The Job Descriptive Index (JDI), which has been adjusted for the case of Vietnam, will be employed to measure job satisfaction of production workers at Tân Hiệp Phát Co., Ltd. This adjusted scale includes five facets, viz.
Development of human resource A recent research by Chou (2007) has shown that a variety of discontent at work is the primary excuse for employees’ resignation Thus, it raises a need for enterprises to measure job satisfaction of their staff with a view to avoiding the happening of further unexpected resignations and establishing a better policy on human resource management The Job Descriptive Index (JDI), which has been adjusted for the case of Vietnam, will be employed to measure job satisfaction of production workers at Tân Hiệp Phát Co., Ltd This adjusted scale includes five facets, viz salary, relationship with boss, job characteristics, working conditions, and fringe benefits, which will be evaluated by means of the Cronbach’s Alpha, EFA, and regression analyses Then, some solutions are to be recommended Keywords: job satisfaction, salary, relationship with boss, job characteristics, working conditions, fringe benefits Introduction together with the development of vietnam’s economy, the need for high qualified human resource is very pressing not only enterprises recruit exact individuals for appropriate posts, but they also have to attempt to retain staff’s loyalty which is affected by salary, work conditions, and fringe benefits By helping employees gain their job satisfaction, enterprises can assure a stable workforce for running business tân Hiệp phát trading and service co., ltd (hereunder referred to as “the company” for short) primarily produces and trade in beer, liquor and beverage exit interview questionnaires show that the company’s employees would like a higher salary, mid-shift breaks, a suitable working schedule, safe working environment, and more fringe benefits some extant researches have pointed out the satisfaction with tasks assigned as the most important factor affecting staff’s loyalty (luddy, Economic Development Review - June 2011 2005 & chou, 2007) thus, it is necessary to study job satisfaction in a hope of adjusting policies on human resources management Theoretical fundamentals a Job satisfaction: there have been many concepts of job satisfaction thus far ellickson and logsdon (2002) suggest that job satisfaction is the extent to which employees are fond of their job, and the attitude, either positive or negative, towards the job as per Quinn and staines (1979), job satisfaction is expressed via positive responses to job luddy (2005) places a stress on factors affecting job satisfaction, viz status, supervision of boss, relationship with colleagues, job characteristics, fringe benefits and other rewards (i.e promotion opportunity, better working facilities, organizational structure, etc.) as Kreitner and Kinicki (2007) put it, job satisfaction is reflected via the extent to which an employee like their assigned task; or in other words, *: University of Finance - Marketing Development of it is the affection of an employee towards the job Herzberg et al (1959) puts forth a two-factor theory (i.e motivation-hygiene theory) according to that, motivational factors include achievement, recognition, job characteristics, responsibility, promotion, and growth; hygiene or maintenance factors comprise company policy and administration, supervision, relationship with co-workers, working conditions, pay and benefits, relationship with co-workers, personal life, status and job security essentially, hygiene factors are needed to ensure an employee is not dissatisfied motivation factors are needed to motivate an employee to higher performance according to the equity theory, which was first developed by John stacey adams in 1963, the structure of equity in the workplace is evaluated on the basis of the ratio of inputs (i.e what employees bring to a job) to outcomes (i.e what employees receive from inputs against the perceived inputs and outcomes of others) Inputs consist of experience, effort, and personal competence outcomes include salary, recognition, reputation, sense of achievement, praise, fringe benefits, etc If an employee notices that another person is getting more recognition and rewards for their contributions, even when both have done the same amount and quality of work, it would persuade the employee to be dissatisfied Hackman and oldham (1974) developed the job characteristics model which states that there are five core job characteristics, namely skill variety, task identity, task significance, autonomy, and feedback they impinge on three critical psychological states, viz experienced meaningfulness, experienced responsibility for outcomes, and knowledge of actual results, which in turn influence work outcomes b Measuring job satisfaction: Job Descriptive Index (JDI): minnesota satisfaction Questionnaire by Weiss et al (1967), JDI by smith et al (1969), and job satisfaction survey by spector (1997) are three typical apparatuses which, (especially the two