Teacher Rating Scales of Early Academic Competence (TRS-EAC): Adaptation to Turkish, Validity and Reliability

This study aims to conduct Turkish adaptation, validity and reliability studies of the Teacher Rating Scales of Early Academic Competence. This study also aims to examine and improve early academic competence thoroughly on the basis of teachers’ ratings in Turkey, and lastly to contribute an assessment and measurement instrument to the literature which will enable to conduct further research in the field. This study was designed in a general survey model. The study included normally developing 619 36-72 month-old preschoolers attending to nursery classes and kindergartens affiliated to Afyonkarahisar Ministry of National Education in 2015-2016 academic year. The validity and reliability studies of the data obtained following the implementation of TRSEAC, was completed. The Cronbach’s Alpha, split half test, test-retest reliability of the scale and its subscales were measured. TRSEAC was found to be valid and reliable for 36-72 month-old children. It can be considered that the various studies that can be used TRS-EAC in Turkey will provide important contributions to the measurement power of this test.


Introduction
With technological and scientific changes and developments in the world, expectations in education are also changing. These developments are led by globalization. Today, all these changes also influence rapid living conditions. Thus, there is a need for qualified individuals with properties such as academic success, creative and critical thinking, high-level thinking skills, and high level academic competency (Ulusoy, 2003;Charlesworth, Lind & Fleege, 2003). For qualified individuals, these individuals should be supported during early childhood period when advances in development are the highest. It is known that during pre-school period that is accepted as magical years in terms of development, academic skills can be effectively gained (Krajewski & Schneider, 2009). Additionally, pre-school educational institutions support social development of children in this period and contribute development of skills of children such as sharing, cooperating, problem solving, decision-making. Gaining all these skills supports early period academic competence gains in children (Kokko, Tremblay, Lacourse, Nagin & Vitaro, 2006;Veenstra et al., 2008).
Academic competence concept was introduced by Diperna and Elliot (1999). Based on an elementary school study of Diperna and Elliot (1999), academic competence concept should be emphasized to distinguish pre-school academic skills and other inner-class behaviours that contribute to inner-class success. Inner-class behaviours here were named as "attitudes providing academic success" and related with academic skills (Diperna & Elliot, 2004). Diperna and Elliot (1999) surveyed large scale of skills, attitudes, and behaviours that may indicate pre-school children academic skills. For academic competence field, five fundamental sections were determined by National Education Goals Panel and were considered (Kagan, Moore & Bredekamp, 1995). There fields were: (a) physical competence and motor development, (b) social and emotional development, (c) approach towards learning, (d) language development, and (e) cognitive and general knowledge accumulation (Diperna & Elliot, 1999).
In short, early childhood academic competence concept consists of combination of academic skills and gaining attitudes for academic success (DiPerna, 2004;Rimm-Kaufman, Pianta & Cox, 2000). Early academic skills include early literacy, comprehension, numeracy, creative, and critical thinking skills (DiPerna, 2004;Rimm-Kaufman, Pianta & Cox, 2000;Welsh, Nix, Blair, Bierman & Nelson, 2010). Early literacy skills include oral language skills, pronunciation, alphabet Numeracy skills include comparison, counting, numerical logic, geometry, piece-whole, classification, measurement, sequencing, and prediction skills (Purpura, 2009;Kandir, Can Yasar, Yazici, Turkoglu & Yaman Baydar, 2016). As in other skills, development of numeracy skills rapidly develops during pre-school period. When numeracy skills gained in this period, it creates a foundation for mathematic learning in the following years and help children to know the world better (Linder, Powers-Costello & Stegelin, 2011;Wood & Spelke, 2005). With such numeracy skills children learn piece-whole relationship, comparison, distinguishing similarities and differences, using tables and graphics, interpreting, and problem solving (Celik, 2012). The key concept in acquisition of numeracy skills is numbers. Numeracy skill gain occurs with acquisition of three elements: number, numerical relationships, and arithmetic operations. The concept of numbers includes counting sequence information, counting principles, total number of items in a cluster, pairwise mapping; numerical relations and numeric information compare, and matching the relation of two or more entities to each other; arithmetic relations include understanding and comparing quantitative or quantitative changes, acquiring, predicting and evaluating skills such as addition and subtraction to transition to simple numeracy operations, understanding the patterns, continuing from the pattern, and creating new patterns (Jordan, Kaplan, Olah & Locuniak, 2006;Jordan, Kaplan, Ramineni & Locuniak, 2009).
