Chapter 4 DATA ANALYSIS AND RESULTS
4.2 Scale Verification
4.2.3 Exploring factors analysis
In order to analyze the interdependencies among observed variables, reduce data to a smaller set of underlying summary variables or discover the underlying structure of observed variable with the theoretical constructs, exploratory factor analysis (EFA) for the variable is assessed in this section.
In study 1- Global perceived organizational support of Eisenberger et al. (1986) after running factor analysis apply for 36 statements in the full version scale they figure out the
presumptive perceived support factor accounted for 93.9% of the common variance and it reveals the minimal evidence for existence of another factor with the possible second factor for 6.1 %. From this light, to discover whether it has any underlying structure in the short version with 17 items of POS scale use in this research, the exploratory factor analysis is using.
According to Hair et al (1998), Factor loading - transmission coefficient was a norm to ensure the practical significant and with the sample size approximately from 100 to 350, factor loading >= .55 was considered having practical significance. To test whether running EFA for variable suitable or not, KMO coefficient, Bartlett’s test Total, Variance Explained, Eigenvalue Point checking would be made. The coefficient of KMO (Kaiser Meyer Olkin) - measure of how suited data for EFA should be great than .5 (.5=< KMO <=1), Bartlett’s test demonstrate the observed variables are correlated with each other in overall when sig < .05, Total Variance Explained much achieve value from 50% and Eigenvalue Point should be higher than 1 then the factor will have statistical significance.
The result after using Principal Component Analysis method available of SPSS for POS exploring factor analysis in these tables below:
Table 4.6: KMO and Bartlett's Test of POS Kaiser-Meyer-Olkin Measure of Sampling highly significant Bartlett's Test of Sphericity with sig = .000 < 0.05.
Additionally, in the “Total Variance Explained” table, a ratio of total variance extracted is
data variability.
This table below shows 16 items of POS scale have practical significance with all factor loading are >= 5.9. The special thing can be seen from the results that are the appearance of two elements analytical variable from a consolidated scale of POS. This is a clear indication that the POS independent variable in this study has 2-dimesion structure. The first factor includes 11 items which are POS8, POS17, POS15, POS9, POS7, POS16, POS4, POS11, POS12, POS13, POS1 which are mostly referring to evaluative perceived organizational reward and job condition exist in the working environment. Besides it, the second factor contains 5 items include POS3, POS5, POS2, POS6 AND POS10 that items tending to indicate perceived fairness treatment support from an organization which experienced by the employees. In Eisenberger and his coworker research (2002), they have proven that tangible resources such as fairness treatment, organizational rewards, training, promotions, or favorable job conditions that are antecedents of Perceived organizational support hence these findings above have consistent connection with literature and 2-exploring factors of POS will be use to analyze the relationship with other variables in the next part.
Table 4.7 POS’ Factor Analysis Result Factor
1 2
POS8 - Formosa really cares about my well-being.
.855 .305
POS17 - Formosa tries to make my job as interesting as possible
.838
POS15 - Formosa cares about my opinions .836 .332
POS9 - Formosa is willing to extend itself in order to help me perform my job to the best of my ability
.831
POS7 - Help is available from Formosa when I have a problem
.828
POS16 - Formosa takes pride in my .827
accomplishments at work
POS4 - Formosa strongly considers my goals and values.
.820 POS11 - Formosa is willing to help me when I
need a special favor
.771 .329
POS12 - Formosa cares about my general satisfaction at work.
.741 .358
POS13 - If given the opportunity, Formosa would take advantage of me
.699 POS1 - Formosa values my contribution to its
well-being
.669 .368
POS3 - Formosa fails to appreciate any extra effort from me
.809 POS5 - Formosa would ignore any complaint
from me
.751 POS2 - If Formosa could hire someone to
replace me at a lower salary it would do so.
.693 POS6 - Formosa disregards my best interests
when it makes decisions that affect me
.681 POS10 - Even if I did the best job possible,
Formosa would fail to notice.
.592
Hence, there are 2 hypotheses are adding and the research framework become:
H1a: Perceived fairness treatment support will positively affect job satisfaction.
