This chapter contained the methodology that was adopted for the study. It comprised of the research design and framework, population and sampling procedure, instrumentation, validity and reliability of the research instruments, data collection procedure, data analysis as well as the evaluation of the research methods. The design was however, strongly influenced by the research questions and by the findings from the review of the literature that are contained in chapters one and two respectively.
The study was an explanatory quantitative research work as the intention was to provide credible and trustworthy explanations on the factors behind the high employee turnover in The Gambia service industry using a survey research method. According to Babbie (2011), surveys may be used for explanatory analysis and are essential in studies that have individuals as units of analysis. The data was collected using a questionnaire containing closed-ended questions in the Likert’s scale format where the respondent selected from a list of answers.
Research Framework and Hypothesis
The study was conducted in The Gambia and attempted to dissect and establish the relationship between organizational culture, compensation, job satisfaction and turnover intention in the country’s service industry. As indicated earlier, employee turnover in the sector is currently at a record’s high. The research framework below is the product of the literature review on the subject as well as the research questions. The study was conducted within the framework and the variables it contained served as the focal points for the study. It comprised of the independent variables, a mediating variable as well as a dependent variable. The dependent variable was employee turnover intention in The Gambia service industry. The purpose was to diagnose the three-work related parameters namely; organizational culture, compensation and job satisfaction that could influence employee turnover intention.
28
Figure 3.1. Research Framework Research Hypothesis
As a result of the review of literature on the research topic and with the aid of the research purposes, the following hypotheses were formulated. They served as the focal point for the study and are as follows:
Hypothesis 1: There is a positive relationship among organizational culture, compensation and turnover intention.
Hypothesis 2: Organizational culture and compensation are positively correlated to job satisfaction.
Hypothesis 3: Job satisfaction is negatively correlated to turnover intention.
Hypothesis 4: Job satisfaction will mediate the relationship between organizational culture, compensation and turnover intention.
Organizational Culture
Compensation
Job
Satisfaction
Turnover Intention H4
H3
H1 H2
29 Variables
The independent variables for the study were organizational culture, compensation and job satisfaction. In addition, job satisfaction also served as the mediating factor between organizational culture, compensation and employee turnover intention. The theoretical research framework presented in Figure 3.1 above attempts to capture the relationship between organizational culture, compensation and job satisfaction in determining employee turnover and was carefully formulated after a review of current materials on the phenomenon. Based on the researchers knowledge of the industry, environment and culture, we postulated that organizational culture and employees compensation if strategically not handle properly has immense potentials to result to turnover which is been indicated by our first hypothesis (H1). In additional, we believe organizational culture and compensation shared some sort of relationship with employees’ job satisfaction. Against this backdrop, we anticipate that organizational culture and compensation do have a positive relationship with job satisfaction and that is being indicated by our second hypothesis (H2). This is a crucial relationship because we believed even if an organization’s culture as well as the level of compensation is poor or not really good but the employees are contended with their situations, turnover would not pose a serious challenge and this is indicated by our fourth hypothesis (H4). Therefore, job satisfaction is negatively correlated to turnover intention which is implied in our third hypothesis (H3).
Research Method
This study investigated the relationship of three work-related variables, organizational culture, compensation and job satisfaction in influencing turnover in The Gambia service industry took the form of an explanatory quantitative study as the purpose was to explain the causes of employee turnover with respect to these key variables. Instruments developed and used by previous researchers with high reliability scales were used for the purpose of data collection.
Since the population for the study was large and diverse, the study chose a sampling approach that enabled the researchers access a sample group where most was collected and learnt. Upon the collection of the data, statistical analysis followed suit in order to determine the underlining
30
factors which are reported in the next chapter. In order to get a proper insight of the research methodology, a closer look at the systematic approach taken for the study is provided below:
Research Procedure
The research procedure of this study comprises of 10 steps, see Figure 3.2 below:
Step 1: Review of Literature
With particular attention to the research agenda, all materials accessed were reviewed to gain understanding on what has been established on the area as well as the research direction.
Step 2: Construction of Research Questions and Hypothesis
Formulate the research questions and hypothesis with the help of the findings of the literature review as well as the research objectives. They do not only guide but form the fulcrum of the study.