Critical thinking is evaluating ideas and actions of others. Children gain critical thinking throughout their lives. Children will learn not to except all information as it is with their own experience, learn to interpret, evaluate, and constantly develop different perspectives (Heyman, 2008). Since development of critical thinking skills influence reading, reading comprehension, commenting, and numeracy skills, and these skills start to develop in early childhood period, these are included under early childhood academic skills. Since creative thinking skills influence academic success in addition to critical thinking, creative thinking has an important role under academic skills. Creative thinking like critical thinking increases as the interaction of children with the environment increases. Thus, early childhood period is critical for development of creative thinking. Creative thinking is creating new and authentic ideas by using information obtained via observation and experiences. An individual who can think creatively can filter when s/he faces a knowledge with large perspective, organize these information, and create new information due to cognitive flexibility (Gelman & measurement tools for ability to measure different skills, language development (Katz, Onen, Demir, Uzlukaya & Uludag, 1974;Ege, Acarlar & Guleryuz, 1998;Sucuoglu, Buyukozturk & Unsal, 2008), mathematics development (Aslan, 2004;Celik & Kandir, 2011), literacy skills (Yazici, 2010Turan & Akoglu, 2011;Karaman, 2013;Kargin, Ergul, Buyukozturk & Guldenoglu, 2015), academic and linguistic skills (Uyanik & Kandir, 2014), learning approaches (McWayne et al., 2004), social competences (McKee-Agostin & Bain, 1997), emotional competence (Howse, Calkins, Anastopoulos, Keane & Shelton, 2003), and self-control (Normandeau & Guay, 1998) of children.
When these scales were examined, measurement tools that measure different sub-skills of academic competence of children, it is striking that there was a lack of holistic measurement tool that analyze academic success of children with teacher evaluation. Besides, when the academic competence of children was evaluated by using experience and knowledge gained by children after spending a long time in their classes, it was believed that teachers could make a reliable evaluation to determine strong and weak sides of children and report these sides (Reid, Diperna, Missal & Volpe, 2014). In addition to all these information, development is a whole, and it is important to evaluate development of children by teachers that know the complete process (MEB, 2013;Reid, Diperna, Missal & Volpe, 2014). The literature review by Hoge and Coladarci (1989) indicted that there was a strong relationship between teachers' views towards academic success of children and direct measurement.
One of the measurement tools that analyses academic success of children with teacher evaluation is "Teacher Rating Scales of Early Academic Competence (TRS-EAC)" developed by Reid, Diperna, Missal and Volpe (2014). This scale is an important measurement tool to evaluate academic competence of children and identify their weaknesses. When literature was reviewed, TRS-EAC had ease of use and high level of reliability and validity values. Based on these information, to be able to holistically evaluate academic skills of pre-school children and support these skills, it was believed that measuring early academic skills with reliable and valid measurement tool is important. Due to universal development, technological advancement under current conditions, and high intercultural interactions, it can be stated that it is possible to use one measurement tool developed for a society can be used on other societies. Therefore, the purpose of this study is to perform reliability and validity of Turkish adaptation of TRS-EAC developed by Reid, Diperna, Missal and Volpe (2014). Additionally, it was aimed to include a measurement tool into literature that will enable holistic investigation and development of early childhood academic skills in Turkey evaluated by teachers.

Research Design
This study was conducted in order to adapt the TRS-EAC form to Turkish children, which was developed to evaluate the early academic skills of 36-72 month old children, and it is a general survey model. The general survey model is a screening procedure which is conducted on the whole population or the group, sample to be taken from the population in order to pass a general judgment about the population consisting of a large number of elements (Karasar, 2007).