H1b: Perceived organizational reward and job condition support will positively affect job satisfaction
4.2.4 Analysis of Pearson Correlation Coefficient
This part describes Pearson correlation coefficient testing to check the level of linear relationship between two quantitative variables. All the variables in this study will be included to running this analysis as control variables (gender, age, status, kids, working time) POS and two-factor perceived fairness treatment support, perceived organizational reward and job condition support which explored in factor analysis above, job satisfaction, organizational reputation and turnover intention.
Table 4.8 demonstrates the correlation relationship between all the variables in this research. The findings are gender positively correlated with POS ( = .129, p < .05), perceived reward and job condition ( = .134, p < .05), job satisfaction ( = .145, p < .05) and negatively correlated with working time ( = -.159, p < .05), organizational reputation ( = -.143, p < .05). Age correlate positively with status ( = .538, p < .01), working time ( = .254, p < .01) and job satisfaction ( = .201, p < .01). Similarly, marital status is positively
correlated with kids ( = .744, p < .01), working time ( = .162, p < .01), POS ( = .239, p
< .01), perceived fairness treatment ( = .174, p < .01), perceived reward and job condition (
= .242, p < .01) and job satisfaction ( = .229, p < .01). Number of kid correlate positively with working time ( = .178, p < .01), POS ( = .262 p < .01), perceived fairness ( = .178, p
< .01), reward and job condition ( = .254, p < .01), job satisfaction ( = .286, p < .01), while negatively correlative with organizational reputation ( = -.150, p < .05), turnover intention (
= - .192, p < .01). Working time have highly negatively correlate with POS ( = -.202, p
< .01), perceived fairness ( = -.157, p < .01), perceived reward and job condition ( = -.219, p < .01), job satisfaction ( = .174, p < .01) although have positively correlate with
organizational reputation ( = .177, p < .01)
Additionally the relationship between these independent and dependent variables of this research also are revealed. POS, perceived fairness treatment, perceived organizational reward
and job condition all have positive relative with job satisfaction ( = .767, = .532, = 0.773, p < .01) whereas negatively correlate with organizational reputation ( = -.323, = -.351, =
-.277, p < .01) and turnover intention ( = -.453, = -.494, = -.389, p < 0.01). Job satisfaction has strongly negatively correlation with organizational reputation ( = -.234, p
< .01) and turnover intention ( = -.394, p < .01). And lastly, organizational reputation have positively relative with turnover intention ( = .197, p < .01),
Table 4.8 Mean, Standard Deviation and Correlations
Mean SD 1 2 3 4 5 6 7 8 9 10 11 12
1.Gender 1.39 .49
2.Age 30.64 5.58 -.048
3.Status 1.49 .50 .058 .538**
4.Kids .70 .93 .085 .686** .744**
5.Education 1.86 .67 -.019 .088 .022 -.083
6.Working time 2.38 .84 -.159** .254** .162** .178** .019
7.POS 3.95 .61 .129* .105 .239** .262** -.050 -.202**
8.Fairness 3.91 .62 .051 .100 .174** .178** .038 -.157** .792**
9.Reward and Job condition 3.99 .72 .134* .100 .242** .254** -.064 -.219** .965** .623**
10. SAT 3.98 .64 145* .201** .229** .286** -.032 -.174** .767** .532** .773**
11. ORP 2.34 .73 -.143* -.009 -.056 -.150* .000 .177** -.323** -.351** -.277** -.234**
12. Turnover Intention 2.38 .89 -.054 -.101 -.160** -.192** .087 .087 -.453** -.494** -.389** -.394** .197**
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
4.3 Result of Hypotheses Testing
4.3.1 Results on direct antecedent of job satisfaction, turnover intention
Regression analysis is also conducted to test the hypothesis and to define the direction of relations. In this section, regression analysis was used to test the effect of POS include fairness treatment perceived, reward and job condition perceived toward job satisfaction – hypotheses H1, H1a, H1b and job satisfaction toward turnover intention – hypothesis H2 that were proposed in the last parts.
The results obtained from the regression analysis are shown in tables below.
