Developing Factors and Indicators of Growth Mindset for School Administrators in Thailand

The sustainability of educational administration is depending on school administrators’ growth mindset. A growth mindset is detected to have a direct relationship with school achievement and success. Despite being an integral part of school leadership, the growth mindset of school administrators has been ignored by past researchers. Hence, this research is designed to investigate the growth mindset’s significant factors and indicators of high school administrators in Thailand. The researchers employed a quantitative approach survey design. A total of 460 school administrators and teachers participated in a survey using a multi-stage sampling technique. The researchers intended to test whether the identified factors and indicators are fitting with empirical data as the ultimate research outcome. The results revealed that there are a total of 17 indicators derived from the six factors in a growth mindset model. The measurement model of growth mindset is corroborated to the empirical data, with χ 2 =64.875, df=50, χ 2 / df=1.2975, CFI=0.99, TLI=0.99, RMSEA=0.02, and SRMR=0.01. In conclusion, the developed growth mindset model for high school administrators has a goodness-of-fit with the attained data. Finally, the results of this research have successfully proposed a measurement model that would be guidelines for school administrators to grow their positive mindset as our major contribution to the educational administration field.


Introduction
The key success of any school is depending on school administrators' mindset because they are not only setting their growth mindset for long-term achievement, but they must ensure that their leadership will be sustainable in the future (Meador, 2019). Goldstein and Brooks (2007) defined school administrators' mindsets as the assumptions and expectations they have for themselves and others that manage their practices and dealings with others. Goldstein and Brooks further describe the factors and indicators of the mindset of effective school administrators are the ways they cultivate this mindset in their subordinates. Elmore (2019) identified four principles to grow school administrators' positive mindset. These four principles include teaching practices that should be affixed to our core instructional schooling, systemic problems entail systemic solutions, educational leadership is a profession without practice, and powerful practices entail strategies. Elmore explained that school administrators sometimes need to spend a few years to identify the issues for developing not only their leadership mindsets but also the leadership mindsets within their subordinates. For example, those teachers who lead professional learning community activities, chair the grade-level meetings, and organize departmental activities need leadership mindsets, too. Therefore, those teachers should be a part of systemic solutions in order to strengthen the instructional core of schooling (DeWitt, 2020). Dweck (2006) who is a popular researcher in the area of students' mindset has identified the new psychology of success in students' learning. Dweck emphasized the importance of classroom and school environment factors to motivate students' mindset. According to Dweck (2006), there are two methods to investigate students' intelligence or capability, namely a fixed mindset and a growth mindset. A fixed mindset is defined as human beings are born with a specified level of capability and it is rooted. Therefore, individuals with a fixed mindset consider that these undeveloped abilities are stationary and unchangeable. It also called entity theory (Rissanen, Kuusisto, Tuominen, & Tirri, 2019). On the other hand, a growth mindset refers to human beings' effort to develop their capabilities over time by tapping in the attempt and believing sufficient self-confidence to seek innovative strategies to obtain deeper learning through hard work. This means that individuals with a growth mindset consider that cleverness, character, and capabilities can be developed, so-called incremental theory (Rissanen et al., 2019).
Generally, the school administrator's major role is to ensure they are cultivating a culture that comprises factors and indicators associated with the anticipated behavior or outcome to assist teachers to be this best. Therefore, school administrators must help teachers to have growth mindsets and allow teachers to utilize the activities of students with a growth mindset (Guidera, 2014). Guidera found that school administrators' intentional and coordinated actions could make the growth mindset norm change. Moreover, those successful school administrators have shown their capabilities in providing teachers with targeted professional development and coaching on the norms and proposed interference training for the struggling teachers. As a result, Guidera recommended school administrators build schoolwide cultures with growth mindset norms.
High school administrators in Thailand are required to have thoughtful behaviors and a growth mindset to offer optimal learning conditions for students as elucidated in Educational Administration in Thailand 4.0 (Boonmepipit & Jiamjan, 2020). Since school administrators hold the most power in developing a school's organizational structure, teachers' beliefs in their collective ability to help all students grow and learn will be influenced by school administrators' growth mindset (Hanson, Bangert, & Ruff, 2016). Therefore, this research aims to develop a growth mindset model for high school administrators in Thailand. Ultimately, the growth mindset model can assist high school administrators to develop national youth who have the quality that equal to many civilized countries as the planning of Educational Administration in Thailand 4.0.

Research Design and Research Process
The researchers employed a quantitative method using a survey. The strength of employing quantitative data is to generate deeper, and ultimately the research results will be more reliable, actionable, and useful research intuitions (Lavrakas, 2008). Therefore, researchers constructed a survey to test the structural construction between experimental examination and the hypothetical theory of quantitative associations concerning experimental data. The associations are epitomized by path coefficients or deterioration between the growth mindset factors and their indicators. Figure 1 illustrates the research process.

