Mutual benefits for research and practice: Randomized controlled trials in the Hector Children's Academy Program

This article describes how a series of randomized controlled trials (RCTs) were implemented in the Hector Children's Academy Talent Development Program with the goal of generating mutual benefits for research and practice. The Hector Children's Academy Program, founded in 2010 and located in one of the largest states in Germany, Baden‐Württemberg, is a statewide enrichment program for talented primary school students, with a focus on STEMM topics. The program is financed by a private foundation, supervised by the Ministry of Education, and offered by 69 local sites that are hosted by local schools. About 15,000 primary school students (Grades 1–4) attend enrichment courses (more than 23,000 course participations) offered by the Hector Children's Academy Program every year. A unique element of the Hector Children's Academy Program is the role of empirical research in course development. The Hector Core Courses—offered by all local sites—undergo a strict quality assurance process in which RCTs are used to test their effectiveness with regard to central outcomes. After describing the Hector Children's Academy Program, we explain how the Hector Core Courses were developed and incorporated into the program, summarize key findings from the RCTs, and discuss mutual benefits for research and practice.

a few days or weeks; and they are offered by a variety of different stakeholders.
There is some empirical support for the effectiveness of enrichment programs.In the most recent meta-analysis available, based on 26 studies that used field data and were published between 1985 and 2014, Kim 4 found substantial positive effects of these programs on academic achievement and socioemotional development.More specifically, regarding academic achievement, Kim 4 integrated data from 13 studies, most of which analyzed the effects of specific enrichment courses.The meta-analysis yielded a positive effect on students' academic achievement of g = 0.96, 95% CI (0.64, 1.30).However, the empirical evidence for the effectiveness of the enrichment programs cited by Kim 4 is less impressive than it appeared at first glance.
First, all the studies included in the meta-analysis used either quasiexperimental or pre-post designs without randomization, leaving the door open for alternative explanations for the effects (e.g., selection bias).Second, the sample sizes were fairly small with a median of N = 54.Notably, positive effects of enrichment programs on several outcome variables have also been found in some older and some more recently published studies 5,6 that were not included in Kim's review. 4Nevertheless, as is often the case in research on educational effectiveness, the reviewed data, which had been collected in real-world settings-a desirable characteristic in terms of external validity-seemed to have come with some compromises in terms of methodological rigor. 7wever, it has been argued that the need for such compromises should not be taken for granted and that to advance research and practice, the most robust research designs available should be used on a more regular basis. 8In the present contribution, we describe how the systematic use of randomized controlled trials (RCTs) in the development of specific enrichment courses was established as a major pillar of the work of the Hector Children's Academy Program, the largest enrichment program for talented elementary grade students in Germany.In the following, we first describe the major features of the Hector Children's Academy Program.We then describe in some detail how a series of enrichment courses were systematically developed on the basis of the available research and how their effectiveness was tested with RCTs and provide an overview of some central results.
In the final section, we reflect on the mutual benefits and challenges of the research-practice collaboration as implemented in the Hector Children's Academy Program.

Definition of talent and the selection of students into the program
According to the operational definition used by the Hector Children's Academy Program, children are considered highly talented if they have far above-average potential in cognitive and creative areas; in addition, they have to show high general interest or deep interest in one area.The 10% most talented students in each cohort are eligible to participate in the enrichment program, meaning that the program is less selective than programs that follow a traditional, narrower academic giftedness definition, which targets only the top 2% or an even smaller proportion of students. 9The Hector Children's Academy Program is also broader in its inclusion criteria than programs that-in the tradition of Terman 10 -primarily rely on psychometric intelligence (for an example, see Cash and Lin 11 ).This more inclusive approach is in line with theoretical conceptions 2 that emphasize the developmental aspects in children's cognitive and creative biographies.Such developmental approaches are based on research that has highlighted the positive outcomes of individuals who received unexpected opportunities 12,13 and suggests that there is a much larger base of talent than is currently being tapped in highly selective programs.However, it has also been shown to place too much of the burden on teachers and have limitations, such as being biased against girls and students from low socioeconomic (SES) families, 14,15 some of which were also documented for the Hector Children' Academy Program.For instance, using sample sizes of well above 1000 students and including standardized measures for assessing a broad array of students' characteristics (e.g., motivation, school achievement, or general cognitive ability), Golle et al. 6 as well as Rothenbusch et al. 16 found that students from families of low SES were less likely to be admitted to the program than students from families of high SES even when general cognitive ability was controlled for.Furthermore, Rothenbusch et al.'s 16 findings indicated that teacher nominations in the Hector Children's Academy Program were also affected by reference group effects: Students in classes with many highly able students had a lower likelihood of getting nominated for the program than students with the same level of cognitive ability in classes with a smaller number of highly able students.Taking these findings into account, the program is currently re-evaluating its nomination strategy by examining the effects of introducing standardized screening instruments, a strategy that has been shown to be successful in increasing the participation of traditionally underrepresented groups. 17