Population and Sample
The population of this study consisted of 2942 children, aged between three and six years old who were attending to nursery classes and kindergartens affiliated to Afyonkarahisar Ministry of National Education in 2015-2016 academic year. Related items in the scale were evaluated with five-scale Likert type based on "significantly below age, below age, compliant with age, above age, significantly above age". Thus, expression of related behaviours was evaluated based on the age of the children, and as age difference between children will have no effect on analyses, probability sampling method was omitted.
Non-probability (Haphazard) sampling was adopted for sampling in the population of this study. Haphazard sampling method is selection of sample based on sample size of any part of the population (Arli & Nazik, 2001). While statistical calculations of children included in the sample were conducted, 95% margin of error, 99% confidence level and 50% response distribution was used and calculated as 542. Based on calculated sample number in this study, 619 children were included. To measure test-retest reliability, total of 93 children (60 were five year old, 17 were 4 year old, 16 were 3 year old) were re-tested in a kindergarten's four classes one month later.

Data Collecting Tool
General Information Form and TRS-EAC was used to collect data in the research.

General Information Form
In the research, a general information form was developed by the researcher with the purpose of collecting personal information about children, including three questions about the date of birth, gender, and whether or not received pre-school education. General information forms have been filled in by the researcher for each child depending on the information in the child's personal development files at their schools.

Teacher Rating Scale of Early Academic Competence-Teacher Form (TRS-EAC)
Reid, Diperna, Missall and Volpe (2014) developed TRS-EAC by conducting the reliability and validity study in which the form was filled for 440 children by the 60 teachers in official pre-school institutions in USA where Head Start Program was applied. As a result of the reliability and validity tests, it was emphasized that the reliability of the Early Academic Skill Scale and the Early Academic Enablers Scale was found to be .98, which is valid and reliable at a high level. The scale consists of 81 items. TRS-EAC is developed to measure the early academic qualities of 38-70 month old children and consists of a combination of Early Academic Skills Scale (35 items) and Early Academic Enablers Scale (46 items). Early Academic Skill Scale includes "Creative Thinking, Critical Thinking Skills, Numeracy, Early Literacy, Comprehension"; Early Academic Enablers Scale consists of "Approaches to Learning, Social and Emotional Competence, Fine Motor Skills, Gross Motor Skills and Communication" subscales. Every statement about the academic competence in the scale is scored with five-point Likert scale such as significantly below age expectations (1), below age expectations (2), compatible with age expectations (3), above age expectations (4), significantly above age expectations (5). From Early Academic Skill Scale participants had the lowest score of 35 and the highest score of 175; from Early Academic Enablers Scale participants had the lowest score of 46 and the highest score of 230 (Reid, Diperna, Missall & Volpe, 2014).

Procedure
First of all, the permission of Erin E. Reid, one of the researchers who developed scale, was obtained to make the adaptation work of the TRS-EAC. The native language of the form is English. The English form is translated into Turkish by two expert linguist academics who speak English and Turkish fluently. Then the scale has been converted into one form. The created Turkish form has been translated back into English by another linguist. After the back translation, it has seen that Turkish form of the scale was close to the English form. Then, by preparing a translation expert evaluation form, English and Turkish forms of the scale were sent to five academicians who are at least at the doctorate level to obtain expert opinions on whether the scale items conformed to Turkish. In line with the opinions and suggestions received from experts, the scale was given the final shape.

Gathering the Data
Before the study, parents were informed, and their permissions were obtained. Data was collected through teachers. Teachers had filled the TRS-EAC for each child. Data was collected at Afyonkarahisar province center in April of 2015-2016 academic year. TRS-EAC was applied to 50 children by the teacher in 3 different schools in order to determine accurate measurement rate of the intended behaviors. Then, TRS-EAC was implemented by teachers for 619 children 36-72 months of age who were studying at kindergartens and pre-schools of the Ministry of National Education in Afyonkarahisar province center during the spring of 2015-2016 academic year in order to adapt them to Turkish. 35 teachers filled the TRS-EAC for 619 children. The teachers were given the scales at the beginning of April and they were asked to fill it within two weeks, and the forms were collected two weeks later. Thus, all data was collected in two weeks. Completion time ranged from 10 to 20 minutes for each child.

