Preschool Attendance,School Starting Age and Cultural-Financial Resources as Predictors of Adolescent’s Reading Comprehension: Evidence From PISA 2015 and 2018

Abstract

This study investigated whether there are significant differences between reading achievements of 15-year-old students and starting age for preschool and socioeconomic factors according to the results from the PISA 2015 and 2018. The quantitative results show that there is a clear and statistically significant difference between reading test scores of Turkish students in PISA 2015 and in PISA 2018 according to the variables starting age for preschool education, highest parental educational attainment and the number of books at home. Our quantitative analysis shows that students who attended kindergartens from age four have higher average test scores in reading skills than students who cared for only in homes, either by their parents or by others.

Keywords

preschool attendance reading skills preschool starting age socioeconomic variables and cultural influences

Introduction: Preschool Attendance and Adolescents’ Reading Comprehension

Many researchers believe that the first 5 years of the childhood play a formative role in early literacy skills, and that high-quality preschool experiences are strongly correlated with children’s later reading literacy achievements (Pullen & Justice, 2003). Large-scale, longitudinal studies are needed to verify the cause and effect relationship between preschool attendance and adolescents’ literacy attainment. A review of the literature has revealed that a number of studies were conducted on this subject matter in different countries. In a cohort study conducted with a sample size of 11,000 people in England, it was found that preschool education increased test scores moderately at the ages of 11, 14, and 16, and was more beneficial for children with disadvantaged socio-economic backgrounds (Apps et al., 2013). In another study conducted in the Eastern Asia and Pacific with a sample of4,712 children from 36 to 71 months living in Cambodia, China, Mongolia, and Vanuatu, the East Asia-Pacific Early Child Development Scales (EAP-ECDS) were used as the data collection tools. The findings demonstrated that the children in the control group who received early childhood education had significantly better cognitive, language and socio-emotional development compared to those who did not receive any early childhood education (Rao et al., 2019). The results of a study conducted in China with a final sample of7,710 observations for the seventh grade and7,074 observations for the ninth grade has revealed that preschool attendance is associated with a 0.163% increase in standard deviations for the cognitive scores of seventh-grade students while the gain is lower for the ninth-grade students (Zhang, 2017). A study conducted in Australia with a birth cohort of 5,107 infants and a kindergarten cohort of 4,983 four-years-old children provides evidence that preschool education can support cognitive development particularly for disadvantaged children by providing opportunities for learning through play, via directed learning and targeted teaching, and by providing supportive environments and interactions that support learning (Goldfeld et al., 2021). In line with these studies, a number of others have confirmed that the children with preschool attendance display a higher level of academic skills during adolescence compared to other children without preschool attendance (Magnuson et al., 2007; McCormick et al., 2021; Shapiro, 2021). Albeit rare, some studies using the retrospective method report that preschool attendance does not predict reading literacy when confounding factors such as family income, extracurricular reading, parental expectation of academic achievement and educational level, and self-education expectation are controlled. However, these studies show that it is necessary to consider the factors such as the low quality, accessibility and affordability of early childhood education; and more explicit conclusions can be reached through longitudinal studies and triangulation (Li et al., 2020).

The Right Age for Preschool

The best age to start preschool education is a topic of recurring debate among educators, educational scientists, and of course parents, as reflected in the array of school starting age policies internationally. Parents in particular are often ambivalent about the optimum age for children to start schooling and they want the best information available in order to make a final decision (Purtell et al., 2020). On one hand, they have an intense desire to find the safest and most effective way to impart children’s critical life skills and enhance their potentials in learning at nursery school or in kindergarten, while on the other hand they typically worry about their children’s possible socioemotional/behavior problems which associated with the psychological effects of parent-child separation or a tough transition (Mashburn et al., 2018; Vandesande et al., 2019) to the preschool classroom setting. Many parents make a concerted effort to maintain safe and secure attachment with their children. With this in mind, it is therefore important to crystallize an idea of the right age to enroll children on a preschool program. Given the emphasis on school starting age and early use of nonparental care in the previous literature, it was noteworthy that the very early intensive use of center-based care and education can be hazardous for a child’s social and emotional development, even if high quality care is provided (NICHD Early Child Care Research Network, 2006). A recent study has revealed that the early use of nonparental care is associated with negative sociobehavioral outcomes. The above research suggests that children who begin nonparental care earlier and in more formal settings showed low performance in cognitive outcomes and had worse scores on social and behavioral measures than did children who remained in parental care longer (Fram et al., 2012).

A number of infant and young child observations according to the Tavistock model demonstrate that many children experience painful separation feelings during adaptation to nursery school and caregivers appear unable to provide them with security and comfort. The observations also confirm that children receive scant emotional support in their struggle to overcome the painful experience of being separated from their parents (Battaglia et al., 2016; Czada, 2012; Datler et al., 2010; Frisch, 2012). To add to these findings, another study which focused on investigate the cortisol levels of toddlers during their transition to childcare suggested that children had elevated cortisol levels throughout transition, with the peak coming in the separation phase and the onset of childcare (particularly separation from parents) may be demanding for toddlers (Nystad et al., 2021). On the other hand, it should be emphasized that giving children a better start and high-quality pre-school provision at the right age provides a unique opportunity for lifelong learning. Thus far, many studies have focused on this issue. A British cohort study (n = ∼13,000) investigated the association between child care during infancy and later cognition while controlling for social selection and its findings revealed that children who benefit from early childhood programs show higher performance in cognitive outcomes than children who have not enrolled on a preschool program (Côté et al., 2013).

The results of many longitudinal studies carried out in Europe and the USA indicate that children who attended kindergartens as of age 4 have higher average test scores in mathematics and reading literacy than children who cared for only in homes, either by their parents or by others (Fram et al., 2012; Gordon et al., 2013; Temple & Reynolds, 2007). In another study in which the data from 9,185 children used were showed that early nonmaternal care was associated with higher achievement and lower behavior problem scores in childhood and adolescence (Jaffee et al., 2011). Similarly, a study of Latino children in the United States revealed that Latino children in center-based childcare improved over time in cognitive, language, and social skills, whereas children in family childcare stayed the same or lost ground in these areas over time (Ansari & Winsler, 2012).

