Research in Social Sciences and Technology (RESSAT) E-ISSN: 2468-6891 16 Effectiveness of Coaching by School Improvement Partners through Good Relationship and Professionalism Skills Practiced Zalina Mohd Tahir1 & Nabihah Mohd Salleh2 Abstract School administrators consisting of Principals and Headmasters (PGB) play a significant role in leading the school to achieve excellence. To ensure that excellence is achieved within the timeframe stipulated, the PGB received guidance from fellow coaches known as School Improvement Partners (SIP). The coaching process requires mutual understanding between the two parties so that it can work smoothly and effectively (Narimawati, 2007) without any conflict of ideologies. This study is to identify the effectiveness of the coaching of SIP through the good relations and professional skills practiced during the ongoing coaching. A total of 212 PGB as respondents and data were analyzed using SPSS version 23 for the descriptive analysis, while AMOS version 22 is used for advanced statistics Structural Equation Modeling. The study found that the goodness of fit model by SIP coaching is good with reading of chi-square/df=2.203, CFI= .969, p=0.00 and RMSEA=.076. Good relationship with professional coaching sessions have contributed a total of 71% (R2=.71) to the effectiveness of the coaching as well as it helps to increase the PGB confidence in the development of schools. This finding confirms that coaching by SIP directly acts as a catalyst for the success of the school and to complete the needs of mission and vision of the Ministry of Education, Malaysia. Key words: Administration, management, coaching, performance, quality. Introduction The success of an empire is very dependent on the ability of leaders to think critically, analytically and thoroughly. To achieve the status of a developed nation, the effectiveness of thought in designing strategies and implementing an agenda of action is particularly important in the education system. Malaysia as a nation that is moving towards a developed status is concerned with educational attainment among people. One factor that helps to achieve excellence is when a person reaches the level of confidence, excitement and satisfaction in the 1 Ph.D., – Department of Education, Federal Territory of Kuala Lumpur, Malaysia Ministry of Education, Malaysia School Improvement Partner, zalina.mtahir@yahoo.com 2Department of Education, Federal Territory of Kuala Lumpur, Malaysia Ministry of Education, Malaysia, nabihahppgc@gmail.com mailto:zalina.mtahir@yahoo.com mailto:nabihahppgc@gmail.com Tahir & Salleh work. Many studies have been conducted regarding the level of job satisfaction which is linked to the success of an organization. Among them are the working environment and style of administration (Karim, 2008) or the management of the education system (Huda et. al., 2004). Education is one area that cannot be set aside because it contributes to the development of the country. The Malaysian education system should be designed strategically to produce the best for all parties (Zainudin, 2010). Good planning should be realized by the head or administrators in an organization. At school level, the principal or headmaster plays a key role in leading the organization. Next, the responsibilities go to the second layer of the administration and so on to other educators to mold students to achieve the desired results. In administrating an organization especially schools, administrators faced various problems including the satisfaction and comfort of working with others. If they are not comfortable in any case, they are likely to lead to failure in fulfilling their responsibilities (Zalina, 1997). Certainly this brings negative impact to the organization. Thus, each educator is responsible for their respective roles in order to realize excellence. Statement of Problem The transition of change in education in the delivery of knowledge using the 20th century approach to teaching of the 21st century seems to be taking quite a long time in Malaysia. In reality, the change in the approach of delivery of knowledge is very significant in developed countries such as Finland and other European countries. Hence, there should be a move so as education in Malaysia today are at par with the developed countries. The Malaysian government through the Ministry of Education (MOE) is aware of this situation and would continue to look for solutions that are more meaningful and efficient. Hence, the Malaysia Education Blueprint (2013-2025) (MEB) was launched to address the current and future needs of the citizens as a whole. MOE have always emphasized on the development of academic achievement