Data Analyses
The data were analyzed by using the SPSS 16 packet program. The distribution of demographic information was calculated as frequency and percentage values. First and second-level confirmatory factor analysis (CFA) were used to confirm whether the design of the total 81-item scale was validated or not. The aim of second-level factor analysis is to determine whether the intended superstructure has been verified. In other words, it was examined to determine whether one total score could be obtained from the scale. CFA aims to assess the extent of compliance of real data with factorial model which is composed of many observable variables (latent variables). The investigated model could define a construct which was defined by using the data of an empirical study or constructed based on a certain foundation (Sumer, 2000). CFA was conducted to examine whether the factor structure is valid.
Chi-Square Goodness Test (Chi-Square Goodness, χ2) was used in order to assess the validity of the model in confirmatory factor analysis (Cole, 1987;Sumer, 2000). Although there was no consensus on which fit indices were appropriate for the evaluation of the model, it was recommended to use more than one fit index. The most commonly used was Chi-Square Goodness Test (Chi-Square Goodness, χ2).
In order to determine the reliability of the scale, Cronbach's Alpha internal consistency coefficient, split half test, and test-re-test was calculated for sub-dimensions of the two scales and also for both scales-wide. Cronbach's Alpha internal consistency coefficient is an internal consistency estimation method that is appropriate to use when the items are scored as 1-3, 1-4, 1-5 when they are not scored correctly or incorrectly. The split half test was used because of the correlation between the scores of the subjects and the reliability estimate. For split half test, sub-dimensions of scale and all of the items were divided into two identical halves as single-double. Test-re-test was used because the correlation coefficient obtained from two applications was the coefficient of reliability of the scale. The item total correlation explains the relationship between the scores of the children in the study group and the total score of the measurement tool.
Every statement about the academic competence in the scale was scored with five-point Likert scale as significantly below age expectations (1), below age expectations (2), compatible with age expectations (3), above age expectations (4), significantly above age expectations (5). From Early Academic Skill Scale participants had the lowest score of 35 and the highest score of 175, and from Early Academic Enablers Scale participants had the lowest score of 46 and the highest score of 230 (Reid, Diperna, Missall & Volpe, 2014).

Content Validity
The original form of scale has remained true while adapting English version of TRS-EAC been carried out as follows:

First Stage
In the adaptation process of TRS-EAC to Turkish, the test items were firstly translated into Turkish by two English linguist academicians in line with translation-back translation technique, and the scale was translated to Turkish by linguists was converted into one form. The created Turkish form has been translated back into English by another linguist. After the back translation, it was seen that the Turkish Form of the scale was similar to the English form. The scale translated into Turkish has been put into final form was given by an expert Turkish linguist after final examination in terms of sentence structure and clarity of expressions.

Second Stage
Expert opinion has been taken to determine whether the scale items measure the intended content, and to determine whether it is applicable to Turkish. The original form of the test and its Turkish translation were presented to five academicians (expert in preschool education, has at least doctoral level) for their views regarding the validity of the TRS-EAC and the eligibility of Turkish culture. The experts were asked to examine the items on the TRS-EAC form in terms of the research relevantness and clarity, and submit their opinions on the items which the amendment, correction and removal of the materials where they seemed necessary. In addition, in order to ensure that validationreliability study produce success at the desired level, they were asked to evaluate the test items in terms of the convenience to Turkish culture, in accordance with the triple Likert-type evaluation criteria as "Eligible", "Not Eligible" and "Modifiable".

Third Stage
In order to determine the validity of the TRS-EAC, the content validity ratio (CVR) for each item was calculated in the evaluation of five expert opinions. Then, the content validity index (CVI) was determined by taking the averages of the calculated CVRs. This index was used for the determination of whether each item was considered necessary by experts (Yurdugul, 2005). This value was calculated for the compliance level of the items. Since the number of experts was five, it was concluded that items with CVR value greater than 0.59 are thought to provide the coverage validity (Yurdugul, 2005). It has been decided that all items should remain in the scale as a result of positive opinion of all experts about the items. CVR and CVI values were obtained as 1.

Fourth Stage
The pre-application of TRS-EAC, which was finalized in line with the opinions and recommendations of the experts, was realized by applying it to 50 children in three different schools in order to determine the correct measurement of the intended behavior and skills. As a result of pre-application, the analyses result of TRS-EAC were examined, and it was determined that the reliability coefficient and total item correlations were sufficient for the whole test. Following the pre-application, scale-related adjustments were made and the scale was made ready of reliability study and it was applied to the sample which was formed from 619 children ranging in age from 36 to 72 months attending the preschools.

