Factors affecting job satisfaction of software engineers at haveynash vietnam master project in business administration

research in measuring the level of job satisfaction of software engineers at HVN as well as the impact of demographic and organizational factors on job satisfaction.. To measure the over

(PART-TIME)

Tutor’s Name: Dr VŨ VIỆT HẰNG

Ho Chi Minh City

(2010)

UNDERTAKINGS

I would like to undertake that, to carry out this thesis “Factors affecting job

satisfaction of software engineers at Harveynash Vietnam”, I have myself

studied and researched the project’s issue, applied the academic knowledge obtained from the MBAVB program, and discussed with my supervisor

I undertake that this is my own research project, all information, data and survey results are true and never announced in any other research projects

Ho Chi Minh City, 22 February 2011

TRAN QUOC TRI

me their practical experience

I also would like to thank all employees at Harveynash Vietnam who help me to join my survey for this project

TUTOR’S COMMENTS

It’s a pleasure to supervise Tran Quoc Tri for this final task forward completing his master in business management In the course of the project implementation, I’ve formed a good impression of the preparatory work completed by Tri as well as his responsibility and trustworthiness I’d like to highlight key accomplishment of this research report as follows

1 The author succeeded in maintaining cohesion between different parts of the research, from the literature review, the status quo to solutions recommendation

2 The research was developed on the basis of modern and solid theories

3 The author achieved the primary objective of doing a serious work of collecting data and thoroughly analyzing them to provide meaningful information and recommendations

4 The results of this research can be applied in real life situations and address them in a practical way

I would encourage the management of the author’s company to adopt the recommendations of the research to help improve the employee satisfaction at Harveynash Vietnam