and character building of students in schools. Through the MEB, new posts have been created, namely a group of experts who are specially prepared to guide teachers in developing school. These Research in Social Sciences and Technology (RESSAT) 2018: 3(1), 16-35 experts consist of two groups known as School Improvement Partners (SIP) and School Improvement Specialist Coaches (SISC). SIP plays a different role which encompass a coach, mentor and counselor to the administrator of the school, while the SISC is to act as an expert adviser to subject teachers. Both groups have a role as partners to upgrade the school in moving towards achieving excellence. Administrator has very critical role to play in managing a school. Principals and Headmasters (PGB) are responsible in helping to shape the quality of teachers (Azlin Mansor, 2006; Tarman, 2012; Yigit, & Tarman, 2016) and the outcome of students. To enable them to perform effectively, PGB should be responsible for the rise and fall of their administration. According to Mayer (1992), commitment of the administrator refers to the attitude shown in the organizations such as schools. Thus, the relationship between the PGB with SIP should always be maintained in order to avoid further negative prejudice and to create a sense of belonging in creating effective guidance. Comfort and satisfaction is usually achieved when administrators and SIP can share the problems and potential solutions and take a view without prejudice of their respective capabilities. Objective of the Study i. Measuring the level of relationship of Principal/Headmaster (coachee) with School Improvement Partners (SIP) through coaching sessions. ii. Measuring the level of professionalism skills in the performance of SIP in coaching the Principal/Headmaster (coachee) . iii. Measuring the effectiveness of the coaching of SIP towards the Principal/ Headmaster (coachee) in respective school. Research Method Design Review The study conducted is based on quantitative, descriptive and correlation. This method emphasizes the specific objectives for translating research questions in addition to help researchers collect and analyze data (Chua, 2006: 2009; Fraenkel, 2007). This method also Tahir & Salleh aims to get the facts by using the objective measurement and statistical analysis of numerical data to explain actual situations (Yin, 2003). Thus, the researchers developed a questionnaire consisting of two parts. Part A consists of 10 items related to the information and personal details of respondents, while Part B covers three (3) constructs namely; i) 8 items related to the relationship between SIP with PGB; ii) 11 items related to the field of professionalism skills of SIP; and iii) 12 items related to the effectiveness of the coaching of SIP towards PGB. The questionnaire is based on the interval level with Likert scale (1-10) to enable the respondent to make a more accurate assessment (Likert, 1932). According to Zainudin (2012), with more choices, higher degrees of the scale, there will be better validity of the answers. The scale is from "1" as "Strongly Disagree" to "10" as "Strongly Agree". Each of the 31 items of the three constructs showed the total value of reliability (cronbach's alpha) 0.958 (pilot test) and 0.961 (actual study), thus explaining the strength of the items in each construct. According to Hair et. al., (2006) the reliability of the minimum acceptable is 0.60. Table 1.1 shows the value of each construct was studied. Therefore, this study takes the same value as the reliability level of the index and found that questionnaire is considered higher which between 0.958 and 0.961. Table 1.1: Value of Reliability and Number of Items in Every Construct Construct Number of Items Cronbach Alpha (Pilot Test) Cronbach Alpha (Actual Study) Relationship of SIP With PGB 8 items 0.948 0.956 Professionalism Skills of SIP 11 items 0.953 0.950 Effectiveness of coaching 12 items 0.973 0.979 TOTAL 31 items 0.958 0.961 Method of Analysis Demographic analysis is carried out through statistical analysis for social science (SPSS) (Pallant, J., 2010) version 23. Measurement of the degree of relationship, professionalism skills and effectiveness of the guidance towards PGB involve statistical analysis with advanced AMOS 22 which were divided into three stages. The first stage is an exploratory analysis of data known as Exploratory Factor Analysis (EFA) (Hair et. al., 2010: Zainudin, 2012) or there Research in Social Sciences and Technology (RESSAT) 2018: 3(1), 16-35 are researchers who use Principle Component Analysis (PCA) (Hair, et. al., 2006). This analysis is designed to evaluate the psychometric characteristics