Socioeconomic-Cultural Characteristics and Student Achievement

Educational achievement, and its relationship with socioeconomic and cultural background, is one of the enduring themes of educational studies. A long series of quantitative studies has found that different measures of socioeconomic and cultural characteristics are positively correlated with academic achievement and with educational attainment (Bhat et al., 2016; Cheadle, 2008; Cheung & Andersen, 2003; Farkas et al., 1990; Loboda et al., 2017; Perry & Mcconney, 2010). One should note that the relationship between a student’s socioeconomic background and their educational achievement seems enduring and substantial (Thomson, 2018).

Present Study

We conducted a secondary data analysis of 11,130 fifteen-year-old students who participated in PISA 2015 (n = 4731) and PISA 2018 (n = 6399) in Turkey to evaluate the possible impacts of preschool starting age and socioeconomic status for reading skills. Our aim of this study is to understand whether socioeconomic variables and cultural influences significantly predict the reading skills test scores of Turkish students in PISA 2015/2018 and discuss the results by comparing them over the years. What would be the ideal age to start preschool? To what extent do literacy environment at home, parent-child shared book reading and the number of children’s books at home affect academic success? We discuss these questions according to The PISA data set.

Specifically, this study addresses the following questions:

Do the reading scores of students differ significantly according to the preschool starting age?

Do the reading scores of students differ significantly in relation to parents’ education level?

Do the reading scores of students differ significantly according to number of books at home?

Do the variables of parental educational attainment, tangible and cultural family resources, information and communication technology predict the reading scores of students?

In order to successfully achieve the goals one-way analysis of variance (ANOVA) was used to determine whether there was a significant difference in the reading scores of the students according to the variables starting age for preschool education, the educational status of parents and the number of books at home.

Main Challenges in the Education System in Turkey and PISA

Many projects have been prepared and implemented in addition to policies, plans, targets and principles which have been defined and put into practice for improvement of the education system in Turkey; however, the expected results have not been achieved yet. The studies highlighted issues such as the high number of students in classrooms, insufficient number of teachers and pedagogues, and failure to provide equal means and opportunities in terms of access to education in Turkey (Kösterelioğlu & Bayar, 2014; Yaman, 2006). On the other hand, the insufficient share allocated to education from the central budget and the insufficient employment of teachers, along with the financing issues faced by schools, are among the other important problems experienced in the education system. Various studies have pointed out that the public resources allocated to education have increased in recent years, but Turkey still remains far behind the OECD average in terms of public spending per student (Arabacı, 2011; Kılıçaslan & Yavuz, 2019). In addition, Turkey still ranks at the bottom of the OECD country rankings in PISA studies within various categories, including reading skills in particular. Intense efforts are being made to overcome this problem and to ensure equal opportunities in terms of access to education, but no significant improvement has yet been achieved (Şener, 2018).

Turkey participated in PISA studies for the first time in 2003 in order to observe the impact of education policies on students, to make the education system more functional, and to increase the quality of education. PISA results are considered to provide important tips to identify the shortcomings in Turkish education system and benchmark with the other participating countries since the very first years of participation (Şener, 2018).

PISA reading skills data conducted and collected in 2015 and 2018 have been included in this study. Based on the review of the PISA results received in 2015, Turkey scored 428 points on average in the category of reading skills. While the general mean score of the countries participating in the project is 460 points, the average of the OECD countries is 493 points. Turkey is ranked 50th among 72 countries participating in the project and 34th among OECD countries that include 35 countries. Figure 1 provides a comparative presentation of the mean scores in reading category for the countries participating in PISA in 2015 and 2018.

Figure 1.

Ranking of national performance in reading.

Based on the review of the PISA results received in 2018, it can be seen that Turkey scored 466 points on average in the category of reading skills. While the general average score was 453 points for the participating countries, the average of OECD countries was 487 points. Turkey is ranked 40th among 79 countries participating in the PISA survey, and only 34th among 37 OECD countries. Turkey’s average scores in PISA reading skills across the years are shown in Figure 2 together with the average scores of OECD countries and of all countries.

Figure 2.

The average reading scores of Turkey, OECD, and all countries by years.

When the scores Turkey received in PISA 2015 and 2018 reading skills category are compared with the OECD average and the general average score of the participating countries, it is concluded that Turkey has ranked behind. Despite various efforts made in Turkey in the recent years, the inequality of opportunities in education still persists; and the socioeconomic conditions of families, the quality of teachers, and the social, cultural and economic conditions of the region stand out among the factors leading to this inequality (Güner et al., 2014).

Method

This study is a secondary data analysis of data collected by the PISA. Secondary data is data that has already been presented, and the secondary data analyst is not involved in the recruitment of participants or in the collection of the data (Andrews et al., 2012). Secondary data analysis is a research process used in various fields such as health sciences (Ekholuenetale et al., 2020; Kottner et al., 2020) and educational sciences (Logan, 2020; Wotto, 2020), which refers to the secondary analysis of existing data collected by other researchers or organizations. It should be emphasized that “the key to secondary data analysis is to apply theoretical knowledge to utilize existing data to address the research questions. Hence, the first step in the process is to develop the research questions” (Johnston, 2014).

Sample

The data used in this study were extracted from the PISA 2015 and PISA 2018 database. According to the PISA official website (http://www.pisa.oecd.org), participants were 15 years old and from Turkey. In 2015, 1,324,089 15-year-old students were educated in Turkey, however the accessible target population was 925,366 school students. For PISA 2015, schools were selected by the use of a stratified random sampling method in the Classification of Statistical Region Units Level 1 (MEB, 2015) and the students who were to participate were determined by random selection. At the end of the sampling process 187 different schools in 61 cities in Turkey participated in the exam (MEB, 2015). The sample consisted of the 5,895 students specifically, 2,938 females and 2,957 males (MEB, 2015). A total of 6,890 Turkish students from 186 different schools participated in PISA 2018 (MEB, 2019). Of these, 3,396 were female and 3,494 male, suggesting a balanced distribution in terms of gender (Erdem & Kaya, 2021).