Ho Chi Minh City, 22 February 2011

Dr VU VIET HANG

TABLE OF CONTENT

Chapter 1  Introduction 1 

1.1  Overview 1 

1.2  Research problem and research objectives 1 

1.2.1  Research problem 1 

1.2.2  Research objectives 2 

1.3  Research questions and hypotheses 2 

1.3.1  Research questions 2 

1.3.2  Research hypotheses 3 

1.4  Research methodology 4 

1.5  Delimitations 4 

1.6  Structure of the study 4 

Chapter 2  Literature review 6 

2.1  Concepts and definitions of job satisfaction 6 

2.2  Theories relating to job satisfaction 7 

2.2.1  Maslow’s Hierarchy of Need (1943) 7 

2.2.2  Herzberg’s Two Factor Theory (1959) 8 

2.2.3  Vroom’s Expectancy Theory (1964) 10 

2.2.4  Hackman and Oldham's Job Characteristics Model (1974) 11 

2.3  The measurement of job satisfaction 12 

2.4  Researches related to job satisfaction 17 

2.4.1  Organizational factors and job satisfaction 17 

2.4.2  Demographic factors and job satisfaction 20 

2.5  Developing the research model 21 

Chapter 3  Research Methodology 23 

3.1  Research Process 23 

3.2  Population and sample 24 

3.3  Data collection 25 

3.4  Questionnaire design 26 

3.5  Data analysis method 28 

Chapter 4  Data analysis and findings 31 

4.1  Profile of samples 31 

4.2  Means of job satisfaction 31 

4.3  Reliability analysis 34 

4.4  Factor analysis 36 

4.5  Comparison of job satisfaction by demographic factors 38 

4.5.1  Comparison of job satisfaction by gender and marital status 38 

4.5.2  Comparison of job satisfaction by years of experience 39 

4.5.3  Comparison of job satisfaction by length of service and position 40 

4.6  Regression analysis 41 

4.6.1  Assumptions checking 41 

4.6.2  Model evaluation 43 

4.6.3  Evaluation of independent variables 44 

4.7  Summary of research findings 45 

4.8  Revised research model 48 

Chapter 5  Conclusion and implication 49 

5.1  Conclusion 49 

5.1.1  Means of job satisfaction 49 

5.1.2  Personal antecedents of job satisfaction 49 

5.1.3  Organizational antecedents of job satisfaction 51 

5.2  Recommendations 51 

5.3  Contributions 52 

5.4  Limitations and recommendations for future researches 53 

REFERENCES 54 

APPENDIX 1  THE QUESTIONNAIRE 57 

APPENDIX 2  PROFILE OF SAMPLES 64 

APPENDIX 3  RELIABILITY ANALYSIS 65 

APPENDIX 4  SUITABILITY ASSESSMENT FOR FACTOR ANALYSIS 69 

APPENDIX 5  FACTOR EXTRACTION 71 

APPENDIX 6  FACTOR ROTATION 77 

APPENDIX 7  COMPARISON OF JOB SATISFACTION BY DEMOGRAPHIC FACTORS 79 

APPENDIX 8  ASSUMPTIONS CHECKING FOR MULTIPLE REGRESSION 86 

APPENDIX 9  REGRESSION ANALYSIS 87 

LIST OF ABBREVIATIONS

HVN Harveynash Vietnam

IT Information Technology

SPSS Statistical Package for Social Sciences

JDI Job Descriptive Index

JSS Job Satisfaction Survey

MSQ Minnesota Satisfaction Questionnaire

JIG Job in General

LIST OF FIGURES

Figure 1: Outline of the thesis 5 

Figure 2: Maslow’s Hierarchy of Needs 8 

Figure 3: Herzberg’s two factor theory 9 

Figure 4: Vroom expectancy theory 10 

Figure 5: Job Characteristics Model 12 

Figure 6: Six facial expressions from the Faces scale 15 

Figure 7: Proposed research model 22 

Figure 8: The research process 23 

Figure 9: Revised research model 48 

LIST OF TABLES

Table 1: Internal consistency reliability for the Job Satisfaction Survey 14 

Table 2: Items from the Michigan Organizational Assessment Questionnaire Satisfaction subscale 16 

Table 3: Measurement scales of satisfaction level 27 

Table 4: Divisions of facets satisfaction into 36 items 27 

Table 5: Mappings between satisfaction scores and satisfaction levels 32 

Table 6: Means of the items in construct “pay” 32 

Table 7: Summary of satisfaction scores 33 

Table 8: The results of the One-Sample T-test 33 

Table 9: Results of reliability analysis of items 35 

Table 10: Factor analysis results 38 

Table 11: Multicollinearity check results for independent variables 42 

Table 12: Regression model summary and statistical significant of the results 43 

Table 13: Multiple regression results 44 

Table 14: Summary of hypothesis testing results 47 

IT employees also suggested that “IT employees seem to be quicker to change jobs than other employees when they are dissatisfied with their employer” (Hacker, 2003)

The purpose of the study is to understand the satisfaction level of software engineers at HVN and the factors affecting job satisfaction

1.2 Research problem and research objectives

1.2.1 Research problem

Harveynash Vietnam (HVN) is one of the biggest software outsourcing companies

in Vietnam As HVN is rapidly expanding, the number of employee at software division is increasing year over year By the end of December 2009, the total software engineers at HVN is 323, 20% increase compared to the year 2008 However, beside the expansion, the turnover is also increasing In the year 2009, there were 47 software engineers left HVN As job satisfaction is strongly correlated with turnover intentions, studying on job satisfaction is considered as the first step to help retain software engineers at HVN This gives motivation for this

research in measuring the level of job satisfaction of software engineers at HVN as well as the impact of demographic and organizational factors on job satisfaction

1.2.2 Research objectives

In attempting to answer the research problems, this research is directed to the following objectives:

1 To measure the overall satisfaction of software engineers at HVN as well as

the level of satisfaction with facets of their job

2 To determine the demographic factors affecting overall job satisfaction

3 To measure the level of impact of organizational factors on overall job satisfaction of software engineers at HVN. 

1.3 Research questions and hypotheses

Demographic and organizational factors that found to be associated with job

satisfaction will be elaborated in Error! Reference source not found The

research questions and hypotheses are developed based on these factors

1.3.1 Research questions

This research investigates the following research questions:

Q1 What is the overall job satisfaction of software engineers at HVN?

Q2 What degree of satisfaction do software engineers express with each of the

facets of their job such as pay, promotion, supervision, fringe benefits, contingent rewards, working environment, coworkers, nature of work and communication?

Q3 Are there significant differences in job satisfaction of software engineers at

HVN by their personal factors such as gender, marital status, years of experience, present position, and number of years working at HVN?

Q4 What is the level of impact of organizational factors such as pay, promotion,

supervision, fringe benefits, contingent rewards, working environment, coworkers, nature of work and communication on the overall job satisfaction

of the software engineers at HVN?

1.3.2 Research hypotheses

The following hypotheses were proposed for the research

Hypothesis H1: There are significant differences in job satisfaction of software

engineers at HVN by demographic factors such as gender, marital status, years of experience, present position, and length of service

H1a:There is significant difference in job satisfaction between male and female

software engineers

H1b: There is significant difference in job satisfaction between single and married

software engineers

H1c: There is significant difference in job satisfaction between software engineers

having different length of service for the company

H1d: There is significant difference in job satisfaction between software engineers

having different years of experience

H1e: There is significant difference in job satisfaction between software engineers

working in different positions

Hypothesis H2: The organizational factors such as pay, promotion, supervision,

fringe benefits, contingent rewards, working environment, coworkers, nature of work and communication are positively related to the overall job satisfaction

H2a: Pay is positively related to the overall job satisfaction

H2b: Promotion is positively related to the overall job satisfaction

H2c: Supervision is positively related to the overall job satisfaction

H2d: Fringe benefits are positively related to the overall job satisfaction

H2e: Contingent rewards are positively related to the overall job satisfaction

H2f: Operating procedures are positively related to the overall job satisfaction H2g: Coworkers are positively related to the overall job satisfaction

H2h: Nature of work is positively related to the overall job satisfaction

H2i: Communication is positively related to the overall job satisfaction

1.4 Research methodology

In this research, quantitative method is applied to examine the research questions and hypotheses A structured survey questionnaire was administered to collect primary data from software engineers working at HVN.Data were collected using

an online survey designed in Google Docs The Statistical Package for Social Sciences (SPSS) v.16 was utilized to analyze the data

