Factors Affecting Employees’ Turnover Intention in Construction Companies in Klang, Selangor

FGIC2019 FGIC 2nd Conference on Governance and Integrity 2019 Volume 2019 Conference Paper Factors Affecting Employees’ Turnover Intention in Construction Companies in Klang, Selangor Suhaidah Hussain and See Huei Xian Faculty of Industrial Management, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Gambang, Pahang, Malaysia Abstract Corresponding Author: Suhaidah Hussain suhaidahhussain@ump.edu.my Received: 5 August 2019 Accepted: 14 August 2019 Published: 18 August 2019 Publishing services provided by Knowledge E Suhaidah Hussain and See Huei Xian. This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the This research is about factors affecting employees’ turnover intention in construction companies. Employees’ turnover intention is known as the organization’s workers’ intent or plan to leave their current working place’s position. Malaysia has scored third highest voluntary turnover rate, which is 9.5% in Southeast Asia year 2015. Most of the construction projects are difficult and complex to manage it. High employees’ turnover rate may influence the construction companies’ productivity and performances. There are many factors that will affect employees’ turnover intention, such as colleague relations, organizational commitment, organizational justice, organizational reputation, communication, and organizational politics. In order to address the issues above, this research was aims to identify the factors affecting employees’ turnover intention and to determine the relationship between the factors and employees’ turnover intention. Therefore, in order to achieve these objectives, a questionnaire survey involving 160 employees conducted to Grade 7 construction company in Klang, Selangor. There was 73 companies’ worker who responded to the survey. The data analysis conducted using SPSS and SmartPLS, and the results showed that organizational politics were mostly caused employees’ turnover intention in construction companies. The findings also showed that communication and organizational politics had a negative relationship with employees’ turnover intention. Results from this research can provide the evidence and bring convince for the construction companies in Malaysia to reduce employees turnover rate. In the future, the scope of the study can be expanded to other states of Malaysia to improve the reliability of this study. Keywords: Employees, turnover intention, construction company original author and source are credited. Selection and Peer-review under the responsibility of the FGIC2019 Conference Committee. 1. Introduction Based on Jayaram (2015), Malaysia scored third highest voluntary turnover rate at 9.5% in Southeast Asia year 2015. Automatic Data Processing (ADP) Workforce Vitality Index 2017 has reported that there are high employees’ turnover problems, which affect construction companies and contractors but unnoticed by them (Palaganas, 2018). In How to cite this article: Suhaidah Hussain and See Huei Xian, (2019), “Factors Affecting Employees’ Turnover Intention in Construction Companies in Klang, Selangor” in FGIC 2nd Conference on Governance and Integrity 2019, KnE Social Sciences, pages 108–131. DOI 10.18502/kss.v3i22.5047 Page 108 FGIC2019 the previous studies such as Thomas (2015), Seong (2015) and Nkomo et al. (2009) showed that salary and fringe benefits are the key factors for employees’ turnover intention in construction companies. Although previous studies focused on factors employees’ turnover intention, there is a lack of investigation of factors affecting employees’ turnover intention in construction companies in Malaysia. Therefore, in this research, it will identify the factors affecting employees’ turnover intention and determine the relationship between the factors and employees’ turnover intention in construction companies in Klang, Selangor. Factors such as colleague relations, organizational commitment, organizational justice, organizational reputation, communication, and organizational politics play an important role in employees’ turnover intention in construction companies in Klang, Selangor. By knowing the factors causing employees’ turnover can help human resource manager in the company to look for better solutions to reduce employees’ turnover rate and thus can increase productivity in construction companies in Klang, Selangor. Therefore, it is very important to find out what factors will cause employees’ turnover intention in construction companies in Klang, Selangor. 2. Literature Review 2.1. Background of the construction industry The construction industry is the field where those companies are responsible for the buildings or engineering projects of the construction. It actually consisted of three types of construction projects which are residential projects, non-residential projects, and engineering projects. Besides that, the construction industry plays a very important role in the socio-economic growth of each country (Rahman, Memon, & Abd. Karim, 2013). Economically, it contributes to major improvement in the overall Gross Domestic Product (GDP) of a country. 2.2. Construction company in Malaysia The construction company is is sector in the Malaysian economy because it contributed 3% to 5% of the aggregate economy GDP over the last 20 years (Khan, Liew, & Ghazali, 2014). Malaysian construction company nowadays is more high-ranking, modernized, and well equipped. As a developing nation, Malaysia has understood the vital role of the construction company not only help in economic growth but also improving the living DOI 10.18502/kss.v3i22.5047 Page 109 FGIC2019 standards and life’s quality of citizen in Malaysia. For Malaysia’s construction company, all contractors must register with the Construction Industry Development Board (CIDB). In tear 1994, Malaysian Government was setting up the CIDB to educate those construction company, to adjust and register the construction companies in seven grades by following the standard given (M.N.A, Majid, Ahamad, & Hanafi, 2012). 2.3. Factors affecting employees’ turnover intention 2.3.1. Colleague relations Colleague relations means the relationship with the co-worker in the organization. The connection with colleagues can also be defined as the social and working undertaken with other people in the workplace. Relationship with the colleague in the workplace can be observed through its attributes such as trust, cooperation, and so on (Buljubasic, 2008). The relationship between the colleagues may influence the working environment. This is because employees will spend a longer time with their colleague during working hours. If employees have a good relationship with their colleague may make a pleasure of the work environment. 