of the instrument which is based on the construction of variables. This technique is also used to get a new validity and reliability of each sample. The second stage is known as Confirmatory Factorial Analysis (CFA), which aims to see the correlation between the variables used. In addition, the analysis of these factors will also determine the appropriateness and validity of items to measure variables and to test the structure of covariance in each variable. Finally, at times this analysis will result in a reduction of data produced and the data is more significant (Hair, et. al., 2010). In the third stage, the analysis of Structural Equation Modeling (SEM), which aims to test the construct of the measurement model achieve its suitability index (fitness indexes). All three stages of the analysis will help researchers to evaluate the psychometric characteristics of the instrument which is based on the construction of independent variables, namely through the construct of the relationship between SIP with PGB, and constructs related to the field of SIP’s professionalism while the dependent variable involves constructs of the effectiveness of coaching. All scales used will be confirmed before proceeding with the next analysis. Population Sample Survey The respondents consisted of Principals and Headmasters serving in schools in the Federal Territory of Kuala Lumpur between the years 2016-2017. Mohd Majid Konting (1998), states that the sampling technique is one of the research strategy where information obtained can be used as the real representative of the population. This study selected a simple random sampling because this method can prevent the occurrence of bias data collection. (Shaver and Norton, 1980). A total of 230 headmasters and principals have agreed to answer the questionnaire in this study. However, only 212 (92.1%) questionnaires were completed by respondents and used as sampling. The study went through a normal process of data collection. Initially, a pilot test was conducted on a total of 58 principals/headmasters selected to respond to the questionnaire. The goal is to Tahir & Salleh identify the level of their understanding of the questions posed. A pilot test was conducted to school heads in two zones namely Bangsar and Pudu. After a week, the questionnaires were collected and re-refining process was made to the questions which are less clearly defined. After carrying out the re-refining of questions and discussions among researchers (SIPs), the researchers have gone to schools to get approval of PGB to be respondents. After approval was obtained, the questionnaires were then distributed to the respondents concerned. At this point of time, they were also given a week to understand and answer the questionnaires. Other SIP colleagues also helped in the effort to distribute and collect questionnaires completed by respondents. Data Analysis The data were analyzed in two ways: i) descriptive analysis that involves frequencies and percentages for explaining; a) demographic profile of respondents; and b) data regarding the level of academic qualifications and experience of PGB; and ii) analysis of the EFA, CFA and SEM to identify the correlation between the variables and the effectiveness of the coaching. Descriptive a) The demographic profile of the respondents indicated a total of 98 PGB (46.2%) were males and 114 PGB (53.8%) were females. Of these, fraction in terms of race, 185 PGB (87.3%) are Malays, 19 PGB (9.0%) are Chinese and 8 PGB (3.8%) Indians. In general, most of the respondents (PGB) are over the age of 41 years, a total of 206 respondents, a high percentage of 97.2%. This proves that the positions of PGB indeed have been held by those who have served long and have extensive experience in the education system. It was found that only four (1.9%) respondents of PGB are between the ages of 31-35 years, while two others (0.9%) aged between 36-40 years. Of the 212 respondents, 201(94.8%) of them are married while only 11 (5.2%) are single. Table 1.2 presents a summary of respondents' demographic data. b) Change in the education era and the importance of seeking knowledge can be seen when a total of 104 (49.1%) respondents have a Bachelor degree, while as many as 48 respondents (22.6%) had Masters. A total of 17 (8.0%) PGBs have qualification at Diploma level and 43 Research in Social Sciences and Technology (RESSAT) 2018: 3(1), 16-35 of respondents (20.3%) are ‘Sijil Tinggi Pelajaran Malaysia’ (STPM) and ‘Sijil Pelajaran Malaysia’ (SPM) holders. Total respondents holding professional qualifications showed that 46 (21.7%) have a first degree in education. Only 8 (3.8%) of them have a Master's degree in