Structure Validity and Reliability
Confirmatory factor analysis was conducted in an attempt to present the validity of the factor structure while adapting the TRS-EAC to Turkish. In this direction, a separate confirmatory factor analysis was applied to the "Early Academic Skills Scale" and the "Early Academic Enablers Scale" which constitute the scale in order to evaluate the structural validity of the scale in the research.
First and second-level confirmatory factor analysis (CFA) was used to confirm whether the design of the total was validated or not. The aim of second-level factor analysis was to determine whether the intended superstructure has been verified. In other words, it was examined to determine whether one total score could be obtained from the scale. CFA aims to assess the extent of compliance of real data with factorial model which is composed of many observable variables (latent variables). The investigated model can define a construct which was defined by using the data of an empirical study or constructed based on a certain foundation (Sumer, 2000). CFA was conducted to examine whether the factor structure was valid.

Early Academic Skills Scale
CFA was applied to assess whether the 5 factors and the 35-item structure of the scale were verified. In the first CFA application, items with a statistically insignificant t value were examined.
When path diagram was examined, all the factors and items were preserved in the scale since none insignificant t value was found in the conducted examination. The fit indexes were χ 2 =2609.68, X 2 /sd= 3.95, RMSEA= 0.074, CFI=0.98, NNFI=0.98, NFI=0.97 and IFI=0.98. When the coefficients showing the relationship between the observed variables model showing the factorial structure of the scale and its factors were examined, it was concluded that all the coefficients were high. When the compliance statistics calculated with CFA considered, it has been decided that the previously determined five-factor structure of the scale yield with collected data in general.
The regression values and t-values of the items are given in Table 1. When Table 2 was examined, it was determined that the obtained regression coefficients, and t values were significant (t> 1.92), and that the model was validated.
Second-level confirmatory factor analysis was performed to show that 5 factors of scale were representative of the superstructure. This examined model was based on the relationships between latent variables obtained in first level confirmatory factor analysis. The analysis also revealed the variances explained by the first level variables of the upper level (second level) variables.
The factor loads (Lambda x, λx), t values, measurement errors (delta δ) between the first level latent variables and the upper level (second level) variables of the model and the explanatory ratios (R²) of the second level variable at the first level variables were presented in Table 2: When the path coefficients and t values between the second level latent variable and first level latent variable were examined, it was found that the relations between the upper latent variable and all the factors were positive and significant (p<0.05). When the variances explained in the first-level variables by the second-level variables were examined, it was determined that the numeracy susceptibility dimension explained the top level at the highest level.
To calculate reliability of Early Academic Skills Scale, split half test, test-re-test, and Cronbach's Alpha reliability were calculated.
For split half test, sub-dimensions of scale and all items were divided into two identical halves as single-double (Table  4). Test-retest reliability that indicated level of stability for time-based repetitive measurements was calculated for all subdimensions of the scale. When Table 4 was investigated, for all sum-dimension and all scale, applied repetitive measurements indicated high level of correlation and test re-test reliability was achieved.
To calculate internal consistency coefficient of Early Academic Skills Scale Cronbach's Alpha reliability were calculated. The internal consistency coefficient Cronbach's Alpha was calculated for the reliability of the scale. The alpha value of the "Creative Thinking" sub-dimension was 0.97; the alpha value of the "Critical Thinking" sub-dimension was 0.97; the alpha value of the "Numeracy" susceptibility sub-dimension was 0.96; the alpha value of the "Early Literacy" subdimension was 0.97; the alpha value of the "Comprehension" sub-dimension was set to 0.94. The alpha value for all of the scale was 0.985. When Table 5 examined according to Early Academic Skills Scale's evaluation criteria, the total item correlations were at a sufficient level. The Table 6 shows average scores of children from the Early Academic Skills scale. It can be seen that the general average scale score of children was 110,56, the average scale score for Creative Thinking sub-dimension was 25,30, the average scale score for Critical Thinking sub-dimension was 38,21, the average scale score for Numeracy subdimension was 16,26, the average scale score for Early Literacy sub-dimension was 14,34, and the average scale score for Comprehension sub-dimension was 35,82.