2.3.2. Organizational commitment The organizational commitment may interfere with the social and personal functions of employee and the effective operation of the organization. It is also an expression of employees’ feeling of psychological attachment towards the organization. It can measure the employees’ willingness to stay in an organization in the future. There are three components of commitment that can be defined as an organization which is identification, organization, and honesty (Raisiene & Vilke, 2014). The organizational commitment is rising from the organization’s principled attitude toward an employee. It also can be described as the involvement of employees in the organization. 2.3.3. Organizational justice According to Greenberg (1987), organizational justice defined as the fairness of resource allocation in an organization based on employee’s perception. This is also referring to decision by management and its action, whether it is according to the moral right with ethical standards, religion, or the law. The fairness can describe in term of employees’ DOI 10.18502/kss.v3i22.5047 Page 110 FGIC2019 salary, performance evaluation, opportunities for promotion, and others. This perception, in turn, can influence employees’ attitude towards management (Fee Yean & Yusof, 2016). If fairness is applied in an organization, employees will also feel satisfied and give a high commitment to the organization. 2.3.4. Organizational reputation Organizational reputation is same meaning as an organizational image which shows the overall attractiveness of the organization. The reputation of an organization will represent its ability to create value, and it passes actions. The good pass actions of an organization will have a good reputation or image in the country, which will make those employees feel proud to work in the organization. Organizational reputation also can be considered as the stakeholder’s opinion or judgment. Organizational reputation demonstrates the ability of the organization to create value. The reputation of the organization is also based on the past behavior of the organization and gives the outlook for the future (Hendriks, 2016). 2.3.5. Communication Communication means the willingness to share non-personal or personal information to others (Eisenberg & Witten, 1987). It also can be defined as the sharing of information between one and others through face-to-face interaction or social media. Communication is an important aspect that helps to improve the decision-making process. Through the exchange of information between manager and employees can choose the best decision to solve any of the problems faced by the organization. At the same time, employees will feel their contributions are being respected and it may reduce the dissatisfaction of employees toward their job. 2.3.6. Organizational politics Organizational politics can be defined as the work behavior of employees in the organization. It is a difficult element to deal with within the organization. Even though it will bring some negative effect on the organization, but it may also bring a positive impact to the organization. The negative impact that organizational politics will cause is ethical concerns such as conflicts over resources. Ferris et al. (1996) view organizational politics as an attitude that is not formally authorized by the organization, dispute, and DOI 10.18502/kss.v3i22.5047 Page 111 FGIC2019 frictions arise in the work environment by opposing individuals or groups against each other or peer-to-peer organizations. On the other hand, those employees with political skills are managed job stress and help the organization meet the strategic goals. 2.4. Employees’ turnover intention Employees’ turnover intention means the employees have thought or plan to voluntarily leave their current organization. Voluntary employee turnover explains that employees’ perception can lead to turnover. Employees’ turnover intention also closely related to the employees’ turnover because the actual employees’ turnover behavior is depending on their intention of leaving the organization (Yang and Wittenberg, 2016). Thus, the employees’ turnover intention is the most immediate determinant of real turnover behavior. Employees’ turnover intention must be taken seriously because it will influence the organizational performance and lower down the efficiency of productivity. 2.5. Relationship between factors and employees’ turnover 2.5.1. Colleague relation & employees’ turnover intention Colleague relationships have a direct negative impact on job stress and realism, while executive support has a direct negative impact on job stress but not realism (Yang et al., 2015). The results of the previous study show that the pressure on employees in the workplace has become increasingly important because it will influence the employees’ turnover intention. Consolidated colleagues and executive support can effectively reduce work stress and improve the status of the aging workforce. At the same time, the employee’s willingness to turnover can be reduced. According to Bateman (2009), it also is shown that there is a significant negative correlation between colleagues’ support and withdrawal intention. Retail organizations should be aware of this relationship and commit to creating a supportive work environment. Thus, it can be hypothesized that colleague relations have a negative relationship with employees’ turnover intention. 2.5.2. Organizational commitment & employees’ turnover intention Organizational support can boost the morale of the employee; as a result, will lead to high organizational commitment and low employees’ turnover intention (Ahmed and Nawaz, 2015). This previous research approves the significant positive relationship DOI 10.18502/kss.v3i22.5047 Page 112 FGIC2019 between job autonomy and organizational commitment which the greater job autonomy leads to more commitment to the current organization needs to provide more autonomy to employees so that employees could decide about their goals and pursue plans to achieve them. Through Jehanzeb, Rasheed, and Rasheed (2013) research on organizational commitment and employee turnover intentions also confirms a strong reverse relationship. Therefore, it can be hypothesized that organizational commitment has a negative relationship with employees’ turnover intention. 