Population and Sampling
The research population was comprised of school administrators and teachers from four provinces, namely Kalasin, Khon Kaen, MahaSarakham, and Roei provinces. All the high schools located in these four provinces are allied with High School Educational Service Area 24 to 27 under the supervision of the Basic Education Commission in Thailand. The researchers employed the rule of thumb proposed by Becker and Ismail (2016) to formulate an adequate sample size (N). The identified sample size is recognized as the presence of classified practice in reaching an adequate probability for the requisite results such as model convergence, statistical precision, and statistical power for particular confirmatory factor analysis (CFA) with empirical data. This is followed by determining the ratio of parameter and samples as 20:1 to fulfill the sample size criteria (Hair, Black, Babin, & Anderson, 2013). Owing to there were 23 parameters in this research that directed to at least 460 as required sample size.
The researchers employed Yamane's (1970) formula at a 95% confidence interval using the multistage sampling technique. Firstly, researchers divided the population into provinces and selected a small sample of relevant separate groups according to school size. This sampling method is allowed significant provinces of the selected samples are split into sub-groups, namely school administrators and teachers at various stages to make it simpler for obtaining primary data. Table 1 displays the dispersal of population and sample groups.

Research Instrument
The researchers employed a survey questionnaire with 95 closed questions as a method to collect quantitative data. A continuous five-choice Likert scale was utilized to evaluate the strength of perceptions. This Thai language questionnaire has six sections, and it was intended to collect information pertaining to respondents' perceptions of growth mindset practice. There were 18 questions about the three indicators of innovative and systematics thinking skills (IST) factor in Section A. This is trailed by Section B to F that was specifically designed by the researchers to gauge information about focus on learning goals (FLG) factor, acknowledgment of success (AS) factor, openminded on change (OMC) factor, being a resourceful and ambitious leader (RAL) factor, and commitment to personal development (CPD) factor, respectively. The following Section B to Section F consisted of three FLG indicators (18 questions), three AS indicators (13 questions), two OMC indicators (12 questions), three RAL indicators (16 questions), and three CPD indicators (18 questions).

Data Analysis
Structural Equation Modeling (SEM) were used to analyze quantitative data. The SEM is an appropriate method to analyze the structural relationship between measured variables and latent constructs because it syndicates factor loading examination and path analysis or multiple regression examination (Gay, Mills, & Airasian, 2011). Moreover, SEM can estimate the multiple and interrelated dependence in a single analysis, namely endogenous and exogenous variables. In this research, the endogenous variable refers to the growth mindset of school administrators and exogenous variables are the conceptualized factors and indicators from the first phase. Consequently, researchers utilized SEM methods to assess how meticulously a hypothetical model fits empirical data to examine the measurement model. The measurement model signifies the hypothesis that denotes how identified factors and indicators join together in corresponding to the hypothesis. Hence, researchers utilized a CFA to examine test the measurement model for its goodness-of-fit.
As mentioned by McDonald and Ho (2002), absolute fit indices mean how appropriately a measurement model fits the empirical data and verifies which projected model has the greatest fit. In this line of reasoning, researchers employed the Comparative fit index (CFI), the Chi-Square statistic (χ 2 ), the Tucker-Lewis index (TLI), the Standardised root mean square residual (SRMR), the Root mean square error of approximation (RMSEA), the Goodness-of-fit statistic (GFI), Normed-fit index (NFI), and the Adjusted goodness-of-fit statistic (AGFI) to analyze the maximum-likelihood estimation and multiple indices of model fit as the variance-covariance matrix.

Results
The preliminary results identified six key factors of growth mindset: (i) Innovative and systematic thinking skills (IST); (ii) focus on learning goals (FLG); (iii) acknowledgment of success (AS); (iv) open-minded on change (OMC); (v) being a resourceful and ambitious leader (RAL), and (vi) commitment to personal development (CPD). Moreover, there are 17 growth mindset indicators and 48 elements which derived from the six key factors with regards to fit the Thai context. Table 2 display the details of the key factors, indicators, elements of growth mindset.

