., 2018; Johnston et al., 2015; Xu & De Arment, 2017), providing a scripted curriculum during a summer school may contribute to the academic outcomes of low-SES students.
In contrast, the summer school studied by Bakle (2010) did not use a scripted or evidence-based curriculum but had teachers prepare daily lesson plans and choose their own teaching materials, to meet the needs of their students. The researchers found no significant effects for math, reading, and language, for neither of the grade levels included in the study. Possibly, these results relate to the absence of a scripted curriculum, which could reinforce the presumption that a scripted curriculum may help support student outcomes.
Context
Although most studies included in this review (k = 16) were conducted in the United States, it is important to highlight several summer schools in different cultural and academic contexts. For example, Bekman et al. (2011) found significant effects of a Turkish summer school on numeracy and literacy outcomes. The intervention group consisted of preschool children from disadvantaged, low-income, multilingual families. The results demonstrate a significant impact of parental education on students’ linguistic skills, meaning that the summer program was particularly effective for students whose mothers had sufficient language knowledge to support the child. In several regions of Turkey in which the research was conducted, the population is predominantly multilingual, speaking Turkish, Kurdish and Arabic (Bekman et al., 2011). This multilingualism, mostly related to cultural context, could explain the starting position of students engaging in the summer school, and could thereby also have had an influence on the study.
Besides this Turkish study, Dutch research shows the importance of context for summer schools. Haelermans, Ghysels, Monfrance, Rud, and Groot (2017) studied the effects of 133 Dutch summer schools for middle and high school students. Students received extra instruction in different subjects, mostly English and Dutch language, mathematics and economics. The major goal of the Dutch summer programs included in table 1 (k = 3), which are subsidised by the government, is to prevent grade repetition. The results of Haelermans et al. (2017) demonstrate that participants in summer school pass on to the next grade
significantly more than nonparticipants. This example from the Netherlands demonstrates how the academic context can influence the goals that are pursued by a summer school (e.g. preventing grade repetition), and possibly the actual attainment of such goals. Hence, taken together, these studies provide indications that the context in which a program takes place can
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 33
impact the program, the participants and the goals. However, based on table 1, no conclusions can be drawn concerning the relation between context and effectivity of summer schools.
Participants
As to participants of summer schools, several characteristics stand out. Most studies include low-achieving students, who need remediation for learning deficits. However, the participants in the summer school studied by Little et al. (2018) are an exception to this population of low achieving students. The researchers used a quasi-experimental design to study a summer school for high potential kindergarten and first-grade students, partly from low-income backgrounds. Licensed teachers provided mathematics instruction for three hours a day during three or four weeks. Twenty students were assigned to either the treatment or control group. Results indicate a statistically significant difference between students who participated and students who did not. As results for low-income students did not differ from the rest of the group, it can be concluded, based on these results, that summer schools may be effective interventions for both low- and high-achieving students, possibly promoting
academic outcomes.
Besides students with various achievement levels, summer schools targeting a low-SES population often attract students from different ethnicities, which is conceptually different from income or parental education. Zvoch and Stevens (2012) studied a five-week summer school program for 76 kindergarten and first-grade students, that was in session for four days a week. Statistically significant differences between control and treatment group were found for oral reading fluency in both kindergarten and first grade. Student ethnicity was the only demographic indicator that was a significant predictor of this change in reading fluency, while no differences were found for income. This study highlights the importance of ethnicity as a characteristic of students. Botorff (2011), who studied the effects of summer
school on math and reading comprehension for four subgroups of participants, supports this conclusion. The researchers found significant effects only for the African-American and special education subgroups, but not for the white and low-income subgroups. However, the majority of the African-American participants were also included in the low-income
subgroup. These two studies provide first indications that ethnicity might have more impact on the effectivity of a summer school than socioeconomic status, contesting the boundaries of the socioeconomic construct that is described next.