education, while 54 respondents (25.5%) have a Diploma in Education and the remaining 104 (49.1%) have only a Teaching Certificate. Table 1.2: Summary of Respondents Demographic Data Demografic Respondent Frequency Percentage Gender Male Female 98 114 46.2 53.8 Total 212 100.0 Race Malay Chinese Indian Others 185 19 08 00 87.3 9.00 3.80 0.00 Total 212 100.0 Age 31-35 years 04 1.90 36-40 years 02 0.90 41 and above 206 97.2 Total 212 100.0 Marital Status Single Married 11 201 5.20 94.8 Total 212 100.0 A total of 191 respondents (90.6%) are educators who have been teachers for 16 to more than 21 years, while 20 respondents (9.4%) have experienced as educators between 6-15 years. However, experience as the head of school administrators are found to be almost a balanced distribution. Respondents who have experienced administering more than 10 years were 65 (30.7%), while 86 respondents (40.6%) have experienced being administrators between 6-10 years. A total of 61 respondents (28.8%) experienced administering other schools less than five years. Considering the size of the Federal Territory of Kuala Lumpur is not too large if compared to other states in this country, the ability to move from home to school is less than 25 kilometers one way. One-way distance from home to school for 41 respondents (19.3%) were between 1-5 kilometers, while 52 of the respondents (24.5%) live between 6-10 kilometers from the school. However, it was found 119 respondents (56.1%) live more than 10 kilometers away from the school. Although the distance is not an issue to senior administrators, other issues Tahir & Salleh such as traffic congestion in the city has resulted in longer time to get to school. Table 1.3 presents a summary of the qualifications and experience of respondents in administering their schools. Table 1.3: Summary of the qualifications and experience of respondents Respondent Frequency Percentage Academic Qualifications (General) SPM/STPM Diploma Bachelor Master 43 17 104 48 20.3 8.0 49.1 22.6 Total 212 100.0 Professional Qualifications Teaching Certificate Diploma in Education Bachelor in Education Masters in Education 104 54 46 08 49.1 25.5 21.7 3.8 Total 212 100.0 Experience as teacher 6-10 years 11-15 years 11 09 5.2 4.2 16-20 years 01 0.5 21 years and above 191 90.1 Total 212 100.0 Experience as administrator 1- 5 years 6-10 years 11 years and above 61 86 65 28.8 40.6 30.7 Total 212 100.0 Distance to school 1-5 km 6-10 km More than 10 km 41 52 119 19.3 24.5 56.1 Total 212 100.0 Analysis of EFA and CFA a) Exploratory Factor Analysis (EFA) Based on the sample size (n=212) the cronbach alpha values for the three constructs reached between 0.950 to 0.979 (Table 1.1). However, when statistical Exploratory Factor Analysis (EFA) is used, the results obtained changed because this analysis produces strength for each construct and its position in these constructs. This analysis was also conducted to minimize data through the cleaning process of the collected data. The main thing to remember is that every item in the construct has a high loading factor value, while with lower value the item will be dropped. The loading factor value of 0.5 or more is considered significant while Research in Social Sciences and Technology (RESSAT) 2018: 3(1), 16-35 values above 0.70 is considered as very appropriate and has a clear structure (Hair, et. al., 2006). Therefore, any item that has a loading factor value of less than 0.50, the item was dropped from the instrument to produce good data. Based on Monte Carlo Simulation Study (Rubinstein, R. Y. and Kroese, D.P., 2007) it was suggested that the appropriate analysis is to maintain the construct that has four or more items in it and each item has a loading value exceeding the reading of .40. However, the researchers took a position by using items that meet the characteristics suggested by Hair et. al. (2010) and Zainudin (2012), namely: • The loading of each item must exceed 0.50. • Each dimension or construct will be maintained if they have more than three items with high loading factor value. • The Bartlett's test is significant at .01 (p <.001) • The Kaiser-Meyer-Olkin (KMO) must exceed .60 • Significant must be p = 0.000 • Communality exceeds .60 Analysis Part B involves three constructs of the relationship between the PGB (coachee) with SIP (coach), professionalism skills and effectiveness of the coaching that includes 31 items. These constructs are as follow: i) Construct of PGB In Relationship With SIP One of the key assumptions evaluated in the analysis of the EFA is to test between correlations (multicollinearity) of items. There are 8 items tested in the construct of relationship with