Retention in Kindergarten for Students with Autism Spectrum Disorder: Improvement, Delusion or Both?

Student retention in kindergarten is a common practice so that students will obtain the necessary school preparation for elementary school. The current study examines the impacts of retention on the students with ASD. A sample of 110 students was taken from kindergartens all over Greece, and it was ob-served that the retention was advantageous, but not for all students. Retention has not actually been beneficial for students from families with low educational attainment as well as students from foreign families. Moreover, these findings apply to students who could not attend extracurricular supporting programs. A serious dilemma came up as a result of all the above about the real benefits of the retention.


Introduction
In Greece, in recent years, the number of applications for retention into kindergarten has significantly increased. The repetition of the kindergarten class, the so-called "retention", is carried out in accordance with paragraphs 7 and 8 of article 6 of Presidential Decree 79/2017 (Greece, 2017) as follows: "Retention in kindergarten for one (1) year may be implemented with the consent of the parents/legal guardians, when it is certified by the Centers of Educational and Counseling Support or a public medico-pedagogical center." Centers of Educa-Psychology 2010). Based on this approach, retention is considered as an attempt to enhance the student's knowledge and skills and at the same time as a means of preventing school failure in their later school career. In other words, this view supports the fact that retention will benefit students, who are not yet able to meet the academic requirements of the next level (Hong & Raudenbush, 2005;Hong &Yu, 2008;Mariano & Martorell, 2013;Peterson & DeGracie, 1987;Smith & Shepard, 1988;Vandecandelaere et al., 2016;Wu et al., 2010). However, there are numerous studies that show that the retention of children in kindergarten does not necessarily improve the academic performance of students at the next level and that there is no significant difference in the school performance in students who were retained in kindergarten compared to those who were promoted. Retention itself, without a supporting framework, can trap the student in a "repetition" of cognitive activities and the expected outcome would be the student withdrawal from the learning process whatsoever (Byrnes, 1989;Goldstein et al., 2014;Holmes, 1989;Mantzicopoulos & Morrison, 1992;Mendez et al., 2015;Pianta et al., 1997;Walters & Borgers, 1995;Yang, 2018). Furthermore, as noted in some cases, retention into kindergarten can be a traumatic experience for students, as they are forced to stay "behind" compared to their peers who are promoted (Hong & Raudenbush, 2005;Jacob & Lefgren, 2009;McCombs et al., 1992;Shepard, 1989;Smith & Shepard, 1988;Peixoto et al., 2016). Despite the discordance of research, kindergarten retention rates have increased significantly in recent decades (Bergin et al., 1996;Cosden et al., 1993;Mantzicopoulos & Morrison, 1992;Pianta et al., 1997;Warren et al., 2014).
In Greece, generally, students with ASD are expected to be retained. It has also been found that for many foreign students with ASD, it is proposed by the educational services of Greece that they should be retained. However, research into the potential benefits of kindergarten retention for students with ASD is lacking. Even the small number of surveys found in the Greek literature related to retention, focus on students without disabilities and mainly investigate the causes for which retention is proposed. Similarly, in international articles, such as Chen et al. (2014), Hong & Raudenbush (2005), Jimerson (1999), Lorence & Dworkin (2006), Moser et al. (2012), the positive effects of retention in strengthening the personality of students with ASD are not sufficiently highlighted. That is why the research was designed and implemented to study whether the retention of students with ASD, in the Greek kindergarten, improves their cognitive and social skills. Based on the above, the following research questions are posed.
Research questions (RQ) • RQ1: What is the impact of retention in students with ASD in kindergarten on their cognitive and social domain? • RQ2: What kind of support during retention in kindergarten correlates with the best outcomes for the students with ASD? • RQ3: Does the attendance of extracurricular supporting programs by the students with ASD affect the results of retention? • RQ4: When a student with ASD comes from a foreign family, what are the Psychology results of retention? • RQ5: How does the educational attainment of the parents affect the effectiveness of retention?