The Goodness-of-Fit of the Growth Mindset Factors and Indicators with the Empirical Data
Researchers started to examine the suitability of data for factor analysis before obtaining estimates of the parameters of the growth mindset model. Two key concerns that must take into account to decide whether the obtained data is suitable for CFA, namely the strength of the association between factors or indicators and sample size (Pallant, 2013). The strength of the association between factors or indicators is measured using Bartlett's Test of Sphericity (Bartlett, 1954) while researchers used Kaiser-Meyer-Olkin (KMO) to verify whether the sample size is sufficient or not. According to Jöreskog and Sörbom (1993), large samples are useful because it is almost impossible for us to reject the null hypothesis even though the chi-square (χ 2 ) is recognized as a standard statistic to evaluate the general fit of the measurement model with the empirical data.
A Bartlett Test of Sphericity is an evaluation of multivariate normality according to data distribution. This means that it is used to verify whether the unique correlation matrix is an identity matrix or not in conformity with the null hypothesis. In other words, if the significant values are more than 0.05 for both factors and indicators imply an identity matrix is produced by the obtained data. It is worth remarking that the factors or indicators have to evaluate at the interval level.
On the other hand, several specialists have recommended different rules of thumb to decide the acceptable KMO value as the measurement to confirm the adequacy of sampling. For example, Kaiser (1974) and Field (2000) determined the acceptable value as more than 0.5 while Pallant (2013) confirmed KMO value must be more than 0.6. The researchers decided to use Hutcheson and Sofroniou's (1999) rule of thumb to decide the acceptable KMO value as shown in Table 3.  Table 4 shows that the sampling size is sufficient and excellent because all the KMO values of factors and indicators are above 0.9 (Hutcheson & Sofroniou, 1999;Pallant, 2013). Besides, Table 4 also shows that collected data were nearly multivariate normal according to the result of Bartlett Test of Sphericity, and excellent sample size was obtained as reflected in KMO value (Hutcheson and Sofroniou, 1999). Therefore, the obtained data could proceed for further examination. This was followed by seeking to attain estimates of the parameters of the growth mindset model, the validity of the identified factors, and their factor loading of the growth mindset. In short, factor loading means the 'relative importance' of the identified indicators that collectively form a specifically identified factor in the growth mindset model of high school administrators that had been considered. The co-variance with the growth mindset factors ranged from 94.30 to 99.10 percent. As illustrated in the following Table 5, the factor loading of all the growth mindset factors are ranged from 0.971 to 0.995 and is statistically significant at 0.01. The factor with the highest factor loading value is innovative and systematic thinking skills. This is followed by open-minded on change, ier.ideasspread.org International Educational Research Vol. 4, No. 1;2021 commitment to personal development, being a resourceful and ambitious leader, and acknowledgment of success. The factor that has the lowest factor loading value is the focus on learning goals. In conclusion, all the key factors are found to be essential constructs of growth mindset for school administrators in high schools. In addition, the co-variance with the growth mindset indicators is in the range of 20.60 to 95.00 percent. As demonstrated in the following Table 6, the factor loading of all the growth mindset indicators are ranged from 0.454 to 0.976 and is statistically significant at 0.01. In this line of reasoning, all the identified indicators are considered important constructs for the growth mindset model.   (2000) Pass In this line of reasoning, it is finalized that the growth mindset model is approved with the empirical data. Hence, the measurement model is accepted according to the above rules of thumb and cut-off values. Therefore, the researchers established precise and significant paths of the growth mindset model as illustrated in Figure 2. Table  8 shows the results of indicators that are attached to the factors of the growth mindset model.

Discussion
A growth mindset model was projected and verified its goodness-of-fit. The results indicated that all six factors have a solid, positive, and significant influence on the growth mindset of school administrators. Moreover, the measurement model revealed that the highest prediction influence is the innovative and systematic thinking skills factor. The second highest prediction influence is open-minded on the change factor. This is followed by the commitment to personal development, being a resourceful and ambitious leader, and acknowledgment of success factors, in this descending order. However, the least capacity of prediction influence is focused on learning goals factor. As a result, good school administrators must grow their expectations through the identified factors and indicators to cultivate teachers' and students' mindset in their schools, as emphasized by Goldstein and Brooks (2007). Besides, this result is consistent with results from previous research (DeWitt, 2020;Elmore, 2019;Meador, 2019). In conclusion, the goodness-of-fit results can assist us to predict future trends and patterns while school administrators are using the growth mindset model.
The crucial results of the current research support our hypothesis that all the six identified factors and 17 indicators of growth mindset accord with state-of-the-art conceptions of positive mindset for school administrators. If school administrators lack knowledge about the mindset phenomenon can hinder them from understanding the consequences of their practices or from rendering teachers and students' performance appropriately (Rissanen et al., 2019). Evidently, a good consideration of the associations between the factors and indicators is important for school administrators to implement the growth mindset model. This is because these research results have successfully provided preliminary indications about their causal associations. Therefore, the researchers recommended that school administrators with a growth mindset are able to make efforts to develop teachers' and students' capabilities and seek innovative strategies to obtain deeper learning according to incremental theory (Rissanen et al., 2019).
The results of factor loading of identified factors and their indicators in the growth mindset model imply that all the factors and indicators are important constructs to grow school administrators' positive mindset. Hence, all the key factors and their indicators are fit well with empirical data at a statistically significant level (Tuksino, 2009). In this line of reasoning, researchers concluded all the factors and indicators appear to be consistent with either growth mindset theory (Dweck, 2006) or previous research studies (DeWitt, 2020;Elmore, 2019;Meador, 2019).
Lastly, researchers suggested to the Thailand Ministry of Education so that the growth mindset model will be included in educational administrators' preparation training program. However, it can be developed as a critical instrument by helping as a reasonable yardstick for professional training development, and for advancing more desirable experimental research on how organizational development in terms of growth mindset can be done.