SIP. Bartlett's test statistically showed the item is significant to the value of χ2 (28) = 2158.467 and p=.000. Value of Kaiser-Meyer-Olkin (KMO) generated is more than .600 (.918), while the value of Measure of Sampling Adequacy (MSA) on each item is more than .60. The next step is to assess the reading of communality that should exceed .60. During a test carried out on rotated component matrix, the complexity factor is applicable to an item (L8). Tahir & Salleh This situation led researchers to drop the item from the constructs even when the value of KMO is higher. The same process of EFA is carried out once again and have produced results with KMO = .914 Bartlett's test showed statistically the item is significant with the value of χ2 (21) = 2086.211 and p=.000. Thus, only seven (7) items are taken for analysis at the next stage. Based on Table 1.4 shows the EFA analysis of the construct 'PGB Relationship With SIP'. This construct produces only one implied dimension based on eigenvalues data greater than 1.0. This finding also demonstrates the said dimension is able to justify more than 84.30% of seven (7) items contained in the overall construct. Table 1.4: Eigenvalues and the Percentage of Variance for the Construct of PGB Relationship with SIP Initial Eigenvalues Component Total % of Variance Cumulative % 1 5.902 84.307 84.307 Total Variance Explained: 84.307% ii) Construct of Professionalism Skills of SIP There are 11 items in the construct of ‘Professionalism Skills of SIP'. Bartlett's test statistically showed that the item is significant to the value of χ2 (55) = 2767.867 and p=.000. The value of Kaiser-Meyer-Olkin (KMO) produced is more than .600 (.893), while the value of Measure of Sampling Adequacy (MSA) of each item is also exceeding .60. When the value of Bartlett's and KMO meet the test conditions, then the next step is to assess the reading of communality that should exceed .60. During the test on rotated component matrix was carried out, no complexity factor occurs and no item is dropped. Thus, all 11 items are accepted to be analyzed at the next level. Table 1.5 shows the analysis of the EFA for 'Professionalism Skills of SIP' construct. This construct also produces only one implied dimension based on eigenvalues data greater than 1.0. This finding also demonstrates the said dimension is able to justify more than 79.65% of seven (7) items contained in the overall construct. Research in Social Sciences and Technology (RESSAT) 2018: 3(1), 16-35 Table 1.5: Eigenvalues and the Percentage of Variance for Professionalism Skills Construct Initial Eigenvalues Component Total % of Variance Cumulative % 1 8.762 79.655 79.655 Total Variance Explained: 79.655% iii) Effectiveness Coaching Construct There are 12 items in the construct of ‘Effectiveness of Coaching’. Bartlett's test statistically showed that the item is significant to the value of χ2 (66) = 4565.235 and p=.000. Value Kaiser- Meyer-Olkin (KMO) generated is more than .600 (.829). The Measure of Sampling Adequacy (MSA) on each item is more than .60. Communality values also exceeded the reading of .60. During the test runs on rotated component matrix, no complexity factor occurred and this indicates that all items received one construct. Thus, the researchers maintain all these items. The same EFA process is carried out once again and has produced similar results. So the researchers maintain 12 items to be analysed at the next level. Table 1.6 shows the analysis of the EFA to construct 'Effectiveness of Coaching'. This construct also produces only one implied dimension by eigenvalues greater than 1.0. This finding also explains that they are able to explain 81.42% of seven (12) items contained in the overall construct. Table 1.6: Eigenvalues and the Percentage of Variance for Effectiveness of Coaching Initial Eigenvalues Component Total % of Variance Cumulative % 1 9.77 81.425 81.425 Total Variance Explained: 81.425% In conclusion, although the KMO value obtained is above .60 and the MSA of each item is greater than .60, the possibility of items dropped still can be occurred. This is because the rotated component matrix test will determine the items in one or multi-dimensional. These findings will lead to the decrease in the number of items despite the higher value of the MSA. This means that each variable can be estimated exactly and meet the criteria suggested at the Tahir & Salleh same time. Finally, after going through the process of EFA, the total number of items in Part B was reduced to 30 items (Table 1.7). Table 1.7: Reliability Value and Number of Items in Every Construct Construct No. of Item Cronbach Alpha PGB relationship with SIP 7 items 0.966 SIP