Step 3: Research Framework Development
Design a theoretical research framework according to the topic and the discussions of the literature review and then clearly identify the independent and dependent variables base on the proposed hypothesis in this study.
Step 4: Research Design
Due to the manner the research questions are structure and formulated, a correlational design was selected due to its effectiveness in enabling the assessment of relationships.
Step 5: Target Population
Based on the topic of the study and the research intention, the population for the study is identified.
31 Step 6: Sampling Procedure
The selection of the sample for the study is carried out using an appropriate sampling technique where most can be obtain and learnt. Moreover, the sample must as well be accessible for the study.
Step 7: Instrumentation Development
Identify instruments to use to collect the data for the study. To select validated instruments that has been used for similar studies with high reliability.
Step 8: Data Collection
Establish contact with the selected population first, and then distribute the questionnaire to them directly.
Step 9: Data Analysis
Digest the collected questionnaires, code them and use SPSS to analyze the results and verify the hypothesis, and then have a discussion.
Step 10: Conclusions and Recommendations
To give a brief conclusion on the main findings of this study as well as the contributions it is posed to make. Finally, to put forward suggestions and recommendations based on the research results.
32 (Commencement)
(End)
Figure 3.2. Research Procedure Review of Literature
Construction of Research Questions and Hypothesis
Research Framework Development
Research Design
Target Population
Sampling Procedure
Instrumentation Development
Data Collection
Data Analysis
Conclusion and Recommendation
33 Review of Literature
A thorough review of the current literature on turnover was conducted with special emphasis on organizational culture, compensation, and job satisfaction and employee turnover. These helped to shape and also influence the formulation of the research hypothesis and enabled the researchers to gain a more better and accurate understanding of the direction of studies in the area as well as what has already been established.
Construction of Research Questions and Hypothesis
Due to the interest and purpose of this scholastic work and the subsequent review of the literature, four research questions were crafted plus four hypotheses for the study in order to strategically exploit the research agenda. The research questions and the hypotheses served as the fulcrum for the study while all factors contributing to the phenomenon been studied were diagnosed.
Development of Framework for the Study
Based on the related review of literature on the research topic and the researcher’s knowledge of the study environment, the framework after consultation with professors was developed. It captured pertinent and seemly unsubstantiated situation responsible for employee turnover in the industry.
Research Design
A correlational research design was chosen to address the crafted research questions using a survey research model. According to Babbie (2011), surveys may be used for descriptive analysis and are useful in studies that have individuals as units of analysis. He also pointed that,
“Survey research in general offers advantages in terms of economy and the amount of data that can be collected. The standardization of the data collected represents another special strength of survey research”.
Therefore, a questionnaire comprising of 58 items were used to collect the required data from the selected participants. A well-structured questionnaire captures statements as well as questions.
The former provides the researcher with ease in designing items while the later can be close-ended wherein the respondents have to choose from a list of possible answers provided. The
34
structure of the provided possible answers could be a source of weakness of the closed-ended questions as those answers might not adequately represent the respondents viewpoint, thus overlooking important responses the respondents might have provided (Babbie, 2011).
Notwithstanding, the uniformity of the responses in closed-ended questions and statements makes it easier for the researcher to process the information gathered. The correlational design was chosen because it is most effective in allowing the researcher to assess the relationship of the variables in question. Moreover, it is useful in identifying problems and drawing sound judgments.
The questionnaire was also self-administered due to its advantages instead of running interviews. The logic behind this approach was triggered by the fact that accessing this particular population was difficult due to the hectic nature of their job. In additional, the risk of bias during interviews would be easily avoided besides the level of comfort and confidentiality the approach guarantees. The self-administered questionnaire guarantees anonymity and would thus serve its purpose factoring the level of risk employees in these environments are at times faced with when they become outspoken or tend to pinpoint work related issues affecting them. In additional, self-administered questionnaires are appropriate in quietly stimulating frank response from respondents, particularly due to the privacy and also the sensitivity of some of the items in the study. The survey was structured in such a way that each question demands to be answered with a ranked response in the Likert scale format. The responses obtained were then coded into numbers and entered on an excel sheet before it was finally exported to the SPSS Windows version 19.0 for statistical analysis.