From Table 4.9, the first finding is an R2 value of the model 1 to test hypothesis H1 was calculated as .588. This value shows the perceived organizational support explained 58.8%
variability of job satisfaction with = .767 (p < .001). It can understand as research model show F value equal to 383.933 which is significant due to the value is less than .001. Hence hypothesis H1 “Perceived organizational support will positively affect job satisfaction” is supported.
Table 4.9 Regression analysis results of direct antecedent on Job Satisfaction, Turnover Intention
Testing Dependent variable Independent
variable Beta R-square F
H1 POS Job satisfaction .767*** .588 383.933***
H1a Perceived fairness Job satisfaction .081
.602 202.565***
H1b Perceived reward and job condition
Job satisfaction .723***
H2 Job satisfaction Turnover intention -.394*** .152 49.341***
*p<.05 **p<.01 ***p<.001
To know the effect of POS to job satisfaction is fully supported or only partial support the analysis relationship of 2- factors of POS which are perceived fairness treatment and perceived organizational reward and job condition were explored in the factor also using this phrase. The result was showed in Table 4.9 that perceived fairness support have no affecting job satisfaction ( = .81, p > .05) whereas, perceived organizational reward and job condition support show positive affect job satisfaction ( = .733, p < .001) and can explain 60.2% of the variance in job satisfaction (R2 = 0. 602, F= 202.565). Thus the supplementary hypothesis H1b is accepted and H1a isn’t accepted so POS is partial effect to job satisfaction. From that finding, the researchers point out company should pay more attention to an important role of organizational reward and job condition in order to enhance job satisfaction.
Job satisfaction shows the significant negative effect to turnover intention from the finding in the same table above. R2 value of the model 2 to test hypothesis H2 was calculated as .152 so job satisfaction may explain 15.2% variability of turnover intention with = .767, F= 49.341, p < .001, it mean the higher level job satisfaction employee have, the less likely they are to leave. In other words, “Job satisfaction will be negatively related to turnover intention” – the hypothesis H2 is accepted.
In conclusion, hypotheses H1, H1b, H2 are all accepted except H1a.
4.3.2 Mediating regression analysis of Job satisfaction
According to Baron & Kenny (1986), Judd & Kenny (1981) mediation is supported when four conditions are met: (1) The relationship between the independent variable and the dependent variable must be significant. (2) The relationship between the independent variable and the mediator/intervening variables must be significant. (3) The relationship between the mediator variable and the dependent variable must be significant. (4) The relationship between the independent variable and the dependent variable must be weaker (partial mediation) or insignificant (full mediation) in the presence of the mediating variable. In
another hand, many recently researchers and typically Shrout & Bolger (2002) point that step 1 is not necessarily required in process of mediator testing so this study uses the guidelines of Shrout & Bolger (2002) with 3 conditions to test a mediating role of job satisfaction.
Hypothesis H3 has proposed that job satisfaction mediates the relationship between POS and turnover intention. Hence, in this testing, perceived fairness support, and perceived reward, job condition support are independent variables, job satisfaction is mediating variable and turnover intention is a dependent variable. The results are demonstrated in Table 4.10.
Table 4.10 Regression analysis results of Job satisfaction mediating Job satisfaction Turnover Intention reward and job condition support show the significant relationship with the mediator factor (= .723, p < .001) and perceived fairness support do not show the direct effect to job satisfaction (= .081, p > .05). Secondly, a result of testing Step 2 pointed that the mediator factor job satisfaction has a significant relationship with dependent variable – turnover intention ( = -.394, p < .001). And finally, Step 3 in this table show a significant mediator of job satisfaction on POS and turnover intention relationship with = - .183, p < 0.05, F=32.628, p<0.001 and R2 = 0. 268.
In detail, perceived fairness treatment dimension show a significance in the direct path from POS to turnover intention (= - .412, p < .001) also significant but decreasing effect in the indirect path from it to turnover intention when job satisfaction was controlled ( = -.398, p
< .001) but it not adapt one condition of mediating testing which is a relationship with mediator factor so job satisfaction do not mediate the relationship between perceived fairness treatment and turnover intention.