One-Way ANOVA Test

In this study, one-way ANOVA test was used to take independent variables such as preschool starting age, parents’ level of education and the number of books at home as categorical variables and in order to analyze, in detail, the impact of the subcategories of these variables on reading skills and to demonstrate the relevant significant differences. Therefore, it was essential to review the variables though to impact students’ reading skills on an individual basis. However, the independent variables such as HISEI, WEALTH, CULTPOSS, HEDRES and ICTRES are index scores and constitute continuous variables. Multiple linear regression analysis was used to determine the power of such variables to predict students’ reading skills to a significant extent and to identify which variable was more important and study questions were prepared in line with the importance and purpose of the study.

Analysis

It should be noted that the outliers were detected using univariate and multivariate detection methods. Firstly, to identify univariate outliers, all variables were converted into Z-scores. Multivariate outliers were assessed by calculating each case’s the robust version of the Mahalanobis distance and a case is considered as a multivariate outlier if the probability associated with its D2 is 0.001 or less. Normality distribution was assessed using skewness and kurtosis. The test statistics and normality results are given in Table 1.

Table 1.

Test Statistics and Normality Results Obtained From PISA 2015 and PISA 2018.

	Variables	$\bar{X}$	Median	Mode(s)	$Skw .$	$Kur .$
PISA 2015	HISEI	36.78	30	12	0.832	−0.327
	WEALTH	−1.41	−1.40	−3.66	−0.089	−0.045
	CULTPOSS	−0.23	−0.13	−1.63	0.047	0.049
	HEDRES	−0.49	−0.72	1.16	0.086	−0.692
	ICTRES	−1.13	−1.08	−0.58	−0.053	−0.144
	PVREAD1	434.76	433.81	302.77	−0.053	−0.284
PISA 2018	HISEI	37.48	28.97	17.00	0.835	−0.753
	WEALTH	−1.36	−1.31	−1.04	−0.119	0.039
	CULTPOSS	−0.77	−0.82	−2.75	0.305	−0.116
	HEDRES	−0.46	−0.69	1.18	0.172	−0.586
	ICTRES	−1.10	−0.94	−0.62	−0.238	0.057
	PVREAD2	467.52	467.95	437.12	−0.023	−0.434

Note. HISEI = the highest level of parents’ occupation; WEALTH = an indicator of family wealth; CULTPOSS = a measure of family cultural possessions; HEDRES = an indicator of availability at home of educational resources; ICTRES = an indicator of information and communications technologies; PVREAD1-2 = reading achievement in PISA 2015/2018.

We have also tested whether or not there is a level of multicollinearity between the variables. In addition, tolerance, variance inflation factor (VIF) and condition index (CI) values were also examined. As shown in Table 2, our results further suggest that the tolerance values are greater than 0.10, the VIF values are less than 10 and the CI values are less than 30, and there is no multicollinearity between the variables.

Table 2.

Multicollinearity Test Results of Variables From PISA 2015 and PISA 2018.

		B	Std. error	Beta	Tolerance	VIF	CI
PISA 2015	Constant	434.131	3.249
	HISEI	0.630	0.055	.171	0.804	1.244	1.825
	CULTPOSS	6.032	1.294	.071	0.764	1.309	2.696
	HEDRES	0.066	1.252	.001	0.528	1.893	3.882
	ICTRES	23.471	2.492	.294	0.184	5.443	7.859
	WEALTH	−3.732	2.427	−.047	0.190	5.273	10.784
PISA 2018	Constant	451.620	2.998
	HISEI	0.943	0.047	.254	0.809	1.235	2.042
	CULTPOSS	8.447	0.954	.118	0.733	1.365	3.081
	HEDRES	−1.887	1.295	−.023	0.524	1.909	3.944
	WEALTH	−9.825	2.187	−.107	0.232	4.314	7.542
	ICTRES	24.695	2.267	.269	0.215	4.652	9.976

The missing values and outliers can drastically change the results of data analysis. The missing data were identified and extreme outliers were excluded from the analysis. The final analytic sample included 4,731 (PISA 2015) and 6,399 (PISA 2018) participants. Finally, one-way analysis of variance (One-Way ANOVA), A Pearson’s product-moment correlation coefficient and a multiple linear regressions analysis were used in the analysis of the data.

Findings

Possible Impacts of Preschool Starting Age on Reading Skills

One-way analysis of variance (ANOVA) was conducted in order to determine whether there are statistically significant differences between the reading skills of students according to the independent variable of preschool starting age. The results obtained from the PISA 2015 with a sample from Turkey are given in Table 3.

Table 3.

ANOVA Results According to the Variable of Preschool Starting Age According to PISA 2015 With a Sample From Turkey.

Age	n	$\bar{X}$	SD	Source of variance	Sum of squares	d.f	Mean square	F	p	df.
(1) ≤2	77	401.70	96.81	Between groups	269,702.069	5	53,940.414	8.083	.000	1–2 1–3 2–4 2–5 2–6
(2) 3	80	473.92	100.83	Within groups	5,198,344.079	779	6,673.099
(3) 4	148	448.93	91.26	Total	5,468,046.148	784
(4) 5	160	428.66	75.94
(5) 6≤	160	423.57	67.81
(6) Never	160	427.14	71.39

From the results shown in Table 3 we can observe that the scores of reading skills of the students in PISA 2015 differs significantly according to the age of starting preschool: $F (5, 784) = 8.083, p = . 000, η^{2} = . 049 .$ However, the Dunnett’s t test was also used to evaluate that differences among independent variable and determine which means differed significantly. Our analysis indicates that the average scores of reading skills of the students who started preschool education at the age of three was higher than the reading skills scores of the students who started preschool education at the age of two and younger. Our findings also reveal that the reading skills scores of students who started preschool education at aged 5 and 6 years were higher than the scores of students who had never received preschool education. On the other hand, it should be pointed out that the average scores of reading skills of students who started preschool education at the age of four was higher than the scores in mean reading scores of students who started preschool education at the age of two and below. Another statistical method used in the interpretation of the findings of one-way analysis of variance is η²(eta-square) effect size. Values of 0.01, 0.06, and 0.14 for eta-square are interpreted as small, medium, and large effect sizes, respectively. When the effect size of η²(.049) for the mentioned difference is evaluated, it can be said that the effect size is low.