Participants
The study comprised of 110 students who were diagnosed with ASD according to the Diagnostic and Statistical Manual of Mental Disorders (5 th ed.; DSM-5; American Psychiatric Association, 2013), of whom 80 were boys (72.7%) and 30 girls (27.3%). Moreover, 20 students come from foreign families and 90 come from Greek families (see Table 1). The students come from foreign families were born in Greece and speak Greek as a second language. The age of the students at the beginning of the retention was from 5 years and 9 months to 6 years and 9 months with an average of 6 years and 4 months. Upon completion of retention, the age of the students was from 6 years and 6 months to 7 years and 6 months with an average of 7 years and 1 month of age. In Greece, during the school year 2019-2020, the students with ASD who were retained in kindergarten were 165. Thus, 66.66% of the total number of students with ASD who were retained participated in this research.

Research Tool
The psychotechnic tool we used to collect our data are Autism Treatment Evaluation Checklist (ATEC), created by the American Research Institute for autism (Edelson & Rimland, 1999). The internal consistency reliability is high (0.94 for the total score). ATEC can provide comparative results of the impact an intervention on students with ASD may have. In other words, it is not a diagnostic tool but an evaluation one. It consists of 4 scales that study Speech-Language-Communication, Sociability, Sensory-Cognitive Awareness and Health/physical/Behavior. Of the four scales, the first three were used. That is, the Speech-Language-Communication scale, the Sociability scale and the Sensory-Cognitive Awareness scale. The first scale consists of 14 questions, the second one of 18 and the third one of 20. They are all closed-format questions, and the answers are given on a 3-point Likert scale (1 = Not True, 2 = Little True, 3 = Very True). For the study of the cognitive domain, we used the first scale, which assesses Speech-Language-Communication and the third scale, which assesses students' sensory-cognitive awareness. These two scales constituted of a new variable, the "level in the cognitive domain" variable. After Cronbach's alpha reliability analysis, a high internal consistency was obtained with a = 0.85. The variable measured students' performance before and after kindergarten retention. The difference between the values of the "cognitive domain after retention-cognitive domain before retention" led to the variable "cognitive domain change". We followed the same process for the analysis of the social domain too. Specifically, for the study of students' social domain we used the second scale, which assesses sociability. Taking the final and initial measurements into consideration, that is, the results after and before the retention, we created a new variable that resulted from the difference of the two values "sociability after retention"-"sociability before retention", which was called "social domain change". At this point, concerning the sociability scale, it should be noted that the Cronbach's alpha reliability analysis showed a high internal consistency of a = 0.81.
As far as the change in the cognitive domain and the change in sociability are concerned, when the deviation is negative, there is improvement, while when the deviation is positive, it shows deterioration. Initially, a pilot study was conducted, in which 25 teachers participated, who had taught retained students. After utilizing the relevant observations and suggestions from the participants, the final research tool emerged.

Data Collection Procedure
The survey was conducted from September 2019 to June 2020 and all rules of conduct were adhered to, such as consent of the participants, assurance of anonymity and the possibility to withdraw from the survey any time they wished.
Participants were also informed in detail about how to complete the questionnaires. Data, concerning the students who are going to retain, were given to us by the Centers of Educational and Counseling Support. The questionnaires were completed by the special educators of the students who were retained. Of these, 93 were women (94.8%) and five were men (5.2%).
Of the women 10 (10.8%) were between 24 -30 years old, 39 (42%) between 31 -40 years old, 29 (31.2%) between 41 -50 years old and 15 (16%) between 50 -60 years old. Of the men one (20%) was between 20 -30 years old, three (60%) between 31 -40 years old and one (20%) between 41 -50 years old. As far as the Psychology special educators are concerned, 45% of them had a master's degree in special education, 30% had two years of professional training in special education, 20% had attended an annual specialization seminar in special education and 5% had Ph.D. in special education.