Target Population
The study was conducted amongst employees of the Gambian service industry and employees of three well established financial services constituted the participants in this study. Moreover, the accessibility to this group posed no major problem as they are largely concentrated in the urban centers, the Greater Banjul Area.
35
Sampling Procedure
Babbie (2011) stated that a sample is a special subset of a population observed in order to make inferences about the total population itself. Since I was unable to meet the entire employees in the service industry, the sample for this study was taken from the financial service industry which is large and vibrant in the country. Besides it is also one of the top players of the service industry. Participants were selected through purposive sampling techniques as a result of knowledge of the study group and environment. Babbie (2011) again, postulated that “at times it is appropriate to select the sample based on the researcher’s knowledge of the population, its elements, environments, and the nature of the research intentions”. Moreover, since the study chose a non-probability sampling approach, the researcher Kidder (1981) stressed its viability in situations wherein the researchers had limited resources in time, finances or both and therefore could not conducted the study with a more extensive population.
The sample size for the study was 152 participants drawn from the various levels of confirmed employees in the Gambia financial service industry. A total number of 230 questionnaires were dispatched to participants in their various workplaces. Out of this number 157 completed questionnaires were collected. However, 5 out of the number were declared invalid and therefore were discarded as a result of either missing data or multiple checking on the same questions. As a result 152 questionnaires out of the total administered and collected were used for this study, representing 66% of the total questionnaires administered.
Eligibility Criteria
For participating in the study, individuals must have worked for their current organizations for a period of a least a year. This gave some level of credibility to the study as it implied that participants for the study were confirmed employees in their various organizations and therefore familiar with a variety of issues in their workplaces. A confirmed staff or employee is a staff who is fully employed and not serving a probation period. Generally it takes six months to be confirmed as a staff in the industry but in some rare instances it could last close to a year.
Demographic Data
Demographic data were collected on participants’ gender, age, marital status, educational level, and years of employment in present organization, as well as position or level in the organization.
36
Despite not the main objective of the study, these items helped provide information on the role these demographic items played with regards to turnover and turnover intention.
Instrumentation Development
The instruments used for this research project are the Organizational Culture Questionnaire (OCQ), measuring organizational culture, the Minnesota Satisfaction Questionnaire (MSQ) measuring job satisfaction, the Pay Satisfaction Questionnaire (PSQ) measuring employee compensation and the Michigan Organizational Assessment Questionnaire (MOAQ) measuring turnover intention. These four instruments were developed by previous researchers on the various areas. A combination of these scales plus the 6 demographic questions gave a total of 58 items arrange in an orderly manner. A description of each of these research instruments follows below:
Organizational Cultural Questionnaire
The Organizational Culture Questionnaire (OCQ) is a survey instrument developed by the Holistic Management Limited in 1999 and was subsequently improved in 2000. The instrument contains 15 items providing five options; almost always, often, neutral, sometimes and almost never. With regards to its reliability and validity which has been captured under four areas namely involvement, consistency, adaptability and mission - it has been proven to be reliable and valid for usage in studies bordering organizational culture. In terms of involvement, consistency, adaptability and mission the instrument yielded the following high Cronbach coefficient alpha 0.90, 0.88, 0.87 and 0.92 respectively.
Minnesota Satisfaction Questionnaire
The Minnesota Satisfaction Questionnaire (MSQ) was developed by Weiss, Dawis, England and Lofquist, (1967) to measure overall job satisfaction. The instrument has both a long and a short version with a five scale option grading from very satisfied to very dissatisfied. The MSQ short version was chosen for this study in order to increase participation of service industry employees who are generally reluctant to spent time responding to surveys or completing questionnaires due to the hectic nature of their jobs. The instrument consisted of 20 items from the long version. The long version was developed to quantify an employee’s overall satisfaction
37
with regards to their job. The data on its internal consistency and reliability was assessed using the Hoyt’s analysis of variance method. Its median reliability coefficients were reported to be 0.86 for intrinsic satisfaction, 0.80 for extrinsic satisfaction and 0.90 for overall general satisfaction. The short version MSQ is based on a subset of the long version items and therefore validity as well as reliability can be inferred from the long MSQ version. Several scholars and academics made use of the MSQ short version in their studies involving job satisfaction, yielding sound results (Hirschfeld, 2000; Struempfer & Mlonzi, 2001).