Additionally, the results obtained from the PISA 2018 are given in Table 4.

Table 4.

ANOVA Results According to the Variable of Preschool Starting Age According to PISA 2018 With a Sample From Turkey.

Age	n	$\bar{X}$	SD	Source of variance	Sum of squares	d.f	Mean square	F	p	df.
(1) ≤2	107	447.62	95.35	Between groups	469,898.930	5	93,979.786	11.694	.000	1–2 1–3 2–4 2–5 2–6 3–6
(2) 3	152	514.07	96.78	Within groups	8,912,564.278	1109	8,036.577
(3) 4	214	490.09	96.74	Total	9,382,463.208	1114
(4) 5	214	478.52	93.21
(5) 6≤	214	466.14	77.95
(6) Never	214	454.50	80.96

The findings revealed a statistically significant difference between the reading scores of the students according to PISA 2018 and the school starting age: $F (5, 1114) = 11.694, p = . 000, η^{2} = . 050 .$ It can be stated that the η² effect size of the difference mentioned above is low. Moreover, the findings showed that the students who started preschool at 3 years old have higher levels of reading skills than the students who started preschool at 2 years old and younger. Our findings also suggest that students who started preschool at the age of four have higher levels of reading skills than children who cared for only in homes, either by their parents or by others, and children who started preschool at the age of two and below.

The students’ average scores in Reading Comprehension by their preschool starting ages are provided in Figure 3 on a comparative basis.

Figure 3.

Students’ mean scores in reading comprehension by their preschool starting ages, PISA 2015 to 2018.

Parents’ Education Level: Another Factor Influencing Reading Achievement?

One-way analysis of variance was further conducted to analyze the relationship between the reading skills of the students and the education level of the parents. The results obtained from the PISA 2015 are given in Tables 5 and 6.

Table 5.

ANOVA Results: The Impacts of Maternal Education Level According to PISA 2015.

Education level	n	$\bar{X}$	SD	Source of variance	Sum of squares	d.f	Mean square	F	p	df.
(1) None	563	417.98	67.90	Between groups	301,205.09	4	75,301.27	14.572	.000	1–2 1–3 1–4 1–5 2–4 2–5 3–4 3–5
(2) Primary school	1739	433.82	67.72	Within groups	24,421,538.17	4726	5,167.49
(3) Secondary school	856	431.12	71.00	Total	24,722,743.26	4730
(4) High school	709	442.62	71.63
(5) Bachelor, Master and Ph.D. degree	864	444.74	82.80

Table 6.

ANOVA Results: The Impacts of Paternal Education Level According to PISA 2015.

Education level	n	$\bar{X}$	SD	Source of variance	sum of squares	d.f	Mean square	F	p	df.
(1) None	237	413.71	69.01	Between Groups	911,596.17	4	227,899.04	45.233	.000	1–4 1–5 2–4 2–5 3–4 3–5
(2) Primary school	1384	424.58	66.36	Within Groups	23,811,147.09	4726	5,038.33
(3) Secondary school	1188	424.07	70.12	Total	24,722,743.26	4730
(4) High school	668	452.69	70.17
(5) Bachelor, Master and Ph.D. degree	1254	450.55	77.25

The results, given in Table 5, show that the average scores on reading skills in PISA 2015 differed significantly depending on the education level of the mother: $F (4, 4730) = 14.572, p = . 000, η^{2} = . 012 .$ Results of the Dunnett’s C test indicated that the increases in maternal education will improve their children’s reading achievement. The main findings also suggest that the average scores on reading skills of the students whose mothers had no formal education and did not graduate from primary school are lower than the scores of students whose mothers had attained primary, secondary, and higher education, and had have postgraduate qualifications such as a master’s degree or a Ph.D.

Our results also demonstrated that the average scores on reading skills in PISA 2015 differed significantly depending on the education level of the father: $F (4, 4730) = 45.233, p = . 000, η^{2} = . 036 .$ Similarly, the father’s level of education might ultimately have an impact on reading achievement. Our findings reveal that the average scores on reading skills of the students whose fathers had a high school education or had attained a bachelor’s, master’s or Ph.D. degree are higher than the students whose father had no formal education or had primary and secondary school education.

The results obtained from the PISA 2018 are given in Tables 7 and 8.

Table 7.

ANOVA Results: The Impacts of Maternal Education Level According to PISA 2018.

Education level	n	$\bar{X}$	SD	Source of variance	Sum of squares	d.f	Mean square	F	p	df.
(1) None	655	440.42	75.43	Between groups	927,266.63	4	231,816.66	34.474	.000	1–2 1–3 1–4 1–5 2–5 3–5 4–5
(2) Primary school	1768	465.23	78.26	Within groups	42,996,059.93	6394	6,724.44
(3) Secondary school	1287	466.16	73.83	Total	43,923,326.56	6398
(4) High school	1170	465.78	84.96
(5) Bachelor, Master and Ph.D. degree	1519	484.36	92.55

Table 8.

ANOVA Results: The Impacts of Paternal Education Level According to PISA 2018.