Early Academic Enablers Scale
The CFA was applied to assess whether the scale, which has a 5-factor and a 46-item structure was true or not. In the first CFA applied, substances with a statistically insignificant t value were examined.
When path diagram was examined, all the factors and materials were preserved in the scale since no t value was found which was not meaningful in the examination made. The fit indexes were χ2 = 4421.86, X2 / sd = 4.29, RMSEA = 0.077, CFI = 0.98, NNFI = 0.98, NFI = 0.98 and IFI = 0.98. When the coefficients of the relationship between observed factors and variables of the model showing factorial structure of the scale were examined, it was concluded that all the coefficients were high. Given the compliance statistics calculated with CFA, it has been decided that the aggregate yield generally complements the previously determined five-factor structure. Table 7 shows the regression values and t values of the items.  Table 7 was examined, it was determined that the obtained regression coefficients ,and t values were significant (t>1.92) and that the model was confirmed.
Second-level confirmatory factor analysis was performed to show that 5 factors of scale were representative of the superstructure. The basis for this model was the relationships between latent variables obtained in first level confirmatory factor analysis. The analysis also revealed the variances explained by the first level variables of the upper level (second level) variables.
The factor loads (Lambda x, λx), t values, measurement errors (delta δ) between the first level latent variables in the model and the upper level (second level) variable and the explanatory ratios (R²) of the second level variable in the first level variables were presented in Table 8: When the path coefficients and t values between the second level latent variable and the first level latent variable were examined, it was found that the relations between the upper latent variable and all the factors were positively significant (p <0.05). When the variances explained in the first-level variables by the second-level variables were examined, it was determined that the fine motor size explains the superstructure at the highest level.
To calculate reliability of Early Academic Enablers Scale, split half test, test-retest, and Cronbach's Alpha reliability were calculated.
For split half test, sub-dimensions of scale and all items were divided into two identical halves as single-double (Table  9). When Table 9 was investigated, the Split Half Correlation value of the "Approaches to Learning" sub-dimension was indicated at 0.969, "Social-Emotional Competence" sub-dimension was 0.958, "Fine Motor Skills" sub-dimension was 0.905, "Gross Motor Skills" sub-dimension was 0.963, and "Communication" sub-dimension was 0.966. The Split Half Correlation value for all the whole scale was 0.99. Based on all sub-dimensions of scale and total split half test reliability results, the results were statistically reliable.
Test-retest reliability that indicated level of stability for time-based repetitive measurements was calculated for all subdimensions of the scale (Table 10). When Table 10 was investigated, for all of the sum-dimensions and all the scale, applied repetitive measurements indicated high level of correlation and test-retest reliability was achieved.
To calculate internal consistency coefficient of Early Academic Skills Scale Cronbach's Alpha reliability were calculated. The internal consistency coefficient Cronbach's Alpha was calculated for determining the reliability of the scale. The alpha value of the "Approaches to Learning" sub-dimension was indicated at 0.97, the alpha value of the "Social-Emotional Competence" sub-dimension was 0.97, the alpha value of the "Fine Motor Skills" sub-dimension was 0.98, the alpha value of the "Gross Motor Skills" sub-dimension was 0.98, and the alpha value of the "Communication" subdimension was 0.98. The alpha value for all the whole scale was 0.99.  When Table 11 was examined according to Early Academic Enablers Scale's evaluation criteria, the total item correlations were at a sufficient level.