2.5.3. Organizational justice & employees’ turnover intention In the previous scholar by Ali and Jan (2012), there was a significant relationship between employee’s turnover intention towards distributive justice perception as well as procedural justice perception. The degree of organizational justice and job satisfaction in the workplace has an effect on employees’ attitude and performance (Letchumanan, Apadore, & Ramasamy, 2017). If the organization can recognize and implement organizational justice, the employee’s performance and productivity can be increased if they been treated fairly. Thus, employee turnover also can be reduced. In other words, this can be hypothesized that organizational justice has a negative relationship with employees’ turnover intention. 2.5.4. Organizational reputation & employees’ turnover intention Findings in the Allahyary and Beheshtifar (2013) and Alniacik, Cigerim, Akcin, and Bayram (2011) show that there is a reverse relationship between organizational reputation and employees’ turnover intention. Focusing on organizational identification and images can reduce employee turnover intentions in the organization. Previous research in South Korea conducted by Kwon & Rupp (2013) also presented a negative correlation between corporate reputation and turnover intentions, especially in high-performing companies in the organization. A high degree of recognition of the company’s reputation will reduce the willingness to leave, and vice versa. So, it can be hypothesized that the organizational reputation has a negative relationship with employees’ turnover intention. 2.5.5. Communication & employees’ turnover intention According to Olcer and Ozenir (2017), their research found out that organizational communication has a negative effect on the intention to leave of employees. The research DOI 10.18502/kss.v3i22.5047 Page 113 FGIC2019 did by Al-Tokjais (2016) also results that good communication can forecast turnover intention behavior of employees in an organization. It is proven that communication from manager to employee (downward communication), communication employee to their leader (upward communication) and communication between departments (sideward communication) in an organization will influence for the turnover intention. It is expected that a high perception of communication will reduce the turnover intention occurs in the organization (Yang & Wittenberg, 2016). Thus, it can be hypothesized that communication has a negative relationship with employees’ turnover intention. 2.5.6. Organizational politics & employees’ turnover intention The results from Abubakar and Abdullahi (2017) study also shows that there would be a reciprocal effect of perceived organizational politics on turnover intention. These results underline our postulation that the relationship between perceived organizational politics and turnover intention is complex and reciprocal. This suggests that there is no simple one-directional effect of perceived organizational politics on turnover intention. Whereas perceived organizational politics influence turnover intention, employee turnover can also influence the perception of organizational politics in the work environment. Therefore, it can be hypothesized that organizational politics has a negative relationship with employees’ turnover intention. 3. Methodology 3.1. Research design The quantitative research design was used in this study. Descriptive and correlational design were some of the types of quantitative research design which would be used to solve the research questions. The descriptive design includes the sample population which may be nearly hundreds or thousands of subjects to achieve the validity estimate, and the Correlational design will be used to determine and test for the relationship between independent and dependent variables of this research. The data collection method that would be used in this study was questionnaires method. Both paper-pencilquestionnaires and web-based questionnaires would be applied in this research. Hence, the questionnaires would be prepared and given to the respondents through email or walk-in to the Grade 7 construction companies in Klang, Selangor. DOI 10.18502/kss.v3i22.5047 Page 114 FGIC2019 3.2. Conceptual framework The following Figure 1 is the conceptual framework of the research. Independent Variables Colleague Relations H1 Organizational Commitment H2 Organizational Justice H3 Organizational Reputation H4 Communication H5 Organizational Politics H6 Dependent Variable Employees’ Turnover Intention Figure 1: Conceptual Framework (Source: Authors’ own work). 3.3. Research hypothesis Based on the Figure 1, the hypotheses for this research as follows: H1: The colleague relations have a negative relationship with employees’ turnover intention. H2: The organizational commitment has a negative relationship with employees’ turnover intention. H3: The organizational justice has a negative relationship with employees’ turnover intention. H4: The organizational reputation has a negative relationship with employees’ turnover intention. H5: The communication has a negative relationship with employees’ turnover intention. H6: The organizational politics has a negative relationship with employees’ turnover intention. DOI 10.18502/kss.v3i22.5047 Page 115 FGIC2019 3.4. Data collection In this study, all the data and information will be collected through the articles and journals which were using secondary data collection method and questionnaire survey forms that were distributed which were using the primary data collection method. The main data collection techniques used in this study was a questionnaire from the target respondents through email and walk-in. Questionnaire format such as close-ended questions, length of the questions, and the choices of the questions had been set to make sure the data can easily gather from the respondents. In this questionnaire design process, the questionnaire separated into three sections, which were Section A: Demographic, Section B: Factors affecting employees’ turnover intention, and Section C: Employees’ turnover intention. Objective questions, subjective questions, and Five Point Likert scale were used in the questionnaire. The levels of measurement used for the variables in this research was ordinal scale because the Five Point Likert scale that uses in the questionnaire such as strongly disagree, disagree, neutral, agree and strongly agree were an ordinal scale of measurement. 