COURSE DEVELOPMENT: A SERIES OF RANDOMIZED CONTROLLED TRIALS
In the beginning, the courses offered at the Hector Children Academies were largely developed and taught by local course instructors, and each of the many different courses was typically offered by only one local site, an approach we termed a "grassroots strategy." 6Not surprisingly, the courses covered a broad array of topics, targeted a diverse set of outcomes, and used a wide variety of educational approaches.To examine the general effectiveness of the program, Golle et al. 6 used a propensity score matching approach (which allows researchers to control for selection effects) to compare outcome variables from a total of eight domains (general cognitive abilities, academic achievement, investigative vocational interest, epistemic curiosity, creativity, self-control, self-concept, and social competence) between students attending the enrichment program (N = 423) and students who did not attend the program (N = 2328).Overall, we found positive effects of the program on school grades at the end of Grade 3 but no statistically significant effects on any of the other targeted outcomes.Given the large variability across the courses that were offered, these findings should perhaps not be surprising.Nevertheless, the findings triggered a streamlining process in terms of course development in the Hector Children's Academy Program.

Hector Core Courses: Talent development on a scientific basis
In recent years, and in terms of the development of the enrichment courses, the grassroots approach was supplemented and replaced to some extent by a systematic process of developing a core set of To evaluate the effectiveness of the courses, RCTs are systematically used.An RCT in education is an experiment in which the effects of an intervention are evaluated relative to another condition (i.e., a control group, which undergoes another intervention or does not F I G U R E 1 Phases of intervention design, stepwise implementation, and review in practice (figure from Herbein et al. 25 ; based on Gottfredson et al. 23 and Humphrey et al. 24 ).undergo an intervention) in authentic learning environments with random assignment to the intervention or control group.The decision to assess the effectiveness of courses in a series of RCTs was based on two considerations.First, in terms of practical relevance, the goal was to implement only courses of high quality and demonstrated effectiveness, and RCTs are believed to be the gold standard for evidence-based decision-making in education. 20Second, although the number of RCTs has grown steadily in recent years in research on education, 21 RCTs have not been as prevalent in research on talent development.Here, the current knowledge on the effects of enrichment classes is still based on a limited number of studies, most of which are characterized by methodological limitations.Hence, by using a series of RCTs, the research associated with the Hector Children's Academy Program is committed to the goal of contributing to both research and practice. 22