Moreover, the sample size in this research is quite small due to time limitation A large sample will make the findings more trustworthy

Finally, the study was conducted at HVN so the results and implications of the study are applicable for only HVN However, the research model and research methods used in this study are not limited to the context of HVN

1.6 Structure of the study

This thesis consists of five chapters Chapter 1 provides an overview of the

research context The research problem and questions are identified, and the research model and hypotheses are constructed Brief introduction about the research methodology used for this study as well as delimitations of this study are

also discussed Chapter 2 presents the literature reviews carriedout for this

research Concepts and definitions of job satisfaction, together with theories relating

to job satisfaction, the measurement of job satisfaction and researches related to job

satisfaction are discussed.Chapter 3 deals with the methodology of research employed in this thesis Chapter 4describes the analysis of the data and testing of the hypotheses of the proposed model and the last chapter, Chapter 5, presents

conclusions and implications of the research.Figure 1provides a conceptual outline

of this thesis

Figure 1: Outline of the thesis

Source: Developed for this research

Chapter 1

Introduction

Chapter 2 Literature  review

Chapter 3 Research  methodology

Chapter 4 Data analysis  and findings

Chapter 5 Conclusion  and  implication

CHAPTER 2

LITERATURE REVIEW

2.1 Concepts and definitions of job satisfaction

Job satisfaction has been defined in a number of ways According to Vroom (1967, cited in Kaya, 1995), job satisfaction is the reaction of the workers against the role they play in their work.Rue and Byars (1992, cited in Luddy, 2005) referred to job satisfaction as an individual’s mental state about the job This definition was expanded by Greenberg and Baron (1995) who defined job satisfaction as an individual’s cognitive, affective and evaluativereactions toward their job

Schermerhorn (1993) defined job satisfaction as an affective or emotional response towards various aspects of an employee’s work According to the author, causes of job satisfaction include status, supervision, co-worker relationships, job content, remuneration and extrinsic rewards, promotion and physical conditions of the work environment, as well as organizational structure

According to Kreitner and Kinicki (2007), job satisfaction reflects the extent to which individuals like their jobs It is employee’s feeling toward his/her job

Spector (1997) defined job satisfaction as the extent to which people like or dislike their jobs This definition suggests job satisfaction as a general or global affective reaction that individuals hold about their job This view is supported by Ellickson and Logsdon (2002) who defined job satisfaction as the extent to which people like their work

According to Cherrington (1994, cited in Luddy, 2005), the concept of job satisfaction could be understood as either facet satisfaction or overall satisfaction Facet satisfaction refers to the tendency for an employee to be more or less satisfied with various facets of the job such as pay, the work itself, compensation, career

opportunities… Overall satisfaction refers to a person's attitude toward his or her job that cuts across the various facets

2.2 Theories relating to job satisfaction

Several theories have been proposed to explain the concept of job satisfaction, including Maslow’s hierarchy of need, Herzberg’s two factor theory, Vroom’s expectancy theory and Job Characteristics Model by Hackman and Oldham

2.2.1 Maslow’s Hierarchy of Need (1943)

One of the most widely recognized theory is Adam Moslow’s need-based theory of motivation Maslow's hierarchy explains human behavior in terms of basic requirements for survival and growth These requirements, or needs, are arranged according to their importance for survival and their power to motivate the individual According to the author, people’s need could be classified into five levels of needs: physiological needs, safety needs, social needs, esteem needs andself-actualization needs The details of these 5 levels were shown in Figure 2 Physiological needs are the first and lowest level of needs These are the basic requirements for human physical survival They include such essentials as food, water, shelter, oxygen, and sleep Safety needs include the needs for safety, protection against threads, fair treatment, job security, etc.Social needs refer to the needs of love and being love, accepted, part of the group or organization, etc whereas esteem needs refer to the needs for being respected, being recognized and being appreciated Finally, self-actualization needs, the highest in Maslow’s pyramid, refer to the desire to become more and more what one is, to become everything that one is capable of becoming

Maslow stated that lower needs must be reasonably well satisfied in order for the person to focus on higher needs, but he noted that complete satisfaction of a given need may not be possible or necessary Moreover, once a need is fulfilled, it disappears as the individual moves on to the next level

Hierarchy

kipedia.org

Theory (1

uenced by vation fact

o pointed rather no jction but n

of Needs

g

1959)

two factotors whilstout that tjob satisfa

no dissatisf

ors Satisfa

t dissatisfathe opposaction; andfaction

action andaction is aite of job

d similarly,

d

a

b ,

Figure 3: Herzberg’s two factor theory

Source: Developed for this research

Hygiene factors are based on the need for a business to avoid unpleasantness at work If these factors are considered inadequate by employees, then they can cause dissatisfaction with work Hygiene factors include company policies, supervision, salary, inter-personal relations and working conditions

Motivator factors are based on an individual's need for personal growth When they exist, motivator factors actively create job satisfaction Motivator factors include achievement, recognition, the work itself, responsibility and advancement