3.5. Population and sample size The respondents in this study will focus on the upper or lower level of employees who were working in Grade 7 construction companies in Klang, Selangor. There had a total population of 160 Grade 7 construction companies in Klang, Selangor found from the Centralized Information Management System (CIMS) that had registered for Construction Industry Development Board (CIDB). Based on Krejcie and Morgan (1970), the sample size for the population of 160 Grade 7 construction companies in Klang, Selangor were 113 of respondents. Those respondents for this research will involve all different gender and age. Besides that, probability sampling was used in this research because the respondents are chosen through random sampling. Each of the respondents has the same probability of being chosen, and they will be randomly chosen from the population. 3.6. Data analysis The Pearson Correlation and Multiple Regression Analysis have been conducted to identify the factors affecting employees’ turnover intention and determine the relationship between factors and employees’ turnover intention in construction companies. Before conducting the Correlation and Regression Analysis, reliability and validity test DOI 10.18502/kss.v3i22.5047 Page 116 FGIC2019 have done in order to make sure whether the Five Point Likert scale questions in the research questionnaire are reliable and the data collection in this research are valid. Moreover, the normality test also has done in order to confirm the data collected are normally distributed and achieved the underlying assumption for most of the statistical analysis. Descriptive analysis has been conducted to summarized and describe the data collected by measure the mean and standard deviation to make the results of the research easier to understand. 3.7. Survey procedure 3.7.1. Pre-test The pre-test will be conducted to get some advice from the experts to improve the questionnaire. In this research, the questionnaire was given to three experts who have the related field or background toward construction or human resources to undergo this pre-test. Those experts have identified the questionnaire’s problems such as unclear sentences or the questionnaire taking too long to administer. The comments from the experts had been recorded down, and correction is done at the same time before proceeding to the pilot test. 3.7.2. Pilot test The pilot test is a rehearsal of the research study which allows testing the study approach with a small number of test respondents before the main study being conducted and ensuring all the respondents understands the questions from the questionnaire in the same way. Based on Connelly (2008) and Hertzog (2008), 12 questionnaires were given to the workers who work in construction companies in Malaysia other than Klang areas to conduct the pilot test. The start and end time have been recorded, so the duration to complete each survey can be estimated. During collecting the questionnaires, feedback about the problems of the questionnaires also collected from the respondents too. After that, the questionnaires that need to be given to the real respondents for this research have been adjusted and improved. Other than that, the pilot test also being used to measure the reliability and consistency of the questionnaire by seeing the Cronbach’s Alpha coefficient generated from SPSS software. The results of Cronbach’s Alpha coefficient value shown in Table 1 for all variables were more significant than 0.7, which means acceptable with the internal consistency and reliability. Hence, there DOI 10.18502/kss.v3i22.5047 Page 117 FGIC2019 was no item from each variable need to be deleted, and it can proceed to the primary respondents’ distribution for this research. Table 1: Pilot Test for Variables. Variables No. of Items Cronbach’s Alpha Colleague Relations 5 0.769 Organizational Commitment 5 0.778 Organizational Justice 5 0.726 Organizational Reputation 5 0.830 Communication 5 0.774 Organizational Politics 5 0.865 Employees’ Turnover Intention 5 0.781 3.7.3. Main survey The corrected and improved questionnaire would be distributed to the workers who were working in the Grade 7 construction companies in Klang, Selangor which had been registered for the CIDB. There are 160 sets of the questionnaire had been given to the construction companies through email and walk-in. Each Grade 7 construction companies would only be given one set of questionnaire which means each of the respondents will be represented by different construction companies. However, there was only received 73 respondents out of 160 populations. The response rate was 45.63%, which considered good and acceptable (Keller, 2014). The total respondents collected was 64.60% of the sample size (113 respondents) needed for this research. 4. Results 4.1. Respondent demographics Table 2 shows that most of the respondent’s gender that participated in this research are female, which has 75.3% from a total of 73 respondents. Besides that, there were only 18 males, which also represented 24.7% of the total respondents involved in this research. Most of the respondents’ age was in between 21 – 25 year’s old, which was 31.5% of the total respondents. There were only 2 respondents who were 2.7% of the total respondents were from the age of 20 years old and below. For education level, Bachelor’s Degree level has the most number of respondents, which is 39.7% while Doctorate level has the fewer respondents which only one person out of a total of 73 respondents in this research. Most of the respondent’s race was build up from DOI 10.18502/kss.v3i22.5047 Page 118 FGIC2019 Chinese, which has 56.2% of the respondents. However, Indian respondents had the lowest amount, which is 6.8% of the total population. Moreover, most of the respondents (47.9%) had 2 years and below working experience in their current working construction companies. There are only 8 respondents who have working experience of 11 years and above with a percentage of 11.0% of total respondents. Other than that, most of the respondents who participated in this research are holding admin position in their current working construction companies, which are 47.9% of the total research respondents. The others 17.8% of the respondents’ position in their current working place was such as senior executive, financial admin, electrical engineer, purchasing executive, senior contract executive, chemist, sales coordinator, receptionist, and quality control admin. 