Systematic, stepwise approach in development and evaluation
The current definitions of RCTs do not specify the extent to which an intervention must be theoretically justified and the stages in which it should be developed and empirically tested.However, for RCTs with a basis in educational research or educational psychology, stage models have been proposed to systematize the development process. 23,24Ts in the Hector Children's Academy Program typically follow the logic described by Herbein et al. 25 (see Figure 1).In this model, the first step is to identify the needs of a specific target group (Phase 1).Obviously, various stakeholders, including both practitioners and researchers, contribute to this step.In the Hector Children's Academy Program, there are general guidelines in terms of the course content (i.e., focus on STEMM subjects), duration (8-12 weeks for a minimum of 45 min per session), or group size (6-10 children).Whereas the leadership at local sites sometimes has suggestions for the development of new courses.In the past, specific content as well as pedagogical approaches and the competencies to be acquired (e.g., computational thinking) have typically been decided upon by the involved researchers on the basis of recommendations from the relevant literature.
The second step focuses on the design or conceptualization of the intervention or treatment (Phase 2).Generally, this phase is crucial in terms of whether an RCT will primarily be an evaluation of a specific treatment or whether the RCT can also contribute to theory development.Existing RCTs differ widely in the extent to which theo-retical knowledge and empirical findings are used.Interventions that are based on strong theories on teaching and learning clearly defy the notion that RCTs are contributing to a naïve positivism 23,24 in education by not paying enough attention to understanding the learning processes that underlie all effective treatments.With regard to the Hector Core Courses, we decided to always use sound theories on learning and instruction as a basis for developing a course, albeit the theories differ in terms of, for instance, their origin (e.g., educational research, educational psychology, and developmental psychology) and their domain-specificity or domain-generality.
The conceptualization of the intervention is followed by a pilot study (Phase 3), which can include elements of design thinking approaches as well as the implementation of the intervention in a small number of academies, thus allowing researchers to gauge the practicability and acceptance of the intervention in the targeted real-world environment.
For each of the Hector Core Courses, a detailed course manual is used, and if necessary, changes are made to this manual as a consequence of the feedback received in the pilot phase.The pilot study is followed by stringent empirical testing across several phases.Phase 4, the efficacy study, has a strong emphasis on internal validity.Accordingly, high priority is given to implementing the intervention exactly as it was planned.Therefore, in the development of the Hector Core Courses, researchers or individuals who collaborate closely with researchers (e.g., student assistants) teach the courses and strictly follow the guidelines as described in the course manual.In line with the assumption that efficacy studies implement the factors that cause changes in students in a "pure" fashion that resembles a laboratory experiment in psychology, sample sizes are typically fairly small, rarely exceeding 100 students.whether the intervention is implemented as intended by the developer: so-called implementation fidelity, 23 which is typically assessed via instructor self-reports, learner reports, or observer ratings.Low implementation fidelity threatens the internal validity of the study and makes it impossible to determine whether the (lack of) effects can be attributed to the originally planned intervention.All current empirical studies with Hector Core Courses have included fidelity measures.
If an intervention demonstrates success in Phases 1-5, it can be further disseminated in practice (Phase 6: scaling-up).For instance, the goal could be to achieve widespread use within the original context or, alternatively, to expand the targeted context.Scaling-up also means that research can again address questions of acceptance, utility, and effectiveness in a larger context.Furthermore, issues such as sustainability and efficiency might be addressed. 23As in Phase 5, implementation fidelity is a major issue in the scaling-up phase.When scaling-up studies include larger numbers of course instructors and students than effectiveness studies, there is also the possibility to directly examine how differences in implementation fidelity across instructors and students impact the effectiveness of a course. 26

Comparison groups, treatment fidelity, and outcomes
By now, empirical results regarding the investigation of the Hector Core Courses are available for a total of 17 studies (see Table 1).
Whereas an exhaustive description of the design of all these evaluation studies is beyond the scope of the present article, we briefly address three very important methodological aspects: comparison groups, treatment fidelity, and outcome measures.

Comparison groups
The bulk of our studies have used a waitlist control group design, meaning that participants in the waitlist control condition participated in the course after the central outcome variables were assessed in both the treatment and control groups.Students in the waitlist control condition were typically invited to attend another, unrelated course from the program, but they did not have to do so.In one of the studies, the participants were randomized to participate in two different Hector Core Courses, which served as control groups for each other.

Treatment fidelity
In all studies in which trained course instructors instead of the course developers themselves provided a course, we assessed implementation fidelity-the extent to which the course was implemented as intended. 24Typically, eight dimensions of implementation can be distinguished, and accordingly, fidelity can be assessed.Thus, in efficacy, effectiveness, and scaling-up studies, multiple dimensions of the fidelity of treatment implementation were assessed to determine how well the instructors implemented the critical components of the courses. 32,33The assessment of treatment fidelity focuses on aspects of fidelity that refer to structure or processes 34 with a focus on adherence (i.e., compliance with the exercises in the course manual) and quality of delivery (i.e., the manner of implementation and thus how well the exercise was delivered 24,35,36 ).To increase the fidelity with the original script, also with regard to aspects of implementation and critical instructional components, all instructors were given a scripted course manual and took part in 4-8 h of training implemented by the original course developers.The manual contained information about the theoretical background as well as detailed explanations of all course units and time frames. 25tcome measures A variety of tests are being used to evaluate the Hector Core Courses, and the outcome measures include a fairly broad range of aspects of knowledge, skills, and competence.Priority is given to available standardized tests, but they are typically supplemented by additional measures constructed by the researchers responsible for the respective course.For a detailed description of the instruments that are being used in our studies, readers are referred to the original articles.