Hygiene factors caused dissatisfaction by their absence, but did not cause motivation by their increased presence Motivation factors are needed to motivate

an employee to higher performance

According to Schermerhorn (1993), Herzberg’s two-factor theory is an important frame of reference for managers who want to gain an understanding of job satisfaction and related job performance issues Schermerhornasserts that Herzberg’s two-factor theory is a useful reminder that there are two important aspects of all jobs: what people do in terms of job tasks (job content), and the work setting in which they do it (job context) Schermerhorn suggested that managers should attempt to always eliminate poor hygiene sources of job dissatisfaction in the workplace and ensure building satisfier factors into job content to maximize opportunities for job satisfaction

Like other theories, there are criticisms for this theory such as Hygiene factors and motivators vary depending on the types of individuals involved and the nature of the

work examined There is no evidence to prove that highly satisfied people are also high performers (Sultan, 2004)

2.2.3 Vroom’s Expectancy Theory (1964)

Vroom's expectancy theory argues that motivation is based on values and beliefs of individuals and examines motives through the perception of what a person believes will happen It is based on the belief that employee effort will lead to performance and performance will lead to rewards (Vroom, 1964) The theory states that individuals can be motivated if they believe that there is a positive correlation between the efforts they put in and their performance and when that favorable performance leads to a reward Consequently, the reward helps satisfy an important need and the desire to satisfy that need is strong enough to make the efforts worth wile

Figure 4: Vroom expectancy theory

Source: Developed for this research

According Vroom, the conditions for motivation can be summarized as

Motivation = V.I.E

E (expectancy): The person's perception that effort will result in performance In other words, the person's assessment of the degree to which effort actually correlates with performance

I (instrumentality): The person's perception that performance will be rewarded In other words, the person's assessment of how well the amount of reward correlates with the quality of performance

V (valence): The perceived strength of the reward or punishment that will result from the performance If the reward is small, the motivation will be small, even if expectancy and instrumentality are both perfect (high)

Like others, there are some criticisms on Vroom’s theory such as research evidence

so far has not supported the expectancy theory, and this is a very complicated theory

2.2.4 Hackman and Oldham's Job Characteristics Model

(1974)

Job characteristics model, proposed by Hackman and Oldham (1974) is a widely used framework to particular job characteristics impact on job outcomes, including job satisfaction According to the this model, there are five core job characteristics (skill variety, task identity, task significance, autonomy, and feedback) which impact three critical psychological states (experienced meaningfulness, experienced responsibility for outcomes, and knowledge of the actual results), in turn influencing work outcomes (job satisfaction, absenteeism, work motivation, etc.) The first three dimensions (skill variety, task identity, and task significance) combine to create meaningful work The jobs that possess autonomy give the worker a feeling of personal responsibility for the results If a job provides feedback, the worker will know how well he or she is performing The more that these three elements are present, the greater will be the employee’s motivation, performance, and satisfaction

Figure 5: Job Characteristics Model

Source: Developed for this research

2.3 The measurement of job satisfaction

Green (2000) pointed out that unlike productivity, absenteeism, and turnover, job satisfaction is present only inside an individual’s mind and cannot be measured directly Job satisfaction could be measured indirectly using both global and facets measures Global measures focus on overall feelings about the job and are used to predict behavior such as quitting Facet measures focus on satisfaction with specific aspects of the job and used to diagnose strengths and weaknesses at an organization

or in a workgroup(Ironson et al., 1989) Facets might include satisfaction with workers, fringe benefits, pay, supervision… Some studies measures both global and specific facet job satisfaction and also calculate the relationship between the global satisfaction and the facets satisfaction

co-The most widely used method to measure job satisfaction is using self-report questionnaires Popular facets measures include the Job Descriptive Index (JDI; Smith et al., 1969), the Job Satisfaction Survey (JSS; Spector, 1997), and the Minnesota Satisfaction Questionnaire (MSQ; Weiss et al., 1967) Overall job

satisfaction are usually measured using global scales such as the Job in General Scale (JIG;Ironson et al., 1989), the Faces scale (Kunin, 1955), the Job Satisfaction Scale (Brayfield and Rothe, 1951) and the Michigan Organizational Assessment Questionnaire satisfaction sub-scale (Cammann et al., 1979)

One of the most popular measures of job satisfaction is the Job Descriptive Index (JDI; Smith et al., 1969) The JDI assesses job satisfaction with five facets, including work, pay, promotion, supervision and coworkers The scale contains 72 items with either 9 or 18 items per subscale Each item is an evaluative adjective or short phrase that is descriptive of the job Respondents read items and evaluate whether each describes their job by answering with a “yes”, “no” or “uncertain” The most limitation of JDI is that it assesses only five facets, although that these five facets are the most widely used in most scales Another limitation is that the JDI is copyrighted and a fee is required for its use

Although the JDI is widely used and has good reliability and validity evidence, Smith and her colleagues continue to put more effort to improve it The most recent version of the JDI, published in 2009 includes a abridge scale called Abridge Descriptive Index containing 38 items to measure both facets satisfaction and overall satisfaction