4.2. Descriptive analysis Based on Table 3, organizational politics had the highest mean of 3.1425, which showed that the employees’ turnover intention in respondents’ construction companies are mostly affected by organizational politics factor. Organizational reputation had the lowest mean of 3.0137, which implied that the employees’ turnover intention in construction companies were low influence by organizational reputation. Since all the mean value for each variable are more than 3 (Neutral), which means most of the respondents are neutral or agree toward those factors can affect the employees’ turnover intention in construction companies. 4.3. Normality test According to Curran et al. (1996), a samples are normally distributed if the absolute value of Skewness not more than 2 (within -2 to 2) and the absolute value of Kurtosis not greater than 7 (within -7 to 7). Based on the SPSS output in Table 4, it was shown that the highest absolute value of Skewness is 0.771 which less than 2 and the highest absolute value of Kurtosis is 1.068 which does not more than 7. Since all the statistic value of Skewness for each item are within -2 to 2 and all the statistic value of Kurtosis for each item are within -7 to 7, it can be confirmed that the samples are normally distributed. Besides that, the outlier has existed if there have any z-score value is less than -3.29 or more than 3.29 (Tabachnick & Fidell, 2007). After running the z-score in SPSS software, all the z-score values had been check, and all values are in the range of -3.29 to 3.29. Thus, there does not have any outlier exist. DOI 10.18502/kss.v3i22.5047 Page 119 FGIC2019 Table 2: Demographic Analysis. Demographic Components Gender Age Education Level Race Work Experience Position Number Respondents Percentage (%) Male 18 24.7 Female 55 75.3 20 years old and below 2 2.7 21-25 years old 23 31.5 26-30 years old 20 27.4 31-35 years old 17 23.3 36-40 years old 7 9.6 41-45 years old 4 5.5 46-50 years old - - 51 years old and above - - SPM 11 15.1 STPM 3 4.1 Diploma 20 27.4 Bachelor’s Degree 29 39.7 Master’s Degree 9 12.3 Doctorate 1 1.4 Others - - Malay 27 37.0 Chinese 41 56.2 Indian 5 6.8 Others - - 2 years and below 35 47.9 3-6 years 20 27.4 7-10 years 10 13.7 11 years and above 8 11.0 Accountant 7 9.6 Site Supervisor 4 5.5 Project Executive 6 8.2 Admin 35 47.9 Human Resource Executive 8 11.0 Others 13 17.8 4.4. Reliability analysis 4.4.1. Validity analysis SmartPLS software had been used to obtain the validity analysis. In SmartPLS, there is two validity that needs to be done, which are convergent validity and discriminant validity. To evaluate the convergent validity of the research data, the outer loadings DOI 10.18502/kss.v3i22.5047 Page 120 FGIC2019 Table 3: Descriptive Analysis. Variables N Mean Standard Deviation Variance Colleague Relations 73 3.0164 0.75554 0.571 Organizational Commitment 73 3.0521 0.78459 0.616 Organizational Justice 73 3.0521 0.67250 0.452 Organizational Reputation 73 3.0137 0.64406 0.415 Communication 73 3.0219 0.74428 0.554 Organizational Politics 73 3.1425 0.75754 0.574 Employees’ Turnover Intention 73 3.0466 0.82564 0.682 Table 4: Skewness and Kurtosis of Each Item. Item 1 2 3 4 5 CR OC OJ OR C OP ETI Skewness 0.066 0.771 0.196 -0.149 0.402 0.302 0.342 Kurtosis -0.949 0.492 -0.345 -0.938 -0.629 -0.304 -0.317 Skewness 0.320 0.452 0.247 -0.218 0.361 -0.056 0.056 Kurtosis -0.640 -0.039 -0.166 -0.955 -0.664 0.219 -0.637 Skewness 0.678 0.144 0.058 0.290 -0.220 -0.169 -0.383 Kurtosis 0.153 -0.360 -0.392 -0.157 -0.489 -0.262 -0.271 Skewness 0.233 -0.053 0.179 0.074 0.177 -0.400 0.004 Kurtosis -0.717 -0.857 -0.047 -0.772 0.442 -0.405 -0.895 Skewness -0.065 -0.145 0.345 0.235 0.108 0.084 -0.115 Kurtosis -0.689 -1.068 -0.658 -0.835 -0.901 -1.018 -0.209 Note: CR = Colleague Relations; OC = Organizational Commitment; OJ = Organizational Justice; OR = Organizational Reputation; C = Communication; OP = Organizational Politics; ETI = Employees’ Turnover Intention of each variables’ questions, and the Average Variance Extracted (AVE) need to be measured. The outer loadings should be 0.708 or higher, and the AVE value must be 0.50 or higher, so it reflects a good convergent validity of research data (Hair Jr et al., 2017). Those items that have outer loadings less than 0.708 will be deleted one by one until the variables do not contain any items that less than 0.708. Those items deleted (11 items) and remained (24 items) have shown in Table 6. Table 6 shows that all the outer loadings of each item remained are more than 0.708, which means that the latent variable should be able to explain at least 50% of each item’s variance. Not only that, AVE values for all variables are more significant than 0.5, which also can be said that all items for the variables have high levels of convergent validity. Since all-composite reliability for each variable are more than 0.7 after deleted items process, which means all the variables in this research are still acceptable with high internal consistency and reliability. Figure 2 shows that the PLS path model after item deleted process. DOI 10.18502/kss.v3i22.5047 Page 121 FGIC2019 Table 5: Convergent Validity. Item CR OC OJ OR C OP ETI 1 0.815 Deleted 0.889 Deleted 0.908 0.816 0.784 2 0.789 0.716 0.916 Deleted 0.910 0.779 0.888 3 0.874 0.851 Deleted 0.728 0.869 0.870 Deleted 4 0.780 0.916 Deleted 0.890 Deleted 0.800 0.856 5 Deleted Deleted Deleted 0.772 Deleted 0.883 AVE 0.665 0.668 0.815 0.700 0.803 0.667 0.745 0.919 Composite Reliability 0.888 0.889 0.898 0.874 0.924 0.889 0.921 Note: CR = Colleague Relations; OC = Organizational Commitment; OJ = Organizational Justice; OR = Organizational Reputation; C = Communication; OP = Organizational Politics; ETI = Employees’ Turnover Intention To assess the discriminant validity of the items, the latest main approach needs to be measured the Heterotrait-Monotrait ratio (HTMT). All the HTMT values should not more than the appropriate threshold level which is 0.85, and the HTMT confidence interval should not include the value 1 for all combinations of variables (Hair Jr et al., 2017). Table 6 shows that the highest HTMT value is 0.815 which means all the HTMT values are lower than the relevant threshold value of 0.85 and there are neither of the HTMT confidence intervals was including the value 1. Since the conservative HTMT threshold of 0.85 already supported the discriminant validity. The result of bootstrap confidence interval HTMT also clearly agreed toward the discriminant validity of the research variables. Seem all the model evaluation criteria have been met, it providing support for the measures’ validity. Table 6: Discriminant Validity. CR OC OJ OR C OP ETI Colleague Relations Communication 0.588 Employees’ Turnover Intention 0.418 0.758 Organizational Commitment 0.610 0.613 0.667 Organizational Justice 0.546 0.538 0.615 0.786 Organizational Politics 0.203 0.510 0.798 0.493 0.414 Organizational Reputation 0.622 0.815 0.708 0.755 0.613 0.573 HTMT confidence interval does not include 1 Yes Yes Yes Yes Yes Yes Yes Note: CR = Colleague Relations; OC = Organizational Commitment; OJ = Organizational Justice; OR = Organizational Reputation; C = Communication; OP = Organizational Politics; ETI = Employees’ Turnover Intention DOI 10.18502/kss.v3i22.5047 Page 122 FGIC2019 Figure 2: PLS Path Model After Items Deleted (Source: Authors’ own work). 