Professionalism Skills 11 items 0.947 Effectiveness of Coaching 12 items 0.978 TOTAL 30 items 0.963 b) Confirmatory Factor Analysis (CFA) with the Measurement Model Having found the relationship among the three constructs are strong with high cronbach alpha value, the CFA is then used to test the appropriateness and validity of the measurement model proposed. CFA is used to validate the constructs that have gone through the process of EFA. The validity is measured to determine the item is actually representing a specific construct. Item confirmation can be made by using a diagnostic model that involves a few key points to be tested (Hair, et. al. 2006). Next, the CFA is also used to explore the item if there is a significant modification that needs to be made before the overall construct is combined into a structural model (Hair, et. al., 2006). Table 1.8 shows the diagnostic model in the CFA analysis process suggested by some researchers and suitable as a reference. Table 1.8: Diagnostic Model In CFA Process Diagnostic Model Requirement Reference Modification Index >4 >10 Hair, et. al., 2006 Fassinger, 1987 Standardised Residual <2.5 –no problem >4.0 –possible problem Hair, et.al., 2006 Squared Multiple Correlations or Reliability >0.3 Hair, et.al., 2006 Standardised Coefficients <1.0 Min & Mentzer, 2004 In all three constructs (PGB relationship with SIP, Professionalism Skills and Effectiveness of Coaching) used in this study resulted in a one implied dimension, the CFA is used to make the confirmation. The construct validity was measured to determine the extent to which the items produced truly represent the resulting construct. CFA analysis was conducted using AMOS Research in Social Sciences and Technology (RESSAT) 2018: 3(1), 16-35 software version 22. Each model constructed is considered fit when the index reading reached the proposed stage (Hair, et. al., 2006), namely: i) The value of Chisq/df between 2 and 5 is considered good and acceptable fit for the model and the data used. ii) Comparative Fit Index (CFI), which exceeds the threshold value of 0.90 indicating a good fit and appropriate, and; iii) Root Mean Square Error of Approximation (RMSEA) of less than 0.08 indicates an estimated error that is reasonable and acceptable. Browne and Cudeck (1993) suggested a reading of between 0.08 to 0.1 is also considered reasonable and acceptable. Since all these constructs are one dimension, then the construct is in the form of first order and the method of combining all of the constructs can be made. This combination method is highly recommended method of measuring the measurement model using the CFA simultaneously (Zainudin, 2012). A total of 30 items of three variable constructs have reached a cronbach alpha above 0.90 with the correlation among the constructs of less than 0.85. Figure 1.1 shows the analysis of standardized hypothesized Measurement Model that combines all of the constructs. Tahir & Salleh Figure 1.1: Hypothesized Measurement Model Model of this construct is tested and the results showed the value of Chisquare/df= 8.88, CFI=.715, P=.000 and RMSEA= .193. All the fitness indexes in these constructs did not meet the proposed criteria, although the loading factor exceeds .50, Squared Multiple Correlation (SMC) exceeded .3 (Hair et. al., 2006). Thus the model needs to be revised and combined with free parameter estimate carried out to obtain better covariance reading. Items with less than .50 loading factor will be dropped and only the best fitness indexes will be retained and used for the next process. Figure 1.2 shows the Revised Measurement Model after the 15 items from all three constructs are dropped. Items dropped are 3 items from the ‘Relationship’ constructs (L1, L2, L7), 5 items from the ‘Professionalism Skills’ constructs (P2, P3, P4, P7 and P10) and 7 items from the ‘Effectiveness of Coaching’ constructs (M1, M2, M3, M4, M5, M6 and M8). Research in Social Sciences and Technology (RESSAT) 2018: 3(1), 16-35 Figure 1.2: Revised Measurement Model After the test is carried out repeatedly by dropping an item at a time, eventually the values of the fitness indexes fit the value of Chi-square / df = 2.203, CFI = .969, P=000 and RMSEA= .076. The results of these tests show that each item in each construct tested has its own strength. No items are redundant that require a combination of the free parameter estimate except for M12 and M11 which are combined. After a combination of these two items is made, the value that results from the test statistically showed that the standardized regression weight is significant and fulfills the rules of thumb (Hair, et. al., 2006) as the validity of all constructs