The Job Satisfaction Survey (JSS; Spector, 1997) assesses 9 facets of job satisfaction, including pay, promotion, supervision, fringe benefits, contingent rewards, working environment, coworkers, nature of work, and communication The scale contains 36 items and uses summatedrating scale format, the most popular format for job satisfaction scales Each of the 9 facets subscales contains4 items and the satisfaction score for each facet can be computed by summing up the score of the 4 items JSS uses the Likert scale containing 6 choices from 1 (strongly disagree) to 6 (strongly agree) with items written in both directions, requiring about half of them to be reversed score

Each of the items is a statement that is either favorable or unfavorable about an

aspect of the job Respondents are asked to select one of the 6 numbers that

corresponds to their agreement or disagreement about the item.The positive items

are score from 1 to 6 whilst the negative items are scored from 6 to 1 By summing

up the 4 items, each facet score ranges from 4 to 24 The total score ranges from 36

to 216

The author reported the coefficient alphas ranged from 0.60 for the coworkers

subscale to 0.91 for the total scale The JSS scale was originally developed for the

social service sector However, it has been widely used in many studies on various

fields

Table 1: Internal consistency reliability for the Job Satisfaction Survey

Subscale Coefficient Alpha Test-Retest Reliability

One of the problems with the JSS scale is that it measures the overall satisfaction by

summing up the score of all the facets Scarpello and Campbell (1983) found that

individual questions about aspects of the job did not correlate well with a global

measure of overall satisfaction Based on the results, the authors argued that faceted

and global measures do not measure the same construct In other words, the whole

is not the same as the sum of the parts

Another popular scale is the Minnesota Satisfaction Questionnaire (MSQ; Weiss et al., 1967) The scale is available in two forms, the long form with 100 items and the short form containing 20 items It measures 20 facets, includingactivity, independence, variety, supervision (human relations), supervision (technical), social status, moral values, security, social service, authority, ability utilization, company policies and practices, compensation, advancement, responsibility, creativity, working conditions, coworkers, recognition and achievement

MSQ allows one to measure both intrinsic and extrinsic satisfaction score Intrinsic satisfaction refers to the nature of job tasks themselves and how people think about the work they do Extrinsic satisfaction concerns whether employees are satisfied with aspects that are extrinsic, or external to their job tasks, such as pay and benefits

Overall satisfaction can be measured using either single-item or multiple-item scales Many researchers believe that a single item is enough to measure satisfaction Wanous, Reichers, &Hudy (1997) concluded that singleitem of job satisfaction may work well in many situations The most famous scale of this type is the Faces scale developed by Kunin in 1955 As its name implies, the Faces scale uses 6 facial expressions as its response format Respondents are asked to circle the face that reflects the way they feel about their job

(JIG; Ironson et al., 1989) Due to the fact that JDI does not contain an overall measure of job satisfaction, the JIG scale was designed to be used in conjunction with JDI It contains 18 items which are adjectives or short phrases and respondents use the same response options as on the JDI Internal consistency for the scale ranged from 0.91 to 0.95 across different samples Correlations of the JIG with other global job satisfaction scales ranged from 0.66 to 0.80, providing evidence of the validity for the scale (Ironson et al., 1989)

The Michigan Organizational Assessment Questionnaire (Cammann et al., 1983) contains three items using the 7-point Likert scale The items of the scale are shown inTable 2 The scale is simple and short, which makes it ideals for use in questionnaires that contain many scales The responses are number from 1 to 7, respectively, but the second question is reversed scored The items are totaled to generate the overall satisfactions score The authors report an internal consistency reliability of 0.77

Table 2: Items from the Michigan Organizational Assessment Questionnaire

Satisfaction subscale

1 All in all I am satisfied with my job

2 In general, I don’t like my job

3 In general, I like working here

Source: Spector, 1997

Judge and colleagues (2000) used a short form of the Brayfield and Rothe (1951) Job Satisfaction Scale.The scale uses 5 items that appears to work well as a measure

of overall job satisfaction:

- I feel fairly satisfied with my present job

- Most days I am enthusiastic about my work

- Each day at work seems like it will never end

- I find real enjoyment in my work

- I consider my job to be rather unpleasant

Five-point Likert scale was used and the authors reported the internal consistencies

of 0.80, which is fairly high

According to Spector (1997), there are two good reasons to use multiple-item scale rather than single-item scale First, and most important, multiple-item scales are more reliable than single-item scales The reason is that respondents may take mistake when they fill the questionnaire With single-item scale, the error can move the respondent’s result from one end of the scale to the other With multiple-item scale, the effect of the error could be reduced because it only contributes a portion

to the final result Second, multiple-item scale allows making more complete assessment of satisfaction People’s feelings toward their job may not be simple Job satisfaction is a broad issue, and it includes many aspects A single item may not cover all aspects of the job and seems to over-simplify people’s feelings toward their job Multiple items allows for more specific questions and help to make a more complete assessment of the job