Education level	n	$\bar{X}$	SD	Source of variance	Sum of squares	d.f	Mean square	F	p	df.
(1) None	238	438.27	78.90	Between groups	1,891,601.57	4	472,900.39	71.939	.000	1–4 1–5 2–4 2–5 3–4 3–5 4–5
(2) Primary school	1359	451.01	77.96	Within groups	42,031,724.98	6394	6,573.62
(3) Secondary school	1606	452.49	76.20	Total	43,923,326.56	6398
(4) High school	1135	478.56	76.86
(5) Bachelor, Master and Ph.D. degree	2061	487.42	88.92

The results suggest that the scores of reading skills of the students in PISA 2018 differs significantly according to the maternal education level: $F (4, 6398) = 34.474, p = . 000, η^{2} = . 021 .$ In each case, Dunnett’s t-test indicated that the increases in maternal education will improve their children’s reading achievement and the results are in very good agreement with our previous analysis of PISA 2015. It can be concluded from the results that the average scores on reading skills of the students whose mothers did not graduate from primary school are lower than the scores of students whose mothers had attained primary, secondary, higher education and had have masters or doctorates etc. It should also be emphasized that the students whose mothers had attained a Bachelor’s, Master’s, or Ph.D. degree have higher reading scores than the students with lower maternal education levels.

As can be seen in Table 8, the scores of reading skills of the students in PISA 2018 differs significantly according to the fathers education level: $F (4, 6398) = 71.939, p = . 000, η^{2} = . 043 .$ In parallel with our findings of PISA 2015 dataset, the results of the Dunnett’s C test indicated that the students whose fathers had attained a Bachelor’s, Master’s or Ph.D. degree have higher reading scores than the students with lower father education levels. It should be noted that the η² effect size of the differences is low.

The Importance of the Number of Books at Home

One-way analysis of variance (ANOVA) tests were performed to ascertain whether there are significant differences between reading scores and number of books at home. The results obtained from the PISA 2015 and 2018 are given in Tables 9 and 10.

Table 9.

ANOVA Results: Number of Books at Home According to PISA 2015.

Number of books	n	$\bar{X}$	SD	Source of variance	Sum of squares	d.f	Mean square	F	p	df.
(1) 0–10	354	395.00	62.69	Between groups	1,873,504.29	5	374,700.86	73.898	.000	1–2 1–3 1–4 1–5 1–6 2–3 2–4 2–5 2–6 3–5 3–6 4–5
(2) 11–25	354	410.53	72.12	Within group	9,654,276.90	1904	5,070.52
(3) 26–100	354	444.77	68.54	Total	11,527,781.19	1909
(4) 101–200	354	459.61	70.84
(5) 201–500	317	482.84	74.71
(6) >500	177	466.41	83.84

Table 10.

ANOVA Results: Number of Books at Home According to PISA 2018.

Number of books	n	$\bar{X}$	SD	Source of variance	Sum of squares	d.f	Mean square	F	p	df.
(1) 0–10	524	415.87	71.97	Between groups	4,539,108.05	5	907,821.61	150.621	.000	1–2 1–3 1–4 1–5 1–6 2–3 2–4 2–5 2–6 3–4 3–5 3–6 4–5 4–6
(2) 11–25	524	438.06	74.15	Within group	17,334,162.13	2876	6,027.18
(3) 26–100	524	472.70	73.50	Total	21,873,270.19	2881
(4) 101–200	524	498.42	80.67
(5) 201–500	524	521.65	79.76
(6) >500	262	521.38	91.69

The results, given in Table 9, show that the scores of reading skills in PISA 2015 differs significantly according to the variable number of books at home: $F (5, 1909) = 73.898, p = . 000, η^{2} = . 163 .$ Additionally, Dunnett’s C test was carried out to verify significant differences between the variables. The test indicated that the number of books in the homes of low-reading groups was low, and as the number of books in the home increased, reading success also increased. The average reading scores of students who have more than 500 books are higher than the students who have 26 to 100, 11 to 25, and 0 to 10 books in the house. The scores of students who have 201 to 500 books are higher than the students who have 101 to 200, 26 to 100, 11 to 25, and 0 to 10 books at home.

Similarly, we found that the scores of reading skills in PISA 2018 differs significantly according to number of books around the house: $F (5, 2881) = 150.621, p = . 000, η^{2} = . 208 .$ Our analysis has shown that the number of books does increase student performance (reading scores) in PISA. In addition, there was no difference found between groups with having “more than 500 books” and “around 201–500 books” at home. Furthermore, the findings showed that there was a significant difference between having more books and having few books in the home and it should be noted that effect size of the differences is large.

Other Family Background Variables

For purposes of this study, family background variables include variables related to the highest level of parents’ occupation (HISEI); an indicator of family wealth (WEALTH); a measure of family cultural possessions (CULTPOSS); an indicator of availability at home of educational resources (HEDRES); an indicator of information and communications technologies (ICTRES) categorized into a five-level. A Pearson’s product-moment correlation coefficient was calculated to determine whether there is a significant relationship between reading achievement in PISA 2015 (PVREAD1) and the continuous variables HISEI, WEALTH, CULTPOSS, HEDRES and ICTRES. Table 11 illustrates the results:

Table 11.

Correlation Results Between Variables According to PISA 2015.

Variables	1	2	3	4	5	6
(1) HISEI	-
(2) WEALTH	.42**	-
(3) CULTPOSS	.28**	.42**	-
(4) HEDRES	.32**	.63**	.43**	-
(5) ICTRES	.41**	.89**	.38**	.66**	-
(6) PVREAD1	.29**	.32**	.21**	.25**	.35**	-

p < .01.

As shown in Table 11, positive but weak correlations were found between the variables HISEI and CULTPOSS, HISEI and PVREAD1, CULTPOSS and PVREAD1, and HEDRES and PVREAD1. There were significant but moderate correlations between HISEI and WEALTH, HISEI and HEDRES, HISEI and ICTRES, WEALTH and CULTPOSS, WEALTH and HEDRES, WEALTH and PVREAD1, CULTPOSS and HEDRES, CULTPOSS and ICTRES, HEDRES and ICTRES, and ICTRES and PVREAD1. Notably, there was a strong, significant, positive correlation between WEALTH and ICTRES.