formers), are widely employed to measure job satisfaction (Green, 2000) as smith et al (1969) put it, job satisfaction is measured on facets namely salary, promotion and training opportunity, relationship with boss and peers, and job characteristics price (1997) sup- human resource poses that JDI is a useful tool for measuring job satisfaction because of the fact that JDI possesses a good content, and well-established and reliable concepts (Kerr, 1995) It is just within the last two decades of the 20th century that JDI has been employed in some 600 published researches (ajmi, 2001) luddy (2005) utilized JDI to investigate job satisfaction amongst 203 employees at a public health institution in Western cape, africa; and found out that occupation, ethnic group, gender, educational level, seniority, age and status have profound impacts on job satisfaction In vietnam, traàn Kim Dung (2005) employed JDI and maslow’s hierarchy of needs to measure job satisfaction with a result that there emerges two more factors namely fringe benefits and working conditions c Research model: almost researches using JDI have reflected job satisfaction on grounds of salary, promotion and training opportunity, relationship with boss and co-workers, and working conditions Besides, fringe benefits and work conditions which have been identified in the research by traàn (2005) will also be included in the model accordingly, the research model can be summarized as in figure Figure 1: Recommended research model Research results a Methodology: a qualitative research will be conducted so as to explore, adjust and modify observed variables Economic Development Review - June 2011 Development of which are to be used to test concepts accordingly, some groups of eight (four males and four females for each) will be established according to nonprobability sampling for group discussion research participants are those who have been working for the company for more than one year, ages from 20 to 35, and have the monthly salary of over vnD2 million there are four workers, two operating ones, and two maintaining ones for each group accordingly, a preliminary questionnaire has been developed In order to assure the questionnaire is easily comprehensible and applicable, a pilot research must be conducted (saunders et al., 2007) twenty questionnaires have been sent to twenty employees who meet the above-mentioned requirements; and face-to-face interviews have been done to make sure that they can easily comprehend all the words of the questionnaire then, the official questionnaire will be worked out the full-scale study is conducted quantitatively face-to-face interviews with 300 employees of the company in combination with the official questionnaire will help collate data b Sample features: of 300 questionnaires collected, there are 13 unusable ones due to lack of responses to some questions among the remainder, 69.7% are males and 30.3% females those aging from 20 to 25 occupy 54.4%, from 25 to 30 representing 42.9%, and the remainder for those over 30 years old In terms of salary, some 51.2% of the population gain from vnD2 million to vnD3 million per month counting to october 2010, those who have been working for the company from one to three years represent 48.8%, from six months to one year occupying 42.5%, from three to five years accounting for 7%, and over years making up 1.7% In terms of educational level, those with the post-secondary education account for 48.1%, with the junior high school certificate making up 20.6%, with the senior high school occupying 30%, and with higher education just representing below 1% Besides, in terms of status, around 44.6% are workers, 38% are operating staff, 15% are maintaining ones, and 2.4% are others c Testing the scale and research model: the cronbach’s alpha and efa are employed to test the scale the former will be utilized first 10 Economic Development Review - June 2011 human resource with a view to leaving out unfit variables; and efa is used for testing the remainder many researchers agree that when cronbach’s alpha range is from 0.6 onward, the scale is usable; and variables with the correlation coefficient being smaller than 0.3 will be rejected (nunnally and Bernstein, 1994) Table 1: Cronbach’s alpha coefficients Satisfactions Cronbach's alpha Salary (SAL) 747 Promotion and training opportunities (TRA) 833 Relationship with boss (SUP) 799 Relationship with colleagues (COL) Job characteristics (CHA) Working conditions (CON) 876 797 820 table shows that most variables, save sup_1 and cHa_1, retain a correlation coefficient larger than 0.3 thus, sup_1 and cHa_1 will be left out the cronbach’s alpha of the scale is larger than 0.7; to put it another word, the scale is fit for the research In conducting the efa, the principal component analysis with the support of varimax rotation is used and factor is only extracted at points where eigenvalue is equal to or greater than the scale is accepted when the total variance extracted is equal to or larger than 50% efa results in table shows that 31 variables have been