Statistical Analysis
The data analysis was based on quantitative methods of measurement. Mean values and standard deviation, as well as one sample-test were used to measure the quantitative variables. Subsequently, cluster analysis was conducted. The purpose of cluster analysis was to classify students into categories according to their response to retention. Thus, a hierarchical cluster analysis was conducted  Table 2 (Table 3). Therefore, as it is revealed, retention works positively for students with ASD both in the cognitive and social domains. It is clear, thus, that there has been a greater improvement in the cognitive domain of students with ASD.

Results
S. Loukeris, S.-G. Soulis Psychology RQ2 results. However, the study showed that the results of the retention are also related to the "type of help that students receive in the classroom". Participants were asked about the support provided to students. According to the answers, 46.3% of students (N = 51) had the help of public parallel support, 28.2% (N = 31) had the help of private parallel support, 15.5% (N = 17) attended the integration class, while 10% (N = 11) had no further support. To correlate the variables "student improvement" and "classroom support" the x 2 -independence test was applied. The result showed a statistically significant dependence on the two variables. In other words, student support in the classroom is an important factor in improving students [x 2 (2) = 25.78, p < 0.001/Fisher exact test: p = 0.001]. Specifically, of the students who were retained without any kind of support in the classroom, 81.8% had no improvement, while 18.2% showed improvement only in the cognitive field. Also, students who had public parallel support at 50.9% did not show a significant improvement in any domain, 23.5% had a significant improvement in both domains, while 25.6% showed improvement in the cognitive domain only. Of the students who had private parallel support, 16.1% showed a significant improvement in both domains, 66.3% had improvement in the cognitive domain, while 17.6% had no improvement in any domain. Finally, of the students who attended the integration class, 58.8% showed significant improvement in both domains. In conclusion, it was noted that the improvement of the students is directly related to the type of support they received. Of the three types of support provided in kindergarten, the most effective was the support provided by the integration class (Table 4).
RQ3 results. In addition, the impact of the afternoon supporting programs on the retention of students with ASD was studied. The x 2 -independence test shows a statistically significant dependence on the two variables. In other words, when students attend supporting programs after school, along with retention, there is a significant factor in student improvement [x 2 (2) = 32.88, p < 0.001]. Specifi-Psychology cally, when students attend additional extracurricular supporting programs, along with retention, there is a significant improvement in both domains corresponding to 34.4% and improvement in the cognitive domain at 35.6% (Table  5).
RQ4 results. To study whether the improvement of a retained student is related to whether he/she comes from a foreign family, the participants were asked about it. Their answers showed that 81.8% (N = 90) of the students came from Greek families while 18.2% (N = 20) came from foreign families. To decide the correlation between the variables "student improvement" and "student from a foreign family" the x 2 -independence test was applied. The result shows a statistically significant dependence on the two variables. Specifically, when students belong to foreign families, they do not improve [x 2 (2) = 23.78, p < 0.001]. In other words, when students belong to foreign families, there is no improvement in any domain. On the contrary, the native students who are retained show a significant improvement in both domains at a rate of 33.3%, and especially in the cognitive domain at a rate of 35.5% (Table 6).  RQ5 results. Furthermore, it was studied whether the educational attainment of the parents is related to the improvement of the students with ASD. Firstly, we created three groups. The first group includes parents who have graduated junior high school (nine school years), the second group includes parents who have graduated high school or technical school (12 school years) and the third group includes parents who have a bachelor or postgraduate studies.
Concerning the educational attainment of the fathers, the findings were the following: 5.5% (N = 6) are junior high school graduates, 12.7% (N = 14) are high school graduates, 12.7% (N = 14) technical school, 56.3% (N = 62) have a bachelor and 12.8% (N = 14) have completed postgraduate studies. Then, to decide the possible correlation between the variables "student improvement" and "educational attainment of the father", the x 2 -independence test was applied.
The result shows a statistically significant dependence of the two variables. The improvement of the students depends significantly on the educational attainment of the father [x 2 (4) = 61.67, p < 0.001]. Specifically, students whose father was a junior high school graduate did not improve in any domain at 100%. Students whose father has graduated high school had improvement in the cognitive domain at 17.9%. Finally, those students whose father has a higher education degree showed significant improvement in both domains by 42.1% (Table 7).
As far as the educational attainment of the mothers is concerned, it was observed that 8.2% (N = 9) are junior high school graduates, 29.1% (N = 32) are high school graduates, 49.1% (N = 54) have a bachelor and 13.6% (N = 15) have completed postgraduate studies. According to the statistical correlation test of the variables "student improvement" and "educational attainment of the mother", a significant dependence was found [x 2 (4) = 41.16, p < 0.001]. In other words, student improvement depends significantly on the mother's educational attainment. More specifically, students whose mother has graduated junior high school showed no improvement in any domain at 77.8%. Students whose mother Psychology has graduated high school did not improve in any domain at 78.1%. In contrast, those students whose mother has a higher education degree showed a significant improvement in both domains by 39.1% (Table 8).
The statistical analysis showed that retention had a positive effect on students with ASD. However, foreign students with ASD as well as students with ASD who come from parents with a low education background, do not seem to benefit much from retention. Similarly, students with ASD who do not have the opportunity to attend afternoon supporting programs have not benefited much from retention. At the same time, it was revealed that the most significant improvement regarding the retention is shown by the students with ASD who visit the integration class.