This research measured both facets satisfaction and overall satisfaction Based on the results, the correlations between the facets and overall satisfaction were calculated to identify which facets most affect on overall satisfaction The facets satisfaction was measured using the JSS while the five items version of the Job Satisfaction Scale was utilized to measure the overall satisfaction

2.4 Researches related to job satisfaction

2.4.1 Organizational factors and job satisfaction

Many organizational factors such as pay, promotion, supervision, coworkers… were found to have significant association with job satisfaction

Boeve (2007) conduct a research to examine the job satisfaction factors for physical assistant faculty using the Herzberg’s two-factor theory and the JDI scale Job satisfaction factors were divided into two categories: intrinsic factors about the respondents (work itself and opportunities for advanced) and extrinsic factors about the institutional faculty support (salary, supervisory support, and coworker relations) Cronbach’s alpha was computed to measure the internal consistency of the five JDI factors and the overall job satisfaction scale The mean and standard deviation for each factor was documented Spearman’s correlation was computed for the JDI factors’ relationship to overall satisfaction Multiple regression analysis was used to determine the predictors of overall satisfaction

The results from the research showed that four of five JDI factors (work itself, coworker relations, advancement opportunities, and supervisory support)are found

to have positive relationships to overall job satisfaction The fifth factor, salary, has

a low relationship and no predictive value to the overall job satisfaction of the physical assistant faculty Besides the four factors from JDI, years of physical assistant education experience is a significant predictor for job satisfaction.Boeve’s research did supported Herzberg’s two-factor theory and Smith’s theoretical framework

The research conducted by Luddy (2005) used the JDI scale (Smith et al., 1969) to measure the levels of job satisfaction experienced amongst employees at a public health institution in the Western Cape region Results indicated that employees at the public health institution in the Western Cape expressed satisfaction with their coworkers, followed by the nature of the work and the supervision they receive Opportunities for promotion and pay emerged as major sources of dissatisfaction Although the research indicated that job satisfaction is significantly related to variables such as work, remuneration, supervision, promotion, and coworkers, the author stated that ongoing research is required in this domain However, the role of other potentially confounding extraneous variables on job satisfaction needs to be contemplated for future research

Steven and John (2008) used the JSS (Spector, 1997) to measure the job satisfaction and assess the job satisfaction factors most significantly related to software developer turnover intentions The overall level of job satisfaction of software developers was reported as 4.05 which can be interpreted as slightly satisfied Supervision, benefits, coworkers, nature of work had a high mean value of 4.827 (SD=1.214), 4.323 (SD=1.123), 4.641 (SD=0.958), and 4.769 (SD=0.993) respectively which indicated that software developers were moderately satisfied with these four facets The results also showed that software developers were slightly satisfied with pay(Mean=3.629, SD=1.301), contingent rewards (Mean=3.850, SD=1.259), working condition (Mean=3.718, SD=0.978), and communication (Mean=3.722, SD=1.128) while they were slightly unsatisfied with promotion (Mean=2.951, SD=1.263)

Using the JSS, Sharaf et al (2008) measured the level of job satisfaction among primary care physicians The results showed that physicians were slightly satisfied with their job (Mean=3.46, SD=0.67) They also found that physicians were moderately satisfied with supervision (Mean=4.62, SD=1.20), coworkers (Mean=4.58, SD=0.86) and nature of work (Mean=4.69, SD=1.06) while slightly satisfied with communication (Mean=3.80, SD=1.09) Physicians were slightly unsatisfied with pay (Mean=2.76, SD=1.26), promotion (Mean=2.56, SD=1.12), fringe benefits (Mean=2.65, SD 1.09), contingent rewards (Mean=2.61, SD=1.15), and operating condition (Mean=2.85, SD=0.71)

Worrell (2004) used a modified version of the 1977 Minnesota Satisfaction Questionnaire (MSQ) to measure school psychologist’s job satisfaction Results indicated that 90% of school psychologists were satisfied or very satisfied with their jobs The findings showed a gradual increase in overall job satisfaction when compared to the 85.7% in 1982 and the 86% in 1992 who reported being satisfied or very satisfied with their jobs Participants in the current sample were more satisfied with their job security, independence, and creativity The only variables

demonstrating a significant relationship with job satisfaction were the intent to remain in current position and supervisor certification

2.4.2 Demographic factors and job satisfaction

Studies investigating job satisfaction indicated that personal determinants such as gender, marital status, years of experience… were significantly associated with job satisfaction

Athanosios (2001) surveyed on personal characteristics and job satisfaction of Greek teachers and found that teachers satisfied with the work itself and supervision, whereas they are unsatisfied with pay, and promotion opportunities The study also revealed that certain characteristics such as age, gender, marital status and years of teaching were significant predictors of job satisfaction

Osagbemi (2003) studied the correlations between personal factors and job satisfaction Collecting a large sample from UK universities, the study found that the rank of an individual and the length of service are significant predictors of job satisfaction Length of service in higher education was found to be negatively related Gender and age were not found to be significantly associated directly with job satisfaction