4.4.2. Pearson correlation analysis If the significance value is less than 0.05, that means the relationship between independents variable and dependent variable are significant. Table 7 shows that all the independent variables had a significance value less than 0.05, which prove that the relations between the independent variables and employees’ turnover intention are significant. DOI 10.18502/kss.v3i22.5047 Page 123 FGIC2019 Therefore, all the independent variables (colleague relations, organizational commitment, organizational justice, organizational reputation, communication, and organizational politics) are considered as the factors that affect employees’ turnover intention. The Pearson Correlation value in Table 7 also explained that all the independent variables are negatively significant relationship toward employees’ turnover intention. Since all the questions in the questionnaire were changed to the negative statement in the early stage of the data analysis, thus, the positive value of Pearson Correlation represent a negative relationship between the independent variable and dependent variable. Table 7: Pearson Correlation Analysis. Employees’ Turnover Intention CR OC OJ OR C OP Pearson Correlation 0.354 0.576 0.511 0.582 0.678 0.691 Sig. (1-tailed) 0.001 0.000 0.000 0.000 0.000 0.000 N 73 73 73 73 73 73 Note: CR = Colleague Relations; OC = Organizational Commitment; OJ = Organizational Justice; OR = Organizational Reputation; C = Communication; OP = Organizational Politics 4.4.3. Multiple regression analysis Standardized beta coefficient (β) is very similar to the correlation coefficient, which ranges is between 0 to 1 or 0 to -1 (Morgan et al., 2013). The closer the value toward 1 or -1, it reflects the stronger the relationship between the independent variable and dependent variable. In order for the independent variables to be significant, the p-value must be less than 0.05 (Frost, 2013). Table 8 shown only two hypotheses (H5 and H6) are supported due to the variables’ beta coefficient are positive and have a significant level of p-values less than 0.05. While, there are four hypotheses (H1, H2, H3, and H4) not being supported because of the significant level of p-values are greater than 0.05. 5. Discussion The purpose of the finding is to identify the factors affecting employees’ turnover intention in construction companies in Klang, Selangor and to determine the relationship between the factors and employees’ turnover intention in construction companies in Klang, Selangor. There are six factors affecting employees’ turnover intention being DOI 10.18502/kss.v3i22.5047 Page 124 FGIC2019 Table 8: Summary of Hypotheses Testing. Hypotheses Relationship CR → ETI H1 H2 H3 H4 H5 H6 Note: *p < 0.05 OC → ETI OJ → ETI OR → ETI C → ETI OP → ETI Unstandardized Coefficient Standardized Coefficient t Sig. Decision B Std. error Beta -0.017 0.093 -0.016 -0.178 0.859 Not supported 0.135 0.103 0.135 1.306 0.196 Not supported 0.126 0.091 0.126 1.381 0.172 Not supported -0.001 0.112 -0.001 -0.005 0.996 Not supported 0.375 0.101 0.371 3.716* 0.000 Supported 0.462 0.088 0.433 5.261* 0.000 Supported Source: Authors’ own work selected from previous research, which are colleague relations, organizational commitment, organizational justice, organizational reputation, communication, and organizational politics. Based on the results from reliability analysis, it shows that all the independent and dependent variables in this research are acceptable with high internal consistency and reliability. Validity analysis results also proved the accuracy of an assessment of the research and providing support for the measures’ validity. This is because all the model evaluation criteria for the convergent validity and discriminant validity have been met. Thus, the research data can be said reliable and valid. Before proceeding to the Pearson Correlation analysis and Multiple Regression analysis, the normality test had been conducted to make sure the sample size distribution was normally distributed. It is because normal distribution is an underlying assumption for most of the statistical analysis (Hair Jr et al., 2010). From the normality test’s result, it was proven that the samples of this research are normally distributed. At the same time, the Pearson Correlation analysis and Multiple Regression analysis can proceed. Pearson Correlation Analysis was being applied to examine the correlation between the independent and dependent variables. According to the results of the Pearson Correlation analysis, all the independent variables are significant and strong negatively correlated to the employees’ turnover intention. Hence, the results proved that colleague relations, organizational commitment, organizational justice, organizational reputation, communication, and organizational politics are the factors which will affect employees’ turnover intention. At the same time, Research Objective 1 had been met. DOI 10.18502/kss.v3i22.5047 Page 125 FGIC2019 There are six hypotheses that had been developed and tested through Multiple Regression analysis to determine the relationship between the factors and employees’ turnover intention. The results of Multiple Regression analysis shown that there are only two out of six hypotheses were supported. H5 and H6 have been supported because from the multiple regression analysis shown that the p-value for H5 and H6 (communication and organizational politics) are less than 0.05 which means communication and organizational politics are significant negative relationship toward employee’ turnover intention. This result was same as the finding from previous researches such as AlTokjais (2016), Olcer & Ozenir (2017), Latif & Saraih (2016), and Abubakar & Abdullahi (2017) which proved that communication and organizational politics was negatively related to employees’ turnover intention. While, H1, H2, H3, and H4 have not been supported because their p-value from the multiple regression analysis was higher than 0.05. The reason causes the finding of this research turn out differently from the expected is maybe due to the different respondents’ background and characteristics of the population in this research. Different respondents’ background may have different expectation toward the factors employees’ turnover intention, which will influence the finding of the research. Thus, the results from this research are reasonable, and Research Objective 2 also had been achieved. 