Chapter Overview
This chapter gives details of the research procedure, research method and gives details of the design of the model or research framework. The framework shows the paths through which the hypotheses are tested. The framework was developed after thorough literature review on what has already been researched concerning the variables under study.
It therefore gives way to answer the research questions. The research procedure outlines the sequence of events that led to the existence of this paper. Under research methods issues of data collection in terms of population and sapling are discussed. Also discussed is the instrumentation method under which validity and reliability of the questionnaires are explained. Data analysis explains the kind of research the researcher used (quantitative), the coding style, the package used in analysing the data as well as the method of analysis the data. The chapter also explains the questionnaire distribution.
Research Procedure
A website was developed for collecting data. The questionnaire was uploaded in a format that allowed respondents to get the instructions on how to answer the questions and what the numbers on the answers meant. This website had an inbuilt database in which all the answers were being stored. No e-mail address was required to send the answers. By simply clicking the submit button, all the responses went straight to the database. The website was developed in a way that by simply clicking at the link, the whole questionnaire would appear. Only the IP address from which the individuals responded could be stored in the database. This was one way of protecting the respondents’ identity. In submitting the questionnaire link to the free sites of the university and to others, an accompanying letter was included to introduce the researcher and the aim of the research.
The questionnaire was submitted to the free forum website of the University of Malawi where people could easily and freely answer the questionnaire. They could also send the link to other individuals as a snow ball questionnaire. This consideration was made
40
because the middle range civil servants being targeted are either graduates of the University of Malawi, Mzuzu University or other universities. Over 90% of them are direct graduates of the University of Malawi and Mzuzu University. The rest are either graduates from other Universities outside Malawi or reached the Grade I level through promotions (Lienert, 2005).
The other consideration was that this study was not targeting every civil servant. The only target were the ones who are on Grade I and Grade H levels which are the entry level positions and one step from the entry level respectively. To this end random sampling would not be used because the target population would not be reached. Hence snow-ball was` a better solution as the right population would be reached so easily.
The questionnaire was mainly adopted from two main sources: the Minnesota Satisfaction Questionnaire (Weiss et al., 1967) and Job Diagnostic Survey (Hackman &
Oldham, 1974, 1975). There were a few modifications to the questions just to make sure that they suit the current study. The modifications were mainly done to make them fit the time lag. Otherwise there were not many modifications to the questionnaires. The selection of the questions took into consideration the underlying theories upon which the model was developed-the 2-factor theories of motivation.
The data were collected for a period of two months. A due date was set on which the researcher would stop collecting the data. After the one month of collecting the data, the researcher downloaded the available data to test the reliability. This acted as the pilot test although no changes were made to the questionnaire.
Research Framework
The model is in 3 stages, called models for easy reference. In the first phase (Model 1) H1 is about Job Satisfaction and H2 is about Employee Commitment measured against Turnover Intentions. The second phase (Model 2) is about the HR practices: Salary strategies, H3; Job Enrichment/Autonomy strategies, H4; and, Job Stability/Security strategies, H5 measured against Job Satisfaction and then Employee Commitment. In this model each hypothesis has two parts (a) and (b). The (a) part is tested against Job
41
Satisfaction while the (b) part is tested against Employee Commitment. In Model 3, H6, H7 and H8 for Salary, Job Enrichment and Job Stability are tested against Turnover Intentions in that order. It is worthy stressing at this stage that the model was arrived through literature review. This also includes the work that was done on a similar kind of study using the same variables.
Figure 3. 1. Research Framework Job Satisfaction
Employee Commitment
Turnover Intentions H2
H1
Salary Job Enrichment
Job
Stability H: 3, 4,5b
H: 3, 4,5a
H: 6, 7, 8 Job Satisfaction
Turnover Intentions
Employee Commitment
42
Research Methods Measures and Data Analysis
Survey items for the constructs derive from commonly accepted theoretical definitions. Thus they are partially influenced by the efforts of other researchers who have already used similar constructs. The research paper uses the Likert-type with five-item scale on all constructs (1 = definitely not to 5 = definitely yes as follows):
1= Very Dissatisfied/ Strongly Disagree (much less than what you expected) 2= Dissatisfied/Disagree(less than what you expected)
3=Neutral (cannot make your mind: neither satisfied nor dissatisfied) 4=Satisfied/Agree (as you expected)
5=Very Satisfied/Strongly Agree (more than you expected)
A consideration to the research was to keep the questionnaire short. The total number of the questionnaire was therefore 24 items. There were 6 demographic items and 18 questions. The respondents were advised to answer all the questions and use their own assessment of their experience in the civil service since these are issues of employee attitudes and perceptions.
The Model was tested using backward multiple regressions using Statistical Package for Social Sciences (SPSS) version 17.0. Backward elimination was chosen because it eliminates the values that are not significant or have the least weight and only keeps the ones that are strong in the last equation.
Population and sample
The total population of the civil servants working on the two positions of Grade I and Grade H is not really clear. The researcher did not manage to get the actual figures due to filing problems. It is estimated that the total number of civil servants is about 191,000 (Malawi Dairy Times News, 2008) which includes all grades and all people with different
43
qualifications. It is estimated that the population of these grades contributes between 3% and 4% since the civil service is highly skewed at the bottom with clerical and unskilled workers.