Overview of courses and empirical evidence of their effectiveness
So far, empirical data on course effects have been collected for a total of nine Hector Core Courses (see Table 1).All but one study has used randomized group assignment, and since 2018, all RCTs with Hector Core Courses have been preregistered.The sequence shown in Figure 1 has been used in almost all the courses.However, in the case of two courses, the efficacy trial phase was skipped, and the evaluation instead began with an effectiveness trial (Müller et al., manuscript in preparation; Tsarava et al., manuscript in preparation).Three courses are still in the efficacy phase, and scaling-up data are available for only two courses so far.As the project is ongoing, more data will be available in the future.The sample sizes for the completed studies range from 58 to 310 students, indicating that sample sizes for all studies were above the median sample size in the studies reported in the meta-analysis by Kim. 4 At the same time, the sample sizes were smaller than planned for some studies, especially in the case of the effectiveness phase of the "Talking about Science" course, which met with organizational problems and a shortage of qualified instructors.
The general approach to data analysis was similar across studies, although there were some differences in detail (see Table 1).To estimate the effects of the courses, a set of regression analyses were run.
In an initial model, a measure of achievement was used as the outcome variable, and course participation and a pretreatment score on the achievement measure were used as predictors.If there was an imbalance in potentially important covariates in the pretreatment stage, these covariates were also included in this initial model.In additional steps (see the respective publications), different course effects were explored.Each study contains more than one outcome of interest.The inclusion of several outcome variables allows some leeway in terms of what constitutes empirical support for the effectiveness of a course.Therefore, beginning in 2018, one or more so-called primary outcomes were defined in the preregistrations, and to be certified as a Hector Core Course, positive effects on at least one of these primary outcomes had to be found.

TA B L E 1
Table 1 reports the effect sizes for one central (primary) outcome for each of the Hector Core Courses in each of the studies completed so far.Effect sizes are reported as Cohen's d, which describes the difference between being a member of the treatment group versus the control group in predicting the standardized outcome variable while controlling for pretreatment differences on the outcome variable.In most of the studies, some additional covariates were used (e.g., to take into account preintervention differences across groups).For the sake of consistency, we report the effects as reported in the original analyses.
Although the findings cannot substitute for a formal meta-analysis of the empirical support for enrichment effects in the Hector Chil-dren's Academy Program, they suggest that the course development strategy yielded effective courses as indicated by statistically significant effects in one or more of the consecutive studies.The data also show that there is considerable variation in terms of the sizes of these effects, with some of the largest effect sizes found for "knowledge" outcomes.Moreover, in line with reviews on intervention effects, effect sizes tend to decrease from efficacy to effectiveness studies and from effectiveness to scaling-up studies.In other words, an increase in external validity by training instructors to implement the course is often coupled with a decrease in effect size.However, such a change does not imply that scaling-up generally results in smaller or zero effects.In fact, for some outcomes not reported here, there were opposing trends, at least descriptively.For instance, in the course "Little researchersworking like scientists," 26 for the outcome variable scientific inquiry competence-which was not assessed in the efficacy trial but was defined as a primary outcome in the scaling-up study-we found statistically significant positive treatment effects in both the effectiveness (effect size = 0.32) and scaling-up (effect size = 0.47) studies.Clearly, more research addressing the variability of effects across different outcomes (e.g., knowledge outcomes vs. measures of competence) and across various phases of the development of the courses is needed.

An example: The "Chemistry Lab" course
To illustrate the course development strategy in the Hector Children's Academy Program, we next describe the development and systematic evaluation of the Hector Core Course "Chemistry Lab."This course is designed to enhance scientific literacy in primary school students in Grades 3 and 4 by means of combining "hands-on" and "minds-on" practices. 37,38In line with other enrichment courses from the Hector Children's Academy Program, the course focuses on small groups of 6-12 students and consists of 10 120-min sessions.Each session covers specific conceptual knowledge by focusing on the core idea of matter and its interactions 38 and inquiry-based methods; this core topic is not part of typical curricula.In order to teach the core idea of matter and its interactions in primary school, it is important to help students develop a coherent understanding of matter on a phenomenal level instead of developing the atomic idea (atoms as discrete units).0][41] The research question was whether At the same time, research can also greatly benefit from such a close research-practice collaboration.In fact, to date, the access to enrichment classes has allowed us to run a set of 17 studies on the effectiveness of nine enrichment courses, which were developed on a scientific basis, with several additional studies under way.In this way, our research adds considerably to the still somewhat weak empirical literature on the effectiveness of enrichment courses. 4Each of the completed studies addressed specific questions related to the effectiveness of a specific intervention, and each yielded some important findings.However, the knowledge gained from our approach goes beyond the single studies.In fact, in forming a package of courses, the whole is more than the sum of its parts, thus potentially allowing us to address some higher-level issues in future research.For instance, as indicated above, there is the possibility to examine in detail the potentially decreasing effect sizes across the different phases of evaluation and to scrutinize some possible reasons for such a decrease-including aspects of treatment fidelity and instructors' expertise.Moreover, the large set of studies will also allow us to take a closer look at differential effects for various outcome variables.For instance, are effect sizes for "knowledge outcomes" larger or smaller than for "competence outcomes"?Similarly, to what extent do transfer effects exist beyond the effects that were at the core of the specific interventions?Answering these questions can contribute to the unwavering debate about transfer effects. 42,43As a final example, the set of studies on the Hector Core Courses might also inform the discussion about replication, in terms of both how our studies represent "successful" and "failed replications" and, more conceptually, the discussion of the extent to which the progression from efficacy trial to scaling-up study constitutes "direct" versus "conceptual" replications. 44spite the reported mutual gains of engaging in the research-