A multiple linear regressions analysis (Enter Method) were done to detect whether or not the variables the highest level of parents’ occupation, family wealth, family cultural possessions, information and communications technologies, educational resources at home predict the reading achievement. The results are given in Table 10.

As shown in Table 12, the variables of the “availability at home of educational resources” (HEDRES) and the family wealth (WEALTH) did not significantly predict the reading scores of students p > .05. The results obtained from the PISA 2015 have provided evidence that the variables of highest level of parents’ occupation (HISEI), family cultural possessions (CULTPOSS) and the information and communications technologies (ICTRES) did significantly predict the reading scores of students. The resulting standardized regression coefficient (β) indicate that ICTRES, HISEI and CULTPOSS are the most effective variables in predicting the reading achievement of students. The indicator of ICTRES was the most important predictor variable for reading scores of students. The direct effect of this variable explains 15% of the reading skills variance: R = .39, F(5,4730) = 171.122, p = .000.

Table 12.

Results of Multiple Linear Regression Analysis for Reading Scores in PISA 2015.

Variables	B	Std. error (B)	β	t	p	VIF
Constant	434.131	3.249		133.622	.000
HISEI	0.630	0.055	.171	11.462	.000	1.244
CULTPOSS	6.032	1.294	.071	4.660	.000	1.309
HEDRES	0.066	1.252	.001	0.053	.958	1.893
ICTRES	23.471	2.492	.294	9.419	.000	5.443
WEALTH	−3.732	2.427	−0.047	−1.538	.124	5.273

Other findings revealed a positive and weak correlations between reading achievement in PISA 2018 (PVREAD2) and the continuous variables WEALTH and PVREAD2; CULTPOSS and PVREAD2; HEDRES and PVREAD2. Table 13 demonstrate the results:

Table 13.

Correlation Results Between Variables According to PISA 2018.

Variables	1	2	3	4	5	6
(1) HISEI	-
(2) WEALTH	.42**	-
(3) CULTPOSS	.30**	.47**	-
(4) HEDRES	.30**	.60**	.44**	-
(5) ICTRES	.38**	.87**	.43**	.67**	-
(6) PVREAD2	.34**	.27**	.25**	.22**	.31**	-

p < .01.

Significant positive (but moderate in size) correlations were found between HISEI and CULTPOSS; HISEI and WEALTH; HISEI and HEDRES; HISEI and ICTRES, HISEI and PVREAD2, WEALTH and CULTPOSS; WEALTH and HEDRES; CULTPOSS and HEDRES; CULTPOSS and ICTRES; HEDRES and ICTRES; ICTRES and PVREAD2. Additionaly, there was a significant, positive correlation between WEALTH and ICTRES.

As shown in Table 14, the variable of the “availability at home of educational resources” (HEDRES) did not significantly predict the reading scores of students p > .05. Our results have provided evidence that the variables of highest level of the parents’ occupation (HISEI), family cultural possessions (CULTPOSS), information and communications technologies (ICTRES) and the family wealth did significantly predict the reading scores of students. The resulting standardized regression coefficient (β) suggest that ICTRES, HISEI, CULTPOSS and WEALTH are the most effective variables in predicting the reading achievement of students. It was observed that the indicator of ICTRES was the most important predictor variable for reading scores of students in PISA 2018. The direct effect of the variable explains 17% of the reading skills variance: R = .41, F(5,6398) = 253.158, p = .000.

Table 14.

Results of Multiple Linear Regression Analysis for Reading Scores in PISA 2018.

Variables	B	Std. error (B)	β	t	p	VIF
Constant	451.620	2.998		150.666	.000
HISEI	0.943	0.047	.254	20.034	.000	1.235
CULTPOSS	8.447	0.954	.118	8.857	.000	1.365
HEDRES	−1.887	1.295	−.023	−1.457	.145	1.909
ICTRES	24.695	2.267	.269	10.894	.000	4.652
WEALTH	−9.825	2.187	−.107	−4.493	.000	4.314

Presentation of Variables and Results With a Regression Analysis

The enter method of multiple linear regression analysis was used on the sample of the PISA 2015 and PISA 2018 Turkey to determine whether the variables such as either parent’s highest occupational status, the family’s financial wealth, cultural possessions, information and communication technology resources and educational resources at home significantly predict reading skills. The results obtained are provided in Table 15.

Table 15.

Multiple Linear Regression Analysis Results for PISA 2015 to 2018 Reading Skills.

Variables	B	Standard error (B)	β	t	p	VIF
Fixed	482.354	1.246		387.164	.000
HISEI	−0.004	0.000	−.101	−11.875	.000	1.034
CULTPOSS	4.917	0.752	.063	6.538	.000	1.328
HEDRES	3.717	0.884	.048	4.205	.000	1.876
ICTRES	27.186	1.386	.327	19.610	.000	3.954
WEALTH	−5.709	1.445	−.066	−3.950	.000	3.947

Note. HISEI = the family’s highest occupational status; WEALTH = the family’s financial wealth; CULTPOSS = the family’s cultural possessions; HEDRES = home educational resources; ICTRES = information and communication technologies resources.

From the review of the values provided in Table 15, it was determined all independent variables significantly predicted the reading skills, p = .000. According to the standardized regression coefficients (β) the order of significance of the variables predicting reading skills are as follows: information and communication technology resources, cultural possessions, home educational resources, the family’s financial wealth, and the family’s highest occupational status. It was identified that the variable of information and communication technology resources was the most important predictor variable. It was determined that these predictor variables explained 13% of the variance in reading skills; R = .36, F (5,12360) = 370.580, p = .000.

Discussion and Conclusions

As could be shown using one-way analysis of variance (ANOVA) and multiple linear regressions analysis, a range of factors contributed positively to reading achievement among Turkish students that had been assessed as part of the PISA project. This study demonstrates that the indicators of preschool starting age, the highest level of parents’ occupation, parents’ education level, number of books at home, family cultural possessions and information and communications technologies around the house can predict the reading achievement of students. In addition, the multiple linear regression method was used to determine whether (or not) reading skills scores from PISA 2015 and 2018 were significantly predicted by the indexes (variables) of highest parental occupation status, highest parental educational attainment, tangible assets, economic and cultural resources at home, information and communication technology. The findings reveal that reading skills scores were significantly predicted by the indexes of highest parental occupation status, cultural resources at home and information and communication technology.