extracted into seven factors namely salary, promotion and training opportunities, relationship with boss, relationship with colleagues, job characteristics, working conditions, and fringe benefits the factor loadings are all larger than 0.5, and thereby having statistical significance for the Kmo coefficient equaling 0.825 (i.e larger than 0.5), efa coincides with data the chi-square of Bartlett test is 4791.971 with the significance at 0.000; and thus, observed variables are correlational the variance extracted reaching 65.03% (greater than 50%) shows that seven extracted factors explain 65.03% variability of data the eigenvalue is equal to 1.217 (greater than 1), therefore the scale is accepted (Gerbing & anderson, 1988) d Regression analysis: Development of human resource Bảng Kết EFA cho thang đo nhân tố thành phần thỏa mãn công vieäc Factor loadings Observed variables SUP_7 798 SUP_5 747 SUP_4 713 SUP_2 668 SUP_6 659 SUP_3 654 CON_2 843 BEN_4 766 CON_5 625 CON_3 607 CON_4 598 CON_1 591 CHA_4 764 CHA_6 757 CHA_5 725 CHA_3 710 CHA_2 528 TRA_4 764 TRA_2 754 TRA_3 705 TRA_1 593 BEN_2 734 BEN_5 672 BEN_1 666 BEN_3 658 COL_1 884 COL_2 872 COL_3 862 SAL_1 833 SAL_2 793 SAL_3 571 Eigenvalue Variance extracted 8.975 2.655 2.199 1.893 1.687 1.533 1.217 28.951 8.566 7.095 6.106 5.443 4.945 3.927 ordinary least square will be employed in the regression analysis along with job satisfaction as the dependent variable and seven extracted factors as independent ones firstly, it is necessary to ponder the linear correlations among variables the correlation coefficients show that the dependent variable retains linear relationships with seven independent ones, where the correlation between job satisfaction and fringe benefits produces the highest coefficient (0.729) while the lowest one (0.270) exists in the correlation between job satisfaction and relationship with colleagues there is also correlations amongst independent variables, even if the coefficient is not high, ranging Economic Development Review - June 2011 11 Development of human resource between 0.187 and 0.546 table indicates that there are five factors that have a linear relationship with job satisfaction (sig < 0.05) Yet, for tran and col having sig.>0.05, they not have a linear relationship with job satisfaction the minute sig value in the f-test (.000) (table 5) shows that variables of the model can explain variability of the dependent variable, fits the data, and is usable as table figures out, due to a small tolerance and vIf not exceeding 10, there is neither multicollinearity nor close corre- Table 3: Regression analysis results (using the Enter method) Unstandardized coefficients Model Variables Standardized coefficients Standard deviation b -.547 585 SAL 200 030 242 6.629 000 761 1.314 TRA 051 037 059 1.387 167 560 1.787 SUP 120 037 117 3.267 001 798 1.253 COL 001 032 001 032 974 890 1.124 CHA 161 041 161 3.973 000 619 1.616 CON 122 038 136 3.255 001 583 1.715 BEN 363 034 443 10.662 000 590 1.694 R2 R2 adjusted Standard Durbin- deviation Watson of estimate coefficient 709 404 2.052 Table 5: Variance analysis results Total sum of Root df mean squares Regression Residual Total 12 VIF 159 845(a) 714 Model Tolerance -.087 Table 4: Regression analysis results R Sig Constant the regression model then includes only five variables, viz sal, sup, cHa, con, and Ben; and can be written as follows: Y = -0.087 + 0.200* X1 + 0.120* X2 + 0.161* X3 + 0.122* X4 + 0.363* X5 the Beta value shows that Ben, sal, cHa, con, and sup sharply affects job satisfaction with the Beta value respectively equaling 0.443; 0.242; 0.161; 0.136 and 0.117 the adjusted r2 of the model equaling 0.709 (table 4) indicates that the independent variables of the regression equation can explain 70.9% variability of the dependent variable Model t b Multicollinearity statistics F Sig squares 114.795 45.937 281 160.731 286 22.959 140.442 000 0.163 Economic Development Review - June 2011 lation among independent variables In addition, a scatter plot of residual and expected values is also worked out with a view to testing the assumption of linear relationship the results show that residues are scattered randomly around and along a diagonal line going through the origins of ordinate axis, and thus the assumption of linear relationship is satisfied the spearman’s rank correlation test for the absolute value of residues and independent variables shows that all sig values are greater than 005; and thus the error variance is kept unchanged examining the normal distribution of residuals through the histogram shows that the distribution of residuals is nearly normal (mean = -.23*10-15 ~ 0; and the standard deviation equaling 0.991) the Q-Q plot indicates that dots are scattered closely to the diagonal line, and the distribution of residuals can be considered as normal; and the assumption of normal distribution of residuals is not violated the d-value checked in the Durbin-Watson table retains the significance at 5%; dl = 1.707 and du = 1.831 the d-value equaling 2.052 (table 4) falls within the domain 2