Discussion
The current study examined the impact of students' retention with ASD in their cognitive and social domain. As it emerged from the analysis of the results, the retention has a positive effect on both domains and depends on the type of education they receive at school, the extracurricular support provided to them, the educational attainment of their parents and if students come from a foreign family. The results of the present study, which demonstrate the improvement of students with ASD due to retention, in the cognitive and social domain are confirmed by the corresponding findings of the research of Heckman (2006) (2000) and Wu et al. (2010). However, this finding is disputed by the research of Brophy (1983), who conclude that retention has negative consequences for students' sociability as it affects their self-esteem and level of communication with their classmates. At this point, it should be noted that the findings of this study confirm the research hypotheses that in students with limited academic performance and social skills, such as most students with ASD, retention has positive results (Chen et al., 2014;Hong & Raudenbush, 2005;Renaud, 2013;Yang, 2018). It was also found that many students who are retained come from foreign families. This finding is consistent with the results of Jimerson et al. (1997) and Goldstein et al. (2014). Similarly, the current study found that a significant number of students who are retained belong to families with low educational backgrounds. This finding is also apparent in the research of Alexander et al. (2003) as well as Frey's (2005).

Conclusion
The research on students who are retained in kindergarten concludes with conflicting results. Many pieces of research highlight its positive benefits, others its negative impact on the mental health of the students, while a small number of studies show no difference in the performance of students who are retained compared to students who are promoted and study in the first grade of elementary school. However, in the present research, the positive effect of the retention in the cognitive and social domain of students with ASD is clearly noted. Of course, this finding does not apply to all students. Foreign students with ASD as well as students with ASD who come from parents with a low education background, do not seem to benefit much from retention. Similarly, students with ASD who do not have the opportunity to attend afternoon supporting programs have not benefited much from retention. At the same time, it was revealed that the most significant improvement regarding the retention is shown by the students with ASD who visit the integration class. Therefore, according to the above, Psychology it is necessary for the Greek State to support inclusive education and plan the appropriate educational interventions, so that retention can be beneficial to every student with ASD.
The present research has some limitations that should be mentioned. Since the sample is small, the results of the present study cannot be generalized. In addition, because of the relevance of the content of the tool used in this study it cannot be safely deduced that pupil improvement is due to retention or due to interaction with other factors such as student development and maturity or the social environment. Moreover, a thorough study of the results of retention is not feasible. This happens because in the current study there is no comparison with a corresponding group of children who attended elementary school, even though retention was proposed, but their parents did not give their consent.