Whereas perceived organizational reward and job condition has positive affecting to mediator factor and have a significant direct path to turnover intention ( = - .131, p < .05) but show a no significance in the indirect path to turnover intention when job satisfaction was controlled ( = .001, p > .05) thus it can conclude job satisfaction fully mediate the relationship between perceived reward and job condition to turnover intention – and job satisfaction only partially support the relationship between overall POS to turnover intention, in other words, hypothesis H3 is accepted.
4.3.3 Moderating regression analysis of Organizational reputation
In order to understand the role of organizational reputation on predicting turnover intention-behavior, in the case of this research that is “bad reputation” of Formosa Ha Tinh Steel, Vietnam, hypothesis H4 “Organizational reputation will moderate the relationship between job satisfaction and turnover intention” was proposed. This section will discuss it.
Table 4.11 Results of Organizational reputation moderating role Turnover Intention
In a causal relationship, if x is the predictor variable and y is an outcome variable, then z
is the moderator variable that affects the casual relationship of x and y if it found to be significant and can cause an amplifying (> 0, p < .05) or weaken (< 0, p < .05) effect between independent and dependent variable.
As shown in Table 4.11, the model testing with 3 steps. Firstly, same as hypothesis H2 - the interaction of job satisfaction and turnover intention was significant (= -.394, p < .001) and it still significant when organizational reputation occur in the regression because the result show effect of job satisfaction decrease (= -.368, p < .001) although organizational reputation wasn’t significance (= .111, p > .05) relate with intending turnover at step 2 in testing model. Step 3 show the moderator role of organizational reputation was significance on the relationship between job satisfaction and turnover intention (=.261, p< .001) Additionally, job satisfaction has a slightly descending affect to turnover (= -.370, p < .001) compare with step 2 and still decreasing affecting to turnover if consider organizational reputation as moderator in the regression compare with step 1 (Δ= -.024). Hence, organizational reputation can amplify the relationship between job satisfaction and turnover intention with the more increasing organizational reputation the stronger affecting of job satisfaction toward turnover intention behavior. In this case, the rising bad reputation of the company can cause to employee’ embarrassment, discomfort, decrease their self-esteem and satisfaction before, finally affect to their turnover intention behavior.
In sum, hypothesis H4 “Organizational reputation will moderate the relationship between job satisfaction and turnover intention” is accepted.
4.4 Comparing the demographic characteristic to POS and turnover intention
This section will describe the result of One-way Analysis of Variance (ANOVA) - an analysis based on influence of single factor to determine the difference mean of POS, turnover intention by using independent variable as demographic characteristics of employee include gender, age, marriage status, kids, education and working time to determine these differences affecting to turnover intention.
4.4.1 Testing demographic characteristic effect to POS
As shown in Table 4.12, the research finding only a significant relationship between the working experience time with POS. A significance level of Levene Statistic is .001 (p < .001) and the result in Table 4.8, working time show a negative correlation with POS with = -.202, p < .01, it mean there are different POS level between different group of working experience employee and the employees have longer working time may feel lower POS level from the company.
Table 4.12 One-way ANOVA result comparing POS level by working experience POS finding only a significant relationship between the number of kids with turnover intention as a table below:
Table 4.13 One-way ANOVA result comparing Turnover intention level by Kids Turnover
Sum of Squares
df Mean
Square
F Sig.
Between Groups 11.011 3 3.670 4.868 .003
Within Groups 201.306 267 .754
Total 212.317 270
The table 4.12 represents that variances of turnover intention are equal or different between a number of kids groups. A significance level of Levene Statistic is .003 (p < .05) so there is existing unequal or different turnover intention level between a different number of kids groups of the employee. Additionally, from table Table 4.8 the finding is there is the negative correlation between a number of kids and turnover intention with = -.192, p < .01 Hence, in conclusion, the higher number of kids, the lower level of turnover intention employees have.
4.5 Research Summary
H1a: Perceived fairness treatment does not affect job satisfaction.
Furthermore, the figure below shows the relationship between all variables which have figured out of this research.