Pay Satisfaction Questionnaire
The instrument, Pay Satisfaction Questionnaire (PSQ) was development by Heneman and Schwab (1985). It is a multidimensional construct comprising of 18 items developed to assess four dimensions of pay namely pay level, benefits, pay raises and structure/administration. This instrument has received wide recommendation for its usage on studies relating to employee turnover. Regarding its validity and reliability, the instrument has undergone countless scrutiny and verification by scholars and academics. In fact, according to Judge (1992), it is the only instrument available to measure dimensions of pay satisfaction effectively and efficiently. The instrument’s coefficient alpha reliability estimate for the overall scale is 0.89.
Turnover Intention
The 2-item turnover intention scale had been adapted from the Michigan Organizational Assessment Questionnaire (MOAQ), an instrument developed by Cammann, Fichman, Jenkins and Klesh, (1979) and has been used to measure turnover intention. As a result of the little modification of the instrument to suit the purpose of the study, the items were pretested using a small number of graduate Gambian students in Taiwan. This led to some minor adjustment of the instrument besides the scrutiny it underwent from seasoned university professors. Since the original instrument had 3 items and the Cronbach alpha reported as 0.87, its reliability and validity could be inferred.
38 Table 3.1.
Items of Instrumentation
Variables Organizational culture
Job satisfaction Compensation Turnover intention
Questions 1 - 13 14 - 33 34 - 50 51 - 52
Data Collection
The questionnaires were taken to the various institutions selected for the study and hand delivered to the selected employees to complete and was subsequently collected after a couple of days. Due to the nature of work of this study group and their level of understanding of the English language the questionnaire was self-administered. This allowed anonymity and helped to reduce interviewer’s influence and thus provided confidentiality to the participants as there were no identifiers on the instrument. Babbie (2011) stated that “survey research is probably the best method available to social scientists interested in collecting original data for describing a population too large to observe directly”. Individual participation in this study was entirely voluntary.
Data Analysis
After the survey exercise was concluded, the questionnaires were coded using a 3 - digit serial number. This was subsequently followed by exporting the responses into an excel sheet using the guideline provided in Table 3.2 below before finally moving them into the Statistical Package for Social Sciences (SPSS) version 19.0 for windows.
39 Table 3.2.
Data Coding
Column No. for coding Item/Section Scale
1 -3 Serial Number Nominal Data
4 – 16 * Part 1: Organizational culture Q1 - Q13 Likert scale - continuous 5-1 17 – 36 **Part 2 : Job satisfaction Q14 - Q33 Likert scale - continuous 5-1 37 – 53 **Part 3: Compensation Q34 - Q50 Likert scale - continuous 5-1 54 – 55 Part 4: Turnover intention Q51- Q52 Likert scale - continuous 3-1 56 - 61 Personal data
53 Gender 1= Male; 2 = Female
54 Age 1= 29 & below; 2 = 30 & above 55 Marital Status: 1= Single; 2 = Married
56 Level of Education:1= College certificate; 2= College degree
57 Length of Service:1= 1 year & below; 2 = > 1 year to 3 years; 3 = > 3 years to 5 years; 4 = > 5years
58 Position: 1= Junior staff; 2 = Management role
*Note: 5 = Almost always; 4 = often; 3= Neutral; 2 = Sometimes; 1= Almost Never
**Also: 5 = Very satisfied; 4 = Satisfied; 3 = Neutral; 2 = Dissatisfied; 1= Very dissatisfied
Analysis of the data were done using SPSS version 19 for Windows. The analysis included descriptive statistics (means, standard deviation, percentages, and frequencies) for the demographic variables of the sample as well as one-way analysis of variance (ANOVA). Pearson correlation coefficient was used to examine the relationship between variables in the study as well as independent sample test on the variables. Linear, multiple and hierarchical regressions were used to test for the influence of the variables and the role of the mediating variable respectively. Both descriptive and inferential statistics were used during the course of reporting