Robbins et al (2003) argue that no evidence exists suggesting that gender impacts

on an employee’s job satisfaction The authors are of the opinion that gender differences can have an effect on the relationship between job dimensions and job satisfaction, but that it does not have a direct impact on job satisfaction Similarly, Donohue and Heywood (2004) could not prove gender satisfaction differences in a study conducted amongst young American and British employees

Sarker et al., (2003) conducted a research on the relationships of age and length of service with job satisfaction The results indicated that employee age is not significantly associated with overall satisfaction level, but that tenure is

John and John (2008) conducted a research on the effects of demographic factors on job satisfaction The study used survey data collected from 356 staff members from Catholic primary schools and found that age, gender, position were related to job satisfaction

2.5 Developing the research model

Based on the theories and researches revised, the research questions and research hypotheses were proposed as described in Chapter 1 The model for the research was also proposed as shown in Figure 7

Five demographic factors studied in this research are gender, marital status, years of experience, length of service and position The factor “age” was not studied because for software engineers, age and years of experience are strongly correlated Most software engineers start their career at the age from 22 to 25 after graduated

Nine facets of job in the JSS scale including pay, promotion, supervision, fringe benefits, contingent rewards, working environment, coworkers, nature of work and communication were studied for the impact on overall satisfaction

Figure 7: Proposed research model

Source: Developed for this research

Overall  satisfaction Pay

Operating  procedures

Years of  experience

Length of  service

Position

e 8: The re

Developed

CH M

First, a draproblem anection Thehis stage Trious sour

of construc

n to Vietnauistics ex

TER 3

LOGY

estionnaireframeworkcts and theamese wasperts and

Once a draft version of the questionnaire was generated, a pilot study was conducted using a focus group of 6 employees The purpose of this step is to make sure that the questionnaire was clear and unambiguous The selected employees were asked to complete the survey and then, a discussion was hold to discuss on the structure of the questionnaire, questions wording… Items that could cause respondents’ misunderstanding or confusing were discussed and the wording was then modified At the end of this step, the final version of the questionnaire was generated

The last step of the research involves activities for collecting and processing data for the main survey The final version of the questionnaire was sent to legible HVN software engineers Results were then imported andanalyzed using statistical analysis packages.First, the reliability and validity of the scale was access by calculating the Cronbach’s Alpha coefficient.Exploratory factor analysis (EFA) was used to do the factor analysis Various tests such as ANOVA, T-test… were performed to assess whether there are significant differences of job satisfaction of software engineers at HVN by their personal factors Regression analysis was used

as the final step A series of multiple-regression analysis was performed to calculate the level of impact of each of the organizational factors on the overall job satisfaction

3.2 Population and sample

The study was conducted at Ho Chi Minh City office of Harveynash Vietnam As mentioned in the first section of the study, the purpose of this study is to measure the job satisfactionand identify the factors affecting software engineers’ job satisfaction at the Software Development Division, HVN The population of the study wasall software engineers working at the Software DevelopmentDivision, HVN The size of the population was 353, including 236 developers, 86 testers, 14 business analyst and 17 project managers At the time of conducting the

study(10/2010), there was 8 employees were on vacation so only 345 software engineers were surveyed

A non-probability sampling design in the form of aconvenience sampling method was adopted and considered to be appropriate to gatherthe data The rationale for using this sampling method was due to the respondents being easily accessible, their availability, as well as it being less time consuming andinexpensive to gather the research information

3.3 Data collection

Using the list of software engineers provided by the Human Resource Department,

an email was sent to each software engineer containing the link to the online job satisfaction survey Information about the purpose of the study as well as the permission from the board of management to conduct the survey was also mentioned in the email

The rational of using the web based online survey is that it has many advantages compared to other survey methods.First, the data is available in real time and can be easily imported to data analysis programs Each time a new response is completed, the data is immediately available and can be imported directlyinto data analysis programs Second, using a web based survey helps reduce the error rates Researcher can add validations to validate the response before submitting Moreover, the respondents in this study are software engineers, who access internet daily and very familiar with web based surveys like this Completing an online form

is more convenient for them than paper based ones

To encourage respondents complete the survey, the researcher explained clearly in the email that the respondents’ name were not requiredfor the questionnaire, and that the information acquired would be kept confidentially and only be used for research purposes.The respondents were given two weeks to complete the survey One week after sending out the request email, a reminder email was sent to remind those had not completed the survey After two weeks, the survey link was

deactivated and the data was filtered and encoded manually before importing to statistical packages for analysis

3.4 Questionnaire design

The questionnaire consisted of 4 parts Part 1 of the questionnaire is the introduction about the study, its purpose, and instruction on completing the survey Some terms used in the survey that are unambiguous and may cause misunderstanding were also clearly explained in this section

Part 2 of the questionnaire was designed to measure the overall job satisfaction and satisfaction with facets of the job The facets satisfaction was measured using the Job Satisfaction Survey (Spector, 1997) For overall satisfaction, the five-item version of the Job Satisfaction Scale (Brayfield and Rothe, 1951) suggested by Judge was utilized