The findings further revealed that the reading achievement of students who started preschool education at aged 5 and 6 years were higher than the scores of students who had never received preschool education and students who attended kindergarten as of age 4 have higher average test scores in reading skills than students who cared for only in homes, either by their parents or by others. Several of our findings appear consistent with past research especially with the results of many longitudinal studies (Ansari & Winsler, 2012; Côté et al., 2013; Fram et al., 2012; Gordon et al., 2013; Jaffee et al., 2011; Temple & Reynolds, 2007). Our other results indicated that the increases in maternal and father education will improve their children’s reading achievement and the variables of information and Communications Technologies (ICTRES), highest level of parents’ occupation (HISEI), family cultural possessions (CULTPOSS) are the most effective variables in predicting the reading achievement of students. A long series of quantitative studies has found that different measures of socioeconomic and cultural characteristics are positively correlated with academic achievement and with educational attainment. Our results support the arguments made by past researchers (Bhat et al., 2016; Cheadle, 2008; Cheung & Andersen, 2003; Farkas et al., 1990; Loboda et al., 2017; Perry & Mcconney, 2010; Thomson, 2018).

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Yunus Pınar

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.

References

Andrews

Higgins

Andrews

M. W.

Lalor

J. G.

(2012). Classic grounded theory to analyse secondary data: Reality and reflections. Grounded Theory Review, 11(1), 12–26.

Ansari

Winsler

(2012). School readiness among low-income, Latino children attending family childcare versus centre-based care. Early Child Development and Care, 182(11), 1465–1485.

Apps

Mendolia

Walker

(2013). The impact of pre-school on adolescents’ outcomes: Evidence from a recent English cohort. Economics of Education Review, 37, 183–199.

Arabacı

İ. B

. (2011). Türkiye’de ve OECD ülkelerinde eğitim harcamaları (Educational Expenditures in Turkey and OECD countries). Elektronik Sosyal Bilimler Dergisi, 10(35), 100–112.

Battaglia

Touchette

É.

Garon-Carrier

Dionne

Côté

S. M.

Vitaro

Tremblay

R. E.

Boivin

(2016). Distinct trajectories of separation anxiety in the preschool years: Persistence at school entry and early-life associated factors. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 57, 39–46. https://doi.org/10.1111/jcpp.12424

Bhat

Joshi

Wani

(2016). Effect of socio economic status on academic performance of secondary school students. International Journal of Indian Psychology, 3(4), 2349–2394.

Cheadle

J. E.

(2008). Educational Investment, family context, and children’s math and reading growth from kindergarten through the third grade. Sociology of Education, 81(1), 1–31.

Cheung

S. Y.

Andersen

(2003). Time to read: Family resources and educational outcomes in Britain. Journal of Comparative Family Studies, 34(3), 413–433.

Côté

S. M.

Doyle

Petitclerc

Timmins

(2013). Child care in infancy and cognitive performance until middle childhood in the millennium cohort study. Child Development, 84(4), 1191–1208. https://doi.org/10.1111/cdev.12049

10.

Czada

(2012). Ein Kind voller Unruhe und Rastlosigkeit –Über die Schwierigkeiten der Eingewöhnung eines Kleinkindes in die Kinderkrippe aufgrund des Einflusses ambivalenter Gefühle der Bezugs- und Betreuungspersonen. Eine Einzelfallstudie über das Erleben von Trennung und GetrenntSein eines Krippenkindes [Master thesis, University of Vienna].

11.

Datler

Funder

(2010). Struggling against a feeling of becoming lost: A young boy’s painful transition to day care. Infant Observation, 13(1), 65–87. https://doi.org/10.1080/13698031003606659

12.

Ekholuenetale

Onikan

Ekholuenetale

C. E.

(2020). Prevalence and determinants of mother and newborn skin-to-skin contact in the Gambia: A secondary data analysis. The Journal of the Egyptian Public Health Association, 95(1), 18–19. https://doi.org/10.1186/s42506-020-00050-1

13.

Erdem

Kaya

(2021). Socioeconomic status and wellbeing as predictors of students’ academic achievement: Evidence from a developing country. Journal of Psychologists and Counsellors in Schools, 31, 1–19. https://doi.org/10.1017/jgc.2021.10

14.

Farkas

Grobe

R. P.

Sheehan

Shuan

(1990). Cultural Resources and School Success: Gender, ethnicity, and poverty groups within an Urban School District. American Sociological Review, 55, 127–142.

15.

Fram

M. S.

Kim

Sinha

(2012). Early care and prekindergarten care as influences on school readiness. Journal of Family Issues, 33(4), 478–505.

16.

Frisch

(2012). Kinder müssen loslassen – Mütter aber auch! Die Bedeutung der Trennung für das Kind und die Mutter im Eingewöhnungsprozess in die Kinderkrippe [Master thesis, University of Vienna].

17.

Goldfeld

Moreno-Betancur

Guo

Mensah

O’Connor

Gray

Chong

Woolfenden

Williams

Kvalsvig

Badland

Azpitarte

O’Connor

. (2021). Inequities in children’s reading skills: The role of home reading and preschool attendance. Academic Pediatrics, 21(6), 1046–1054.

18.

Gordon

R. A.

Colaner

Usdansky

M. L.

Melgar

(2013). Beyond an “Either-Or” approach to home- and center-based child care: Comparing children and families who combine care types with those who use just one. Early Childhood Research Quarterly, 28(4), 918–935. https://doi.org/10.1016/j.ecresq.2013.05.007

19.