Founded
in 2010, the Hector Children's Academy Program remains a statewide enrichment program for talented primary school students (in Grades 1-4) with a focus on Science, Technology, Engineering, Mathematics, and Medical Sciences (STEMM) topics.It is located in Baden-Württemberg, one of the largest states in Germany.The program is financed by a private foundation, implemented and supervised by the State Ministry of Education, and supported by local communities and school authorities.There are currently 69 local sites of the pro-gram, typically hosted by one or more public schools that collaborate with several neighboring schools.Student attendance in the program is free of charge.The free access to the program and the integration of this enrichment program into existing school structures serves the goal of enhancing equal and fair chances for all eligible students to participate in the program.An important characteristic of the Hector Children's Academy Program is the integration of a research partnership with the University of Tübingen and the Leibniz Institute for Research and Information in Education, Frankfurt.In addition to evaluating various elements of the program at local and statewide levels, researchers have codeveloped enrichment courses and offer professional development opportunities for instructors at the academies.

About 15 ,
000 students attend enrichment courses (more than 23,000 course participations) offered by the local sites of the Hector Children's Academy Program every year.Admission to the program is organized by the local sites of the program.In line with the current rules and regulations of the program in almost all of the local sites, students are selected for the program on the basis of teachers' recommendations.This approach has well-known advantages, including being economical in terms of time and money.

2 "
Talent development: Conceptual backgroundTalent development in the Hector Children's Academy Program is guided by the central conceptual pillars found in recent work.2,18Importantly, the program focuses on early phases of talent development, which are characterized by a rather broad orientation plus some first steps toward building competence.According toSubotnik et al., In the earliest stage, it is the job of the teacher to engage the explicit or undeveloped interests of young people in a topic or domain and to engender and capitalize on motivation.At the next stage of development, it is critical that teachers help the individual to develop the needed skills, knowledge, and values associated with the acquisition of expertise in that domain."In line with this conceptualization, selected students are invited to choose from among a broad array of courses that go beyond the traditional curriculum in terms of either their breadth or their width.At least 60% of the courses offered by each academy cover STEMM subjects.Students are invited to attend at least one course per semester; and courses typically consist of 8-12 sessions 90 min long, thus allowing teachers to help students develop competencies and motivational dispositions in the respective areas.Most of the courses are offered outside regular school hours, but some courses are integrated into the whole-day program of local schools.Whereas some of the courses are open to students from Grades 1 to 4, most courses are specifically geared to only one or two grade levels.For instance, there is a German version of the Math Olympiad for third-and fourth-grade students, and consequently, the corresponding course (see below) at the Hector Children's Academy Program addresses these two grade levels.
manual-based enrichment classes, the Hector Core Courses, which are offered across all local sites and are attended by several thousand students (about 4,000 course participations) every year.The Hector Core Courses are characterized by the following features: (1) The courses are based on scientific accounts of what should be taught to talented primary school students and in what way.Somewhat simplified, there are three building blocks for each course-the general framework on talent development 2 as described above, available research on the specific domain to be taught (see the section "An example: The Chemistry Lab Course"), and research on adequate pedagogical approaches for talented children, oftentimes highlighting the need for inquiry-based elements. 19(2) The development and evaluation of the courses follow a transparent sequence with predefined steps (see below).(3) The courses typically cover STEMM domains and consist of around 8-12 sessions of 90-120 min in length to allow for deep involvement with the respective course content.(4) In order to ensure that the effective courses are delivered as intended (i.e., implemented into the Hector Children's Academy Program with high treatment fidelity), a course manual and course materials are provided, and the professional development of course instructors is part of the strategy for implementing the courses.(5) The courses need to be shown to be effective in fostering relevant outcomes using a robust research design before being permanently implemented into the program.