There are two good reasons to choose JSS as an instrument to measure facet satisfaction The first reason is because of its reliability and validity The nine subscales related moderately to well between each other with the internal consistency ranged from a score of 0.60 for coworker to 0.91 for the total scale Test-retest reliability over an 18-month time span ranged from 0.37 (benefits) to 0.74 (operating procedures) Studies using various scales for job satisfaction on a single employee, supported validity A correlation of 0.61 for coworkers to 0.80 for supervision was calculated between 5 of 9 facets of the Job Satisfaction Survey and those of the Job Description Index

Another reason for choosing JSS is that this instrument is available free of charge for research purpose This is what we cannot get from other popular scales such as JDI or MSQ

Each item in this section used a 6-point Likert scale that measured the degree of agreement with the given statement Each item was score from 1 to 6 The score for each facet was calculated by summing up the items that build up the scale and divided by the number of items in that facet About half of the items are worded in

negative format and were reversed scored before summing up with other positive

items

Table 3: Measurement scales of satisfaction level

Satisfaction Level Positive scores Negative scores

Source: Developed for this research

Table 4: Divisions of facets satisfaction into 36 items

Source: Developed for this research

Part 3 of the questionnaire contains items used to measure the overall job

satisfaction The five-item version of the Job Satisfaction Scale (Brayfield and

Rothe, 1951) suggested by Judge and colleagues was utilized To make the overall

satisfaction scale more consistent with the facet scale, the 6-point Likert scale was

used instead of the original 5-point scale The items in this section are in the same format as the items in Part 2 Two of the 5 items are in negative format and were reversed score before summing to generate the overall satisfaction

Part 4 of the questionnaire was designed to survey personal factors of the respondents Five demographic questions, which address respondents’ gender, marital status, current position, years of experience and length of service were used.The questions in this sectionare in either the form of multiple choices or filling

3.5 Data analysis method

The research data were statistically analyzed by means of the Statistical Package for Social Sciences (SPSS) 16.0

The consistency and reliability of the scale were assessed using theCronbach’s alpha coefficient and factor analysis Cronbach’salpha coefficient was used as the first step to identify and remove unreliable items

r N

r N

⋅

−+

⋅

=

)1(1α

Where, α represents coefficient of the reliability

N represents number of items on the scale

r represents average inter-item correlation among the items

Cronbach’salpha coefficient evaluates the items in the scales to determine if they are measuring the same construct or if they are redundant, suggesting which items should be discarded to improve the homogeneity of the scale The more homogeneity the items in the scale are, the higher the Cronbach’s alpha will be Many researchers agree that when Cronbach’s alpha value equals or greater than 0.8 the scale can be considered reliable Cronbach’s alpha value equals or greater than 0.7 is acceptable In this research, to ensure the reliability, only factors with Cronbach’s alpha values greater than 0.7 are considered as reliable In addition, the item-total correlation was also used to check the reliability and only variables had item-total correlation value greater than 0.4 were kept

Once unreliable items were identified, they were removed from the scale and factor analysis was performed Factor analysis is a collection of methods used to examine how underlying constructs influence theresponses on a number of measured variables In this research, factor analysis was used to determine the constructs we can group from the 36 items of the JSS scale and whether the 5 items of the Job Satisfaction Scale measure the construct of overall satisfaction Several steps were carried out for factor analysis.Before conducting factor extraction, the correlation matrix, the Kaiser-Meyer-Olkin (KMO) and the Bartlett’s test of sphericity were checked to ensure that the characteristics of our data set were suitable for factor analysis For a good factor analysis, there should be many items in the correlation matrix with the value of 0.3 or higher; the KMO should have a minimum value of 0.6 and the Bartlett’s test of sphericity should be significant (p < 0.05).Feature extraction was performed using the principal component analysis as the extraction method The Varimax method was used to detect which items clump together to build separate constructs

The next step is to assess whether there are significant differences of job satisfaction

of software engineers at HVN by their personal factors Both parametric and parametric methods were utilized in this step To test whether there are significant differences of job satisfaction by gender, marital status, both T-Test and Mann-

non-Whitney methods were utilized Similarly, the differences of job satisfaction by years of experience, present position and length of service were tested using Oneway ANOVA

Multiple regression was performed in an attempt to determine the level of impact of each of the nine factors on the overall satisfaction The dependent variable is the overall job satisfaction and independent variables are the factors extracted in factor analysis

Y = B0 + B1*X1 + B2*X2 + … + Bk*XkWhere

Y: dependent variable (overall satisfaction)

population, the F-test was utilized

In addition, underlying assumptions of multiple linear regression such as linearity, normality, autocorrelation, constant of variances, and no multicollinearity were checkedto make sure that the data set was appropriate for multiple regression

analysis and that the results of the regression analysiswas trustworthy Normality was checked using the Histogram and Q-Q plot Assumptions about linearity and homoscedasticity (constant of variances) were checked by looking at the

residualsscatterplots The Durbin-Watson statistic was utilized to detect the present

of autocorrelation Finally, multicollinearity was checkedby inspecting the variance inflation factor (VIF)