6. Conclusion and Implications This research is about the factors affecting employees’ turnover intention in construction companies in Klang, Selangor. The purpose of this study is to find out the factors that may influence the employees’ turnover intention in construction companies through identify the factors and determine the relationship between the factors with the employees’ turnover intention. The results from this research help the human resource department in construction companies to focus on those relevant and useful factors to reduce the employee’s turnover rate in the construction companies. The descriptive analysis results from this study showed that organizational politics had the highest mean compared to other factors from this research which also shown that employees’ turnover intention in construction companies are mostly affected by organizational politics factor. Besides that, the Pearson Correlation analysis results show that all six factors (colleague relations, organizational commitment, organizational justice, organizational reputation, communication, and organizational politics) were highly correlated to the employees’ turnover intention. DOI 10.18502/kss.v3i22.5047 Page 126 FGIC2019 Therefore, those six factors were considered as the factors that will affect employees’ turnover intention. Throughout multiple regression analysis results, only two hypotheses (H5 and H6) had been supported which are the communication factor, and organizational politics factor have a significant negative relationship with employees’ turnover intention in construction companies in Klang, Selangor. Hence, the construction companies in Malaysia should take more attention toward the communication and organizational politics factors to control or reduce the employees’ turnover rate in the construction companies. Results from this research provided the evidence and brought convince for the construction companies in Malaysia to reduce the employees turnover rate. This research finding had filled up the previous research gap which lack of investigation in construction companies in Malaysia. Since the scope of the study is focused on workers that work in the Grade 7 construction companies in Klang, Selangor, the findings of this research can only indicate the perspectives of construction companies in Klang, Selangor. The results from this study might not be so useful or applicable to other districts or states of Malaysia. The employees’ turnover intention is only measured in six aspects, which are colleague relations, organizational commitment, organizational justice, organizational reputation, communication, and organizational politics. The employees’ turnover intention may include many other aspects and factors as well, which do not being measured in this research. The main limitation of this study is the actual population size of Grade 7 construction companies in Klang, Selangor. CIMS does not update the current status of each construction companies, and those companies that already close down or move to other places does not being removed from the CIMS list. This might be the difficulty in getting the accurate total number of the population size for the research and cause the inaccuracy of the population size for the Grade 7 construction companies in Klang, Selangor. At the same time, it may cause the number of respondents who do not achieve the sample size of this research. Another limitation of this study is the rejection of filling the questionnaire from the respondents. There are some respondents do not respond to the questionnaire form that had been sent to them through email, and some of the respondents give an excuse to refuse to fill up the questionnaire form being given to them. In order to improve the reliability of this study, the scope of the study can be expanded to other states of Malaysia. In this study, the employees’ turnover intention is only measured by six factors. If there are more factors being measured toward the employees’ turnover intention, it may make the result of the study more accurate. In-depth DOI 10.18502/kss.v3i22.5047 Page 127 FGIC2019 research can also further investigation about how to solve the employees’ turnover intention problem in construction companies based on the factors. This investigation can make the study more precise and informative. Other than that, the data analysis can be described more depth by classified two different groups toward the analysis such as investigate the view from different gender (male and female) toward the factors affecting employees’ turnover intention by using one-sample mean T-test. Furthermore, the target respondents in the future study can be narrow down to specific level such as management level because those measurement variables may only influence toward employees who work in the management level or top management level. Moreover, qualitative research design can be added into the future study. When both the qualitative and quantitative research design involved in the study, it can improve and enhance the results from the finding. Acknowledgement We would like to thank Faculty of Industrial Management and FIM’s Governance and Integrity Centre, Universiti Malaysia Pahang for the financial support by sponsoring this paper to be presented in the FGIC 2nd Conference on Governance and Integrity 2019. References [1] Abubakar, R. A., & Abdullahi, A. I. (2017). Perceived Organizational Politics and Employee Turnover Intention: A Longitudinal Study in Nigeria. Journal on Innovation and Sustainability, 8(1), 12-17. [2] Ahmed, M., & Nawaz, N. (2015). Impact of Organizational Commitment on Employee Turnover: A Case Study of Pakistan International Airlines (PIA). Industrial Engineering Letters, 5(8), 57-69. [3] Ali, N., & Jan, S. (2012). Relationship between Organizational Justice and Organizational Commitment and Turnover Intentions amongst Medical Representatives of Pharmaceuticals Companies of Pakistan. Journal of Managerial Sciences, VI(2), 202-212. [4] Allahyary, M. H., & Beheshtifar, M. (2013). Study the Relationship among Organizational Reputation with Organizational commitment and Employees’ Turnover Intention. International Research Journal of