SES
As discussed, the construct socioeconomic status is composed of many factors, like income, occupation and educational level of the parents. Authors examining summer schools in relation to SES operationalise the construct in different ways, while some do not specify how SES is measured. Since the majority of the studies has low-SES as the population under investigation, not including it as a control variable, no conclusions can be drawn from these studies concerning the influence of SES on the effects of summer schools. However, several studies did include SES as a control variable (k = 7), as is described in table 1. Some of these studies (k = 4) demonstrate that SES is related to the academic outcomes of summer schools. For example, Eidahl (2011) studied the effects of a fifteen-day summer school for low- and high-SES second-grade students. In this study, SES was operationalized by qualification for free- and reduced lunch prizes. Eidahl (2011) found significant effects of the program on students’ reading fluency, accuracy, and comprehension. Besides, the results indicate no significant difference between high- and low-SES students for reading fluency, while children from low-SES families did benefit more from the reading accuracy instruction than high-SES students. This study demonstrates how SES can possibly be an important factor for
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 35
Some authors looked at the different aspects of socioeconomic status. For instance, Borman and Dowling (2006) researched the impact of a seven-week summer school program mainly focused on reading comprehension and vocabulary. Participants in this longitudinal study were 686 kindergarten and first-grade students, mostly from low-income African-American families. The researchers found significant longitudinal effects for reading and writing, when controlling for students’ attendance. Besides, they found that parents’ education level was a significant predictor of the academic growth in terms of reading of students over a three-year period. In supporting of this outcome, Bekman et al. (2011) found a significant relationship between parental education and summer school outcomes. Their results
demonstrate that children show more gains in vocabulary during the summer when they have a mother with at least five years of education, indicating that a literacy-focused summer school could be particularly effective when students’ parents have sufficient language knowledge to support the child. Together, table 1 and the studies described above provide indications that SES may possibly relate to students’ academic outcomes, although it remains unclear whether low-SES students benefit more or less from summer school than high-SES students.
Qualitative studies
Besides these characteristics, several features that contribute to the effectivity of summer schools have been pointed out by qualitative research. Greenman (2013), for
instance, conducted a multiple case study research, to study how experiential service learning summer programs keep academic knowledge flowing for middle school students. Experiential learning programs differ from traditional classroom instruction and can be defined as those placing students in active roles and providing cooperative hands-on activities. Greenman (2013) studied two six-week summer schools for students from grades six to eight. The first
summer school focused on safety, environment, and entrepreneurship. The summer school was in session six hours a day and the 120 participating students were guided by 20 staff members. The second summer school aimed to help students to live healthily and make smart choices. The 50 participants in this summer school received seven hours of instruction each day, provided by 13 staff members. In both summer schools, the majority of the participants were from low-income families and a considerable part was African American or Hispanic American (88% in the first program, 46% in the second). Using observation, interviews, and document review and analysis, Greenman (2013) concludes that both hands-on, experiential summer programs were effective at reducing summer learning loss for students. Affirming the previously drawn conclusion, these results emphasise the importance small class sizes, as both summer programs had classes that consisted of fifteen students at most. Introducing a new characteristic, Greenman (2013) stresses the importance of enrichment in the curriculum, which could make the summer school feel different to students than their experiences during regular schooling. Embedding the learning of academic skills and concepts in hands-on activities might make the program more enjoyable for children.
In different context, Faber, Timmerman, and Kievitsbosch (2014) conducted a qualitative study among fifteen Dutch summer schools for secondary education students. Although this research does not explicitly mention low-SES students, it was included as it provides rich information about summer schools in the Netherlands. It also adds information that can possibly enhance the generalisability of this review, since most other included research studied primary school students. In these summer schools, students who had failed a course during the school year received support in this particular subject, to prevent them from having to repeat a grade. The majority of the students received instruction in the Dutch language, English language or mathematics. The summer schools all had a student-teacher ratio of approximately six to one. The results of this study show that 86% of the participants,
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 37
who would have had to repeat a grade, could still pass on to the next grade after participating in summer school. Besides, several characteristics of summer schools are mentioned that may contribute to the academic outcomes.
First of all, based on interviews with teachers and project leaders, Faber et al (2014) conclude that student motivation may be an important factor for success. Students need to work hard to be able to remediate their learning deficiencies. The staff of a summer school can help promote the outcomes and motivation of students, by providing positive support and small group instruction. Employing external teachers is also mentioned as a characteristic that may promote academic outcomes. This is consistent with the conclusion of Greenman (2013), that it is important for summer school to feel different to students than regular school. Second, Faber et al. (2014) concluded from student evaluations that most summer schools fulfilled students expectations. However, students did emphasise the importance of subject-related expertise of teachers, which some of them had missed during the summer program.
Following up this research is a qualitative study conducted by Bertling et al. (2016). They studied the characteristics, preparation and implementation of 79 Dutch summer schools and the experiences of participating students and their parents. The secondary school summer programs were in session for ten days on average and students could follow two different courses. Based on the results, Bertling et al. (2016) emphasise several characteristics that contribute to the effectivity of summer schools. First of all, the results indicate the importance of proper communication with teachers, students, parents and external organisations. This can help promote support for the intervention among the stakeholders and collaboration between all concerned parties. According to Bertling et al. (2016), the purpose and content of the summer school should be communicated timely towards students and parents and school teachers should be informed in time about the criteria for participation and selection. Second, the researchers conclude that starting preparations timely could promote the effectiveness of a
summer program. Timely preparation allows one to carefully prepare the content, test methods and transmit information to involved stakeholders, which may aid the process of promoting students’ academic outcomes. This also applies to the preparation of curriculum methods. Results of this study emphasise the importance of lesson materials of sufficient quality, supporting the conclusion drawn before on the importance of a scripted curriculum (Beach et al., 2018; Xu and De Arment, 2017).