showed a correlation value less than 0.85. This relationship can be seen when the correlation between the constructs of 'Effectiveness and Relationship' is r=.84, while the correlation between the constructs 'Professionalism and Effectiveness' is r=.69 and the correlation between the constructs 'Relationship and Professionalism' is r=.76. Table 1.9 shows the standard error (SE), the critical ratio (CR) and p-value. All items of each construct tested are considered good by fulfilling the proposed rules of thumb. Tahir & Salleh Table 1.9: Value of Regression Weights for Each Item in Construct Regression Weights: (Group number 1 - Default model) Estimate S.E. C.R. P L6 <--- Relationship 1.085 .030 36.532 *** L5 <--- Relationship 1.000 L4 <--- Relationship .996 .030 33.762 *** L3 <--- Relationship 1.023 .052 19.601 *** M11 <--- Effectiveness 1.117 .040 27.609 *** M10 <--- Effectiveness .760 .062 12.275 *** M9 <--- Effectiveness 1.073 .045 23.663 *** M7 <--- Effectiveness 1.001 .041 24.343 *** P9 <--- Profesionalism 1.063 .090 11.840 *** P8 <--- Profesionalism .957 .093 10.297 *** P6 <--- Profesionalism .774 .078 9.909 *** P5 <--- Profesionalism .663 .084 7.897 *** M12 <--- Effectiveness 1.000 P1 <--- Profesionalism 1.050 .084 12.568 *** P11 <--- Profesionalism 1.000 Findings - Structural Equation Modeling (SEM) In answering the research objectives, the structural model tests involving the analysis of the independent variables (exogenous) with the dependent variable (endogenous) are carried out. Figure 1.3 structural models are built to test the relationship between the independent variables to the dependent variables. There are two exogenous constructs, construct of 'Relationship' which has four items (L3, L4, L5 and L6) and the construct of 'Professionalism Skills' which has six response items (P1, P5, P6, P8, P9 and P11). Next endogenous construct consists of one construct 'Effectiveness' which has five items (M7, M9, M10, M11 and M12). Based on standardized regression weights which measure beta estimate and the standard deviation, it is found that 'Relationship' construct will affect the 'Effectiveness of Coaching' with reading of β=0.73, while the construct 'Professionalism Skills' will affect the 'Effectiveness of Coaching' with reading of β=0.14. A very good correlation is shown among Research in Social Sciences and Technology (RESSAT) 2018: 3(1), 16-35 the constructs because there is no similarity (mirror) in understanding. Construct 'Relationship and Professionalism' has a value of r = .76 which less than .85. Figure1.3: Structural Model of SIPs’ Effectiveness of Coaching Referring to the Squared Multiple Correlations (SMC) for the construct of 'Effectiveness of Coaching' is R2 = 0.71, which means that the construct of 'Relationship' and construct of ‘Professionalism Skills' contribute 71% to the effectiveness of SIP coaching to PGB in schools. SMC value for each item was among the lowest of R2=0.31 (P5) to the highest value of R2 = 0.94 (L6). All of the SMC show the contribution of each item is between 31.0% to 94.0%, and this shows the contribution and strengths of those items in the respective constructs. Similarly the correlation between the constructs of all independent variables have a relationship and lower than the value of 0.85 which r=.76. Tahir & Salleh Overall the suitability index values obtained in this structural model is good and achieve the desired level with reading of chi-square/df = 2.203, CFI=0.969, p= 0.000 and RMSEA=0.076. All items in the independent variables have played a role in the relationship (causal relationship) contributing their individual strength to the dependent variable. Based on the analysis of the findings, it can be concluded that the effectiveness of the coaching of SIP towards PGB is due to the good rapport and professional skills during the coaching sessions. These findings have also fulfilled the research objectives. Conclusion and Implication The results of this study showed that although most administrators are among those aged over 40 years (97.2%) and 40.6% of them had held this position between 6-10 years the cooperation and guidance of outsiders expertise is still needed. The study confirmed that the overall good relationship skills and professionalism skills of SIP has contributed to the effectiveness of the coaching so as to increase confidence in the administration of the PGB. The findings also illustrate that the PGB and SIP should continue to work together to strengthen the management and administration of the school in order to achieve the objectives and goals set. The chance of PGB getting guidance from coaches should be seen as a catalyst to force a shift towards