Figure 4.2: Regression analysis results POS
.723
Turnover Intention Organizational
Reputation
Job Satisfaction
.767
-.183
-.394
Perceived organizational reward and job condition
Perceived fairness treatment
-.412
.261
Chapter 5
CONCLUSIONS AND RECOMMENDATIONS 5.1 Conclusions
As presented in Chapter I, the purpose of this study was to pursuit four research questions: (1) How does POS affect job satisfaction? (2) How does job satisfaction affect turnover intention? (3) What is the relationship between POS, job satisfaction and turnover intention? (4) How does organizational reputation influent job satisfaction and turnover intention process? which identified by four specific objectives: (1) Exploring the role of perceived organizational support to job satisfaction (2) Examining the impact of job satisfaction on turnover intention (3) Figuring out how perceived organizational support affects turnover intention through employee job satisfaction. (4) Figuring out the moderate effect of organization reputation on job satisfaction and turnover intention process that base on the case of Formosa Ha Tinh Steel – a very famous and long established Taiwanese company with a “bad reputation” in recently at Vietnam. This study was developed 4 hypotheses which tested by using a sample of 271 employees have been working at this company and the results gave empirical pieces of evidence to support all hypotheses.
5.1.1 POS and Job Satisfaction
The first finding is POS positively effect to employee job satisfaction. Aligning with studies conducted by Eisenberger et al., (1997, 2001) this research found out an important role of perceived organizational support from organizational human practices to enhance job satisfaction. Furthermore, organizational reward and working condition support show strong evidence to predict how much job satisfaction employees have whereas perceived fairness treatment does not have an affect directly toward job satisfaction which also was found by Thomas (2012). Because of perceived fairness treatment is concern of the ways used to
determine the distribution of resources among employees while the second factor - perceived organizational reward and job condition relating to pay, benefits, and rewards (Greenberg, 2011), hence the reasons to explain this finding is Vietnamese employees satisfaction are not much bothered about the procedural aspects – how fairness the distribution of resources does but their job satisfaction much more affected by the amount of pay or recognition, reward, promotion…that they receive from their company.
5.1.2 Turnover intention
Starting from the enormous brain-drain problem at Vietnam recently, the main aim of this study to enhance theoretical understanding of turnover intention and the result achieve this goal in several ways.
Firstly, hypotheses H2, H3 explain how perceived organizational support and job satisfaction influence intention to quit a job through social exchange theory. In detail, these findings are the low level of job satisfaction is associated with high level of intention to quit.
In other word reasons make the employee quit the job have direct expressions of job dissatisfaction that also was proven in many previous types of research (Maertz & Campion, 2004; Greenberg, 2011). Moreover, the caring, respect, recognize connoted by POS can fulfill employees socio-emotional need, amplify employee’s belief that organization approval and rewards their efforts and performances, then can lead to positive outcomes to both the employee (job satisfaction) and the organization (turnover intention) which explains the reason why job satisfaction fully mediation the relationship between perceived organizational reward and job condition support toward turnover intention, although other hand fairness treatment perceived is not a satisfied but in reality its absence can cause dissatisfaction and strong directly affect to turnover intention too.
Secondly, social identity theory was applied to elucidate how employees think outsiders see them that can lead to decrease satisfaction, self-esteem, embarrassment if their company
has a “bad reputation” then potentially triggering the decision to leave this company – hypothesis H4. That finding emphasized the importance of building a good organizational image brand because it not only affects the external customer but also affects the inside customers – employees who are the most important predictor of organizational success in the era of fierce competition.
5.2 Practical contributions
This research has some implications for managers in organizations.
First, because job satisfaction holds a major role of employee turnover intention, the higher satisfaction employee has, the less likely they are to leave the organization. Hence, managers need to focus on factors that strongly effectively enhance job satisfaction, let the employee more embedded into their jobs then remain with the organization. These factors are human practical actions inside the organization such as develop an effective and unique organizational reward system, provide more opportunities for employee advancement and development, create attractive promotion policies, give employee more chances for growth through an amount of training courses, sharing knowledge and experience programs.
Especially for those employees who have had long working years at the company, they seem
Especially for those employees who have had long working years at the company, they seem