In conclusion, results of qualitative research help in adding aspects to the list of characteristics that contribute to the academic outcomes of summer schools for low-SES students. It may be helpful to enrich the curriculum with hands-on activities, in order to make summer school feel different than regular school (Greenman, 2013). Hiring external teachers who can provide positive support is also a characteristic deemed as contributing to
effectiveness, through the mechanism of inciting a ‘different feel’ at the summer school. Moreover, student motivation may have much influence on the effectivity of summer school for academic outcomes (Faber et al., 2014). Finally, clear communication, timely preparation, and high-quality curriculum are features that may improve effectivity of summer schools, according to qualitative research (Bertling et al., 2016).
Based on the analysed studies described in table 1, the following characteristics of summer schools may contribute to the academic outcomes of low-SES students: high-quality teachers, providing teacher training prior to the program, low student-teacher ratio, scripted and enriched curriculum, clear communication, timely preparation, student motivation, and inciting a ‘different feel’ in summer school. In the next chapter, the focus is on summer schools that aim to promote non-academic skills and the characteristics that may contribute to achievement of these outcomes.
UNPACKING THE EFFECTIVITY OF SUMMER SCHOOLS 39
Chapter 3. Characteristics of effective summer schools that may contribute to the non-academic outcomes of low-SES students in primary and secondary education
Besides aiming to improve academic outcomes, some summer schools also set non-academic goals or have side-effects concerning the non-non-academic development of students. Chaplin and Capizzano (2006) studied a summer school that aimed to improve both academic and non-academic achievement of students from grades one to four. They used a
quasi-experimental design with 835 participants from low-income families to study the effects of a summer school on literacy skills, academic self-perception, social and emotional development and involvement of parents in their children’s life and learning process. Concerning the non-academic outcomes, the only significant outcome was parental involvement, parents started reading more to their children and encouraged them to read more often at home. No effects were found on self-perception or social and emotional development.
The study describes three program-elements that might be associated with the observed outcomes. The first component, using a scripted curriculum, is consistent with the conclusions drawn in the previous chapter. The curriculum provided a step-by-step lesson plan for each grade level and instruction for teachers on how to handle different issues. The second element, focus on parental involvement, entails that parents were encouraged to become involved in their children’s development and learning process. Such encouragement might directly relate to both academic and non-academic outcomes in this study, as parents became more involved with their children’s reading. This conclusion is supported by table 1, which demonstrates that many studies (k = 5) emphasize the importance of parental
involvement for either academic or non-academic outcomes. The third element of the program studied by Chaplin and Capizzano (2006) is community involvement. The students went on weekly field trips and every week a prominent citizen in the community spoke to the students during the summer school. Hence, this study indicates that parental and community
involvement are characteristics that may contribute to non-academic outcomes of summer schools.
In contrast to Chaplin and Capizzano (2006), Greenman (2013) found significant effects of summer schools on several non-academic outcomes. As discussed in the previous chapter, Greenman (2013) used a multiple case-study design to study the effects of two middle school summer programs. Besides acquiring academic knowledge, students developed their 21st-century skills during the program. Students reported that they had improved at skills like problem-solving, critical thinking, collaboration, and communication, which are often seen as elements of 21st-century skills. Other non-academic outcomes were social and emotional growth. For instance, teachers emphasised the development of students’
persistence, willingness to answer questions and to take more risks. They also reported growth in confidence, development of friendships, and behavioural improvement. Parents mentioned that their children developed new friendships and ‘were more willing to try new things’ (Greenman, 2013, p.156).
Similar to the summer school studied by Chaplin and Capizzano (2006), Greenman (2013) concludes that community engagement may contribute to the success of a summer program. The programs studied by Greenman (2013) provide field trips for students and guest speakers from the community. Besides, several of the hands-on learning activities were designed to engage students in the community. This characteristic possibly contributes to the development of 21st-century skills and social and emotional growth. Another relevant element of the summer schools is enrichment in the summer curriculum. As mentioned in the previous chapter, this feature may contribute to the academic outcomes of students, but it may affect the non-academic outcomes of the summer program as well. The programs studied by Greenman (2013) implemented experiential learning, which provided practical, cooperative, and goal-directed activities. Enrichment entailed physical activities in the afternoon, field trips