Güner

Çelebi

Kaya

G. T.

Ve Korumaz

(2014). Neoliberal eğitim politikaları ve eğitimde fırsat eşitliği bağlamında uluslararası sınavların (PISA, TIMSS ve PIRLS) analizi (Analysis of international examinations (PISA, TIMSS and PIRLS) in the context of neoliberal educational policy and educational equity). Journal of History Culture and Art Research, 3(3), 33–75.

20.

Jaffee

S. R.

Van Hulle

Rodgers

J. L.

(2011). Effects of nonmaternal care in the first 3 years on children’s academic skills and behavioral functioning in childhood and early adolescence: A sibling comparison study. Child Development, 82(4), 1076–1091.

21.

Johnston

M. P.

(2014). Secondary data analysis: A method of which the time has come. Qualitative and quantitative methods in libraries, 3(3), 619–626.

22.

Kösterelioğlu Bayar

. (2014). Türk eğitim sisteminin sorunlarına ilişkin güncel bir değerlendirme (Evaluation of problems in the Turkish Education System). The Journal of Academic Social Science Studies, 25(1), 177–187. https://doi.org/10.9761/JASSS2279

23.

Kottner

Everink

van Haastregt

Blume-Peytavi

Schols

(2020). Prevalence of intertrigo and associated factors: A secondary data analysis of four annual multicentre prevalence studies in the Netherlands. International Journal of Nursing Studies, 104, 1–7.

24.

Kılıçaslan

Yavuz

(2019). PISA sonuçları ile Türkiye’de eğitim harcamaları ilişkisi (Relationship between PISA results and education expenditures in Turkey). Bilgi Sosyal Bilimler Dergisi, 21(2), 296–319.

25.

Chen

(2020). Effects of attending preschool on adolescents’ reading literacy: Evidence from the ethnic minority children in China. Children and Youth Services Review, 116, 105211.

26.

Loboda

Vogelbacher

Gawlitzek

(2017). The role of ethnic differences, structural background and process characteristics in the family in preschool children’s language proficiency. Journal of Multilingual and Multicultural Development, 38(6), 558–572.

27.

Logan

(2020). A practical, iterative framework for secondary data analysis in educational research. The Australian Educational Researcher, 47(1), 129–148. https://doi.org/10.1007/s13384-019-00329-z

28.

Magnuson

K. A.

Ruhm

Waldfogel

(2007). The persistence of preschool effects: Do subsequent classroom experiences matter? Early Childhood Research Quarterly, 22(1), 18–38.

29.

Mashburn

A. L.

LoCasale-Crouch

Pears

K. C.

(2018). Kindergarten transition and readiness: Promoting cognitive, social-emotional, and self-regulatory development. Springer.

30.

McCormick

Weiland

Hsueh

Pralica

Weissman

A. K.

Moffett

Snow

Sachs

(2021). Is skill type the key to the prek fadeout puzzle? Differential associations between enrollment in preK and constrained and unconstrained skills across kindergarten. Child Development, 92, 599–620.

31.

MEB. (2015). Turkish Ministry of National Education. PISA 2015 National Report. Milli Eğitim Bakanliği.

32.

MEB. (2019). Turkish Ministry of National Education. PISA 2018 National Preliminary Report. Milli Eğitim Bakanliği.

33.

NICHD Early Child Care Research Network. (2006). Child-care effect sizes for the NICHD study of Early Child Care and youth development. The American Psychologist, 61(2), 99–116.

34.

Nystad

Drugli

M. B.

Lydersen

Lekhal

Buøen

E. S.

(2021). Toddlers’ stress during transition to childcare. European Early Childhood Education Research Journal, 29(2), 157–182.

35.

Perry

L. B.

Mcconney

(2010). Does the SES of the school matter? An examination of socioeconomic status and student achievement using PISA 2003. Teachers College Record, 112(4), 1137–1162.

36.

Pullen

P. C.

Justice

L. M.

(2003). Enhancing phonological awareness, print awareness, and oral language skills in preschool children. Intervention in School and Clinic, 39(2), 87–98.

37.

Purtell

K. M.

Valauri

Rhoad-Drogalis

Jiang

Justice

L. M.

Lin

T. J.

Logan

J. A. R.

(2020). Understanding policies and practices that support successful transitions to kindergarten. Early Childhood Research Quarterly, 52, 5–14.

38.

Rao

Richards

Sun

Weber

Sincovich

(2019). Early childhood education and child development in four countries in East Asia and the Pacific. Early Childhood Research Quarterly, 47, 169–181.

39.

Şener

(2018). Türkiye’de Yaşanan Eğitim Sorunlarına Güncel bir Bakış (A current look at educational problems in Turkey). Milli Eğitim Dergisi, 47(218), 187–200.

40.

Shapiro

(2021). The benefits of prekindergarten programs: Strong findings and open questions. Phi Delta Kappan, 103(2), 8–13.

41.

Temple

J. A.

Reynolds

A. J.

(2007). Benefits and costs of investments in preschool education: Evidence from the child–parent centers and related programs. Economics of Education Review, 26, 126–144. https://doi.org/10.1016/j.econedurev.2005.11.004

42.

Thomson

(2018). Achievement at school and socioeconomic background-an educational perspective. NPJ Science of Learning, 3, 5. https://doi.org/10.1038/s41539-018-0022-0

43.

Vandesande

Bosmans

Maes

(2019). Can I be your safe haven and secure base? A parental perspective on parent-child attachment in young children with a severe or profound intellectual disability. Research in developmental disabilities, 93, 103452.

44.

Wotto

(2020). The future high education distance learning in Canada, the United States, and France: Insights from before COVID-19 secondary data analysis. Journal of Educational Technology Systems, 49(2), 262–281.

45.

Yaman

(2006). Eğitim sistemindeki sorunlardan bir boyut: Büyük sınıflar ve sınıf yönetimi (One dimension of problems in the education system: large classrooms and classroom management). Gazi Üniversitesi Türk Eğitim Bilimleri Dergisi, 4(3), 261–274.

46.

Zhang

(2017). Effects of attending preschool on adolescents’ outcomes: Evidence from China. Applied Economics, 49(27), 2618–2629.