Conversely, in Phase 5 ,
the effectiveness study, a stronger focus is placed on external validity, meaning that the course should be taught in the regular educational environment by regular course instructors.In the context of the Hector Core Courses, training sessions with a duration of up to 2 days are offered for the instructors of the local sites of the program.In these training sessions, the course instructors (1) become familiar with the course manual and are provided with the course materials, (2) are invited to engage in short exercises, and (3) discuss the manual with the course developers.Although the required sample size depends, among other factors, on the expected effect sizes and the number of students per instructor, effectiveness studies in the Hector Children's Academy Program are typically planned to include 100 or more students.A major question in any effectiveness study is such a program could successfully foster students' conceptual, procedural, and epistemic knowledge as well as their motivation to engage in and their interest in science and scientific procedures already at this young age.So far, the development of the course has followed five of the six consecutive steps described above.The needs of the target group were identified by integrating feedback from the academies and by consulting the research literature (Step 1).After a conceptualization phase (Step 2) and a pilot study (Step 3), two experimental studies (Steps 4 and 5) were conducted to investigate the effectiveness of the science program (see Ref. 17).In 2014/2015, we carried out the first multisite RCT with repeated measures (N = 73).Students were randomly assigned to either the course or a waitlist control group.At this stage, the course followed a first scripted draft of the manual.The course instructors were master's students who were in the process of getting their teaching certificates in chemistry and were involved in developing the program.The results showed a significant positive intervention effect on conceptual knowledge (d = 1.2) but no consistent pattern of positive effects on motivational variables.After the first empirical study was completed, the program was adapted slightly to also increase motivation.In the spirit of problembased learning, the sessions were tied more closely to the children's everyday context.The course was structured consistently across all sessions, and the didactic instructions for the course instructors were formulated in a scripted course manual for each session.In 2017/2018, Study 2 was conducted to investigate the effectiveness of the adapted program (N = 247), again in a multisite RCT.For this purpose, course instructors who had experience working with chemicals (e.g., active and former teachers with basic knowledge of chemistry) were systematically qualified to teach the program, and a scripted manual was provided.A larger sample, a more heterogeneous group of course instructors, and different stages of material development characterized the differences between the first and second studies.The results of Study 2 still revealed a significant positive intervention effect on conceptual knowledge, d = 1.06.As in Study 1, the motivational variables as well as epistemic beliefs or procedural knowledge were not significantly affected by the program, thus providing the opportunity to further improve the course in additional cycles in the course development sequence.DISCUSSIONIn the present article, we described how rigorous empirical research can contribute to the development of gifted programs, and we outlined an ambitious attempt to achieve the important goal of bringing together research and practice in the field of talent development.1The strategy of developing the Hector Core Courses implements an integrated approach to putting interventions into practice, ranging from identifying the needs of a target group to scaling-up using RCTs.A huge advantage of this approach is that educational practice is provided with interventions that are grounded in theory and based on empirical evidence.Given the paucity of empirical support for enrichment programs, the breadth of the empirical evidence established for the Hector Core Courses, which by now are a central pillar of the Hector Children's Academy Program, is fairly unique; this evidence in turn has contributed to the acceptance and sustainability of the program whose original runtime had been set for 10 years.
practice collaboration in the Hector Children's Academy Program, such a collaboration also comes with challenges and risks.For example, implementing robust research designs often conflicts with educational practice, potentially lowering the acceptance of the input of the researchers.Conversely, demands by practitioners for support might put strains on the researchers' time budgets for scientific contributions.Even more problematic, changes in priorities in the involved education administrations might severely impede the further development of the program or its evaluation.An increased understanding of the importance of scientifically supported enrichment programs might constitute the best antidote to these risks.
Overview of empirical studies evaluating the effectiveness of the Hector Core Courses.
Note. d = Cohen's d; ns = not significant; Pub stat = Publication status; Ip = in prep; Pre = only preregistration published; Sub = submitted; and Pub = published.The unpublished results are categorized into small (d > 0.2), medium (d > 0.5), and large (d > 0.8) effect sizes and should be considered preliminary.The unpublished manuscripts are available upon request from the corresponding author.†p < 0.10; *p < 0.05; **p < 0.01; ***p < 0.001.a Statistical model: Group assignment and pretest scores as predictors.b Statistical model: Group assignment, pretest scores, and significant covariates as predictors, including an interaction between group assignment and pretest scores.c Statistical model: Group assignment, pretest scores, and significant covariates as predictors.d No randomization to experimental conditions.e Effect size for children in Grade 3. f Effect size for children in Grade 4.