Discussion and Conclusion
The aim of the study was to conduct the reliability and validity work of TRS-EAC developed in the USA by adapting the scale to the Turkish children in order to gain a reliable instrument to literature that can measure the academic competencies of Turkish children as a whole. TRS-EAC has been developed by conducting the reliability and validity study in which the form was filled for 440 children by the 60 teachers in official pre-school institutions in USA where Head Start Program was applied. Dunn (2016) and Cleveland (2017) Alpar (2014) stated that scales that had correlation coefficient of 1 or close to 1 was reliable at a perfect level. Correlation values obtained from this study were between 0.89 and 0.99. Since these values were close to 1, this proves that this measurement tool has high reliability.
When Early Academic Skills Scales Test's re-test reliability results were analyzed, the test/re-test Reliability Correlation value was found to be between 0.994 and 1. When Early Academic Enablers Scale's results were analyzed, the test/retest reliability correlation value was between 0.996 and 1. If the correlation coefficient of the test/re-test approaches +1, it can be said that the related scale is reliable and measurements show similarity (Alpar, 2014). Tavsancil (2006) expressed that to prove the time-based stability of a scale, the correlation coefficient must be positive and high in this applied method, and this should be at least .70 for scales. Since the correlation values obtained from this study were above 0.994, i.e. close to +1, this scale could be considered reliable.
Based on item analyses results for Early Academic Skills Scale's and Early Academic Enablers Scale's evaluation criteria, the total item correlations were at a sufficient level. Reid, Diperna, Missal & Volpe (2014) found the sufficient total item correlation for scales. Buyukozturk (2010) emphasized that positive and high total item correlation sampled similar behaviours in the measurement tool. Therefore, the test measurement tool has high internal consistency. Additionally, items with .30 or higher total item correlation differentiated individuals at a good level and these items should be included in the scale. Since total item correlations obtained from this study were higher than 0.647, these values showed that items differentiate individuals at a good level and these items should be kept in the scale.
It was stated that under early academic skills, early literacy skills was linked with early academic skills of children (Purpura, Schmitt and Ganley, 2017). Studies show that there was a significant relationship between numeracy skills and early literacy and with early academic skills in general (Duncan et al., 2007;Pagani, Fitzpatrick, Archambault, & Janosz, 2010). Therefore, it could be said that reliability results obtained from this study and the average scores of children supported each other. Reid, Diperna, Missal & Volpe (2014) explained creative thinking with generating new ideas while other authors explained critical thinking with reasoning and analytic skills. Accordingly, a validity and reliability study conducted by the authors showed a high reliability in creative and critical thinking sub-dimensions. This result supports the results obtained in this study.
When the fine and gross motor skills sub-dimension was analyzed under the Attitudes for Early Academic Success, it could be stated that this scale had a unique property to relate fine and gross motor skills with academic skills and measuring this relationship (Reid, Diperna, Missal & Volpe, 2014). Reid, Diperna, Missal & Volpe (2014) found high reliability results of fine and gross motors skills sub-dimensions with this measurement tool. It was stated that fine motor skills played an important role for comprehending numeracy and fingers had an important role when there were no visual materials (Noel, 2005). All these information support the high reliability obtained in the fine and gross motor skills sub-dimensions in this study and average scores of children. Social and emotional skills were collected under the same sub-dimension, and it was stated that there was a direct relationship between social-emotional competence and interaction (Senol, 2017;Reid, Diperna, Missal & Volpe;. It can be stated that children with adequate social and emotional competence and high interaction should have higher attitudes towards academic success and results of this study supports this argument.
Based on this information, it could be said that the average scores of children obtained from this measurement tool support explanations given above as well as previous studies.
Results indicated that TRS-EAC was a scale prepared by considering development properties of 36-72 months children to evaluate academic competence of children and high reliability results were obtained after this scale was adopted to Turkish and applied on study group. It could be commented that TRS-EAC was a valid and reliable measurement tool to identify academic skills of children. It is believed that the scale will have significant contribution to literature in this form.

Recommendations
Although high level of reliability and validity results were obtained from TRS-EAC, there were certain limitations in this study. Recommendations related with these limitations are presented:  TRS-EAC is only limited with normal development children in pre-school of Afyonkarahisar. To increase generalizable structure of TRS-EAC, it could be recommended that the scale needs to be applied to larger samples.  Since compliance with age was considered to evaluate TRS-EAC scale items, probability sampling method was omitted. It could be recommended that this study could be constructed with probability sampling method and can be applied to different groups by creating properties such as age, socio-economic status etc.
 Demographical information of teachers was omitted during reliability and validity study of TRS-EAC. Based on certain demographical properties, considering that teachers may have more information regarding development of children, calculation for teacher classification based on these variables can be recommended.  The relationship of sub-dimensions with each other can be investigated by educational interventions.
In addition to limitations of this study, the following recommendation can also be provided:  It could be suggested that making various studies with TRS-EAC in Turkey will significantly contribute to strength of the scale. It could also be suggested that TRS-EAC could be used with other scales in special education, primary education, elementary school, speech and language therapy, psychology, guidance, and consultant, and studies that investigate effects of this scale that was used for preparing developmental intervention and education programs could be planned.