Applied and Basic Sciences, 6(10), 14671478. DOI 10.18502/kss.v3i22.5047 Page 128 FGIC2019 [5] Alniacik, U., Cigerim, E., Akcin, K., & Bayram, O. (2011). Independent and joint effects of perceived corporate reputation, affective commitment and job satisfaction on turnover intentions. Procedia Social and Behavioral Sciences, 24, 1177-1189. [6] Al-Tokhais, A. A. (2016). The Relationship between Communication Effectiveness and Multicultural Employees’ Job Outcomes. Master Thesis, 119-120. [7] Bateman, G. (2009). Employee Perceptions of Co-worker Support and Its Effect on Job Satisfaction, Work Stress and Intention to Quit. Masters of Science in Applied Psychology, 1-52. [8] Buljubasic, E. (2008, June 20). Relationship with co-workers and job satisfaction. [9] Connelly, L. M. (2008). Pilot studies. Medsurg Nursing, 17(6), 411-412 [10] Cronbach’s Alpha: Simple Definition, Use and Interpretation. (2018). Retrieved from Statistics How To: http://www.statisticshowto.com/cronbachs-alpha-spss/ [11] Curran, P. J., West, S. G., & Finch, J. F. (1996). The robustness of test statistics to nonnormality and specification error in confirmatory factor analysis. Psychological Methods, 1, 16–29. [12] Eisenberg, E. M., & Witten, M. G. (1987). Reconsidering openness in organizational communication. Academy of Management Review, Vol. 12, No. 3, 418-426. [13] Fee Yean, T., & Yusof, A. A. (2016). Organizational Justice: A Conceptual Discussion. Procedia - Social and Behavioral Sciences, 219, 798-803. [14] Ferris, G.R. et al. (1996), “Perceptions of organizational politics: prediction, stressrelated implications, and outcomes”, Human Relations, Vol. 49 No. 2, pp. 233-66. [15] Frost, J. (2013, July 5). How to Interpret Regression Analysis Results: P-values and Coefficients. Retrieved from Quality Digest: https://www.qualitydigest.com/ inside/quality-insider-column/interpret-regression-analysis-results-p-values-andcoefficients.html# [16] Greenberg, J. (1987). A taxonomy of organizational justice theories. Academy of Management Review, 12, 9-22. [17] Hair Jr, J. F., Black, J. W., Babin, B. J., & Anderson, E. R. (2010). Multivariate Data Analysis (Seventh Ed., pp.1–758). Edinburgh: Pearson Education Limited. [18] Hair Jr., J., M. Hult, G., Ringle, C., & Sarstedt, M. (2017). A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM) (Second Edition ed.). United States of America: SAGE Publications. [19] Hendriks, M. (2016). Organizational Reputation, Organizational Attractiveness And Employer Branding: Clarifying the Concept. Master Thesis, 23-25. DOI 10.18502/kss.v3i22.5047 Page 129 FGIC2019 [20] Hertzog, M. A. (2008). Considerations in determining sample size for pilot studies. Research in Nursing and Health, 31(2), 180-191. [21] Jayaram, S. V. (2015, December 3). Key Insights to Attract, Reward and Retain Talent in Malaysia: Aon Hewitt’s Views. Retrieved from HR in Asia: http://www.hrinasia.com/recruitment/key-insights-to-attract-reward-andretain-talent-in-malaysia-aon-hewitts-views/ [22] Jehanzeb, K., Rasheed, A., & Rasheed, M. F. (2013). Organizational Commitment and Turnover Intentions: Impact of Employee’s Training in Private Sector of Saudi Arabia. International Journal of Business and Management, 8(8), 79-90. [23] Keller, A. (12 November, 2014). What is an acceptable survey response rate. Retrieved from National Social Norms Center: http://socialnorms.org/what-is-anacceptable-survey-response-rate/ [24] Khan, R. A., Liew, M., & Ghazali, Z. B. (2014). Malaysian Construction Sector and Malaysia Vision 2020:Developed Nation Status. Procedia - Social and Behavioral Sciences 109, 507 – 513. [25] Krejcie, R. V., & Morgan, D. W. (1970). Determing Sample Size for Research Activities. Educational and Psychological Measurement, 607-610. [26] Kwon, K., & Rupp, D. E. (2013). High-performer turnover and firm performance: The moderating role of human capital investment and firm reputation. Journal of Organizational Behavior, 34(1), 129-150. [27] Latif, F. D., & Saraih, U. N. (2016). Factors Influencing Employee Turnover in Private Sector in Malaysia: A Concept Paper. Australian Journal of Basic and Applied Sciences, 10(7), 51-55. [28] Letchumanan, T., Apadore, K., & Ramasamy, M. (2017). Factors Influence Turnover Intention in Commercial Banks Malaysia: A Theoretical Model. Innovative Journal of Business and Management, 13-21. [29] M.N.A, A., Majid, T., Ahamad, M., & Hanafi, M. (2012, October). A Qualitative Study of Precast Plants in Malaysia. Indian Concrete Journal, 86(10), 47 - 58. [30] Morgan, G. A., Leech, N., Gloeckner, G., & Barrett, K. (2013). IBM SPSS for Introductory Statistics (Fifth Edition ed.). New York: Ryouledge Taylor & Francis Group. [31] Nkomo, W., & Thwala, WD. (2009). Problems Facing Construction Companies in Retention of Employees in South Africa: A Case Study of Gauteng Province. CIDB Paper 22, 116-127. [32] Olcer, F., & Ozenir, İ. (2017). The Relationship between Organizational Communication, Organizational Commitment and Intention to leave. European Journal of Business and Social Sciences, 6(5), 30-53. DOI 10.18502/kss.v3i22.5047 Page 130 FGIC2019 [33] Palaganas, M. A. (2018, February 27). High Employee Turnover as a Silent Construction Manpower Crisis . Retrieved from Rensol Recruitment: https://rensol. com/employee-turnover-construction-industry/ [34] Rahman, I. A., Memon, A. H., & Abd. Karim, A. (2013). Significant Factors Causing Cost Overruns in Large Construction Projects in Malaysia. Journal of Applied Sciences, 13(2), 286-293. [35] Raisiene, A. G., & Vilke, R. (2014). Employee Organizational Commitment Development at Voluntary Organizations in Lithuania. Human Resources Management & Ergonomics, VIII, 88-101. [36] Seong, L. C. (2015). The Study on Factors Influencing Employee Turnover in E&E Manufacturing Industry in Northern Malaysia. Pulau Penang: Universiti Sains Malaysia. [37] Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). Boston, MA: Allyn & Bacon. [38] Thomas, J. (2015). Study on Causes and Effects of Employee Turnover in Construction Industry. International Journal of Science and Research, Volume 4(Issue 5), Page 3041-3044. [39] Yang, J., & Wittenberg, P. (2016, June 13). Perceived Work-related Factors and Turnover Intention. Journal of Business Administration, 1-69. [40] Yang, T., Shen, Y.-M., Zhu, M., Liu, Y., Deng, J., Chen, Q., & See, L.-C. (2015). Effects of Co-Worker and Supervisor Support on Job Stress and Presenteeism in an Aging Workforce: A Structural Equation Modelling Approach. International Journal of Environment Research and Public Health, 13(72), 1-15. DOI 10.18502/kss.v3i22.5047 Page 131