excellence. Thus, the top management through the District Transformation Program (DTP) as the District Education Office (PPD) and the State Education Department (JPN) should understand and support the need for this coaching. This is because the SIP has actually acted as a coach, mentor and counselor which will be continued until the school achieved excellence within the timeframe stipulated. It is proposed that a study on the ability of SIP and PGB to transform education in schools be carried out in the future. It is intended that the school leadership capabilities be enhanced continuously which in line with the government’s vision through the Malaysia Education Blueprint (MEB (PPPM), 2013-2025). Research in Social Sciences and Technology (RESSAT) 2018: 3(1), 16-35 References Azlin Norhaini Mansor. (2006). Amalan Pengurusan Pengetua Sekolah Menengah: A Case Study. (Principal Management Practices in Secondary School) PhD Thesis, Universiti Kebangsaan Malaysia. Chua Yan Piaw (2006). Kaedah Penyelidikan (Research Methodology). Kuala Lumpur: McGraw Hill Sdn. Bhd. Chua Yan Piaw (2009). Statistik Penyelidikan Lanjutan Ujian Regrasi, Analisis Faktor dan Analisis SEM (Advanced Research Statistic of Regression, Factor Analysis and SEM). Kuala Lumpur. Mc Graw Hill Education. Dev, S., Qayyum, 2017. Students Perception towards faculty evaluation on Teaching (SET)., Journal of Social Studies Education Research., Volume 8 (3),149-167 Education Department of Federal Territory of Kuala Lumpur (2017). Statistik Guru dan Murid di Sekolah Menengah Kebangsaan Kuala Lumpur (Teachers and Students Statistic at Secondary School in Kuala Lumpur). Fraenkel, J.R. & Wallen, N.E. (2007). How To Design And Evaluate Research In Education. 7th ed. New York: McGraw Hill Hair, Jr. J.F., Black, W.C., Babin, B.J., Anderson, R.E. & Tatham, R.L. (2006). Multivariate Data Analysis. 6th ed. New Jersey: Pearson Prentice Hall. Hair, Jr. J.F., Black, W.C., Babin, B.J., Anderson, R.E. (2010). Multivariate Data Analysis. 7th ed. Englewood Cliff, New Jersey: Pearson Prentice Hall. Huda, B. Z., Rusli, B. N., Naing. L., Tengku, M. A., T. Winn, & Rampal, K. G. (2004). A Study Of Job Strain And Dissatisfaction Among Lecturers In The School Of Medical Sciences, USM. Southeast Asian J Trop med Public Health, 35(91), 210-216. Karim, N. H. A. (2008). Investigating the Correlates and Predictors of Job Satisfaction Among Malaysian Academics Librarians. Malaysian Journal of Library & Information Science, 13(2), 69-88. Likert, R. (1932). The Technique for the Measurement of Attitudes. Archieves of Psychology. 140-152. Mayer, E. (1992). The key competencies report: ‘Putting education to work’. Australia Education Council and Ministers for Vocational Education Employment and Traning, Melbourne. Tahir & Salleh Ministry of Education of Malaysia (1996). Education Act 1996, MOE. Ministry of Education of Malaysia (2013). Pelan Pembangunan Pendidikan Malaysia (Malaysian Education Blueprint), KPM Kuala Lumpur. Mohd Majid Konting (1990). Education Research Method (Kaedah Penyelidikan Pendidikan). Edisi ke-2. Kuala Lumpur: Dewan Bahasa dan Pustaka. Narimawati, S.E.U. (2007). The Influence of Work Satisfaction, Organizational Commitment and Turnover Intention Towards the Performance of Lecturers at West Java’s Private Higher Education Institution. Journal of Applied Science Research, 3(7), 549-555. Pallant, J. (2010). A Step By Step Guide To Data Analysis Using SPSS program. 4th ed. New York: Mc Graw Hill. Shaver, J.P. & Norton, R.S. (1980). Randomness and Replication in Ten Years of the American Educational Research. Educational Researcher, 9 (1), pg. 9-15. Shuib Dahaban (2006). The Effectiveness of Leadership Towards Changing in Education (Kepimpinan Berkesan Ke Arah Perubahan Dalam Pendidikan). Kuala Lumpur: Yayasan Guru Malaysia. Tarman, B. (2012). Effective leadership in culturally diverse schools. Energy Education Science and Technology Part B: Social and Educational Studies, 4(2), 1103-1114. Yigit, M. F., & Tarman, B. (2016). How do different ethnicities approach to the education system and differences in Turkey? Italian Sociological Review, 6(3), 339-353. 10.13136/isr.v6i3.119 Yin, R. K. (2003). Case Study Research; Design and Methods. 3rd ed. Thousand Oaks: Sage. Zainudin, A. (2012). Research Methodology and Data Analysis. 2nded. Shah Alam: Universiti Teknologi MARA Publication Centre (UiTM Press). Zainudin, A. (2012). Structural Equation Modeling Using AMOS Graphic. Shah Alam: Universiti Teknologi MARA Publication Centre (UiTM Press). Zalina Mohd Tahir (1997). Occupational Stress Among Teachers At Secondary School In Malaysia. Master Thesis, University of Aberdeen, United Kingdom.