The other problems are that the system is being updated from manual to electronic yet there are still some ghost workers, more deaths, more exits without notice and other logistical problems. However the total number of respondents was 116. Since there were no clear numbers of the total population (N), it was necessary to get as much data as possible.
Since the questionnaire was being administered electronically and it was snowball, it was difficult to know the number of people who got the link to the questionnaire. However after one month since the distribution of the questionnaire, some data were retrieved from the database to test the re liability. The results listed in Table 3.1 as follows:
Table 3. 1. Data variables by entries and values (n=68)
Variables Entries Percentage
Gender Male 42 61.8
Female 26 38.2
Age range 20-29 46 67.6
30-39 19 27.9
40+ 3 4.4
Marital status Single 47 69.1
Married 20 29.4
Widow/er 1 1.5
Job Grade I 48 70.6
H 17 25.0
Work Experience 1-2 Years 20 29.4
3-4Years 29 42.6
5-6Years 10 14.7
7Years+ 8 11.8
Highest Education Qualification Diploma 9 13.2
College Graduate 36 52.9
Graduate Studies 23 33.8
44 Instrumentation
The instrument used for data collection was a questionnaire. The questionnaire had 2 parts due the language technicalities. The first part was to be answered with denotations of satisfied/dissatisfied while the other part had be answered in an agree/disagree form. This was done in order to make sure that there is subject-verb agreement.
The questionnaire was derived from several questionnaires that have been used before. Such include the (Minnesota Satisfaction Questionnaire [MSQ], 1967) the long form version which was modified in (1977). According to (Weiss, Dawis, England & Lofquist, 1977), this short version can be scored on three dimensions: intrinsic satisfaction, extrinsic satisfaction and general satisfaction. It is also measured on the Likert type scale with five items each. This paper however did not adopt the full version of the MSQ because some of the questions are not applicable to this study. Hence other questions have been adopted from the Valencia survey previously mentioned in this paper. The survey was called A model of high performance work practices (Luna-Arocas & Camps, 2008) which used the Minnesota Satisfaction Questionnaire; the Job Diagnostic Survey (Hackman & Oldham, 1974, 1975);
The Organisational Commitment Questionnaire (OCQ) by (Mowday et al., 1979) and (Hom
& Griffeth, 1991).Thus job satisfaction was measured using the MSQ whereas job enrichment and job stability were measured using the Job Diagnostic Survey. Employee commitment was measured using the Organisational Commitment Questionnaire whereas turnover intentions were measured using (Hom & Griffeth, 1991).
Questionnaire distribution
The questionnaire has been divided into several categories based on the y-variables.
The first seven questions measure Job Satisfaction as follows:
• Questions 1 and 2 measure satisfaction based on salary. The questions were adapted from the Minnesota Satisfaction Questionnaire (Weiss et al., 1967). However question 2 was modified a little bit by removing words that relate to company and inserting words that relate to government institutions in order to suit the situation.
45
• Questions 3, 4 and 5 measure satisfaction based on job enrichment. All the questions were adapted from the Minnesota Satisfaction Questionnaire (Weiss et al., 1967).
• Questions 6 and 7 measure satisfaction based on job stability. Both questions were adapted from the Minnesota Satisfaction Questionnaire (Weiss et al., 1967).
The second set of questions (8 - 14) measure Employee Commitment as follows:
• Questions 8 and 9 measure employee commitment based on salary. Job Diagnostic Survey (Hackman & Oldham, 1974, 1975).
• Questions 10, 11 and 12 measure employee commitment based on job enrichment (autonomy). All questions were adapted from the Minnesota Satisfaction Questionnaire (Weiss et al., 1967) with slight modification to question 10.
• Questions 13 and 14 measure employee commitment based on job stability. Adopted from Job Diagnostic Survey (Hackman & Oldham, 1974 & 1975). Once used by Roberto Luna-Arocas and Joaquin Camps in “A model of high performance work practices and turnover intentions” (January, 2007).
The last set of questions (15-18) measure Turnover Intentions as follows:
• Questions 15 and 16 measure turnover intentions based on job satisfaction. The questions were adapted from the Minnesota Satisfaction Questionnaire (Weiss et al., 1967) with a few modifications in terms of the time lag.
• Questions 17 and 18 measure turnover intentions based on employee commitment.
Question 17 is self-composed modelled after MSQ while question 18 is fully adopted from the Minnesota Satisfaction Questionnaire (Weiss et al., 1967).
46 Distribution of questions and their scales
Salary scale: In my present job, this is how I feel about:
“The amount of pay for the work I do.”
“My salary compared with other employees in similar position in other government institutions other than the civil service.”
“My salary compares well with that of other workers on the same position.”
(3 items: Q 1, 2, 9)
Job enrichment scale: “The challenging nature of my job (tasks & responsibilities).”
“The involvement and treatment by colleagues and supervisor that I get.”
“I can make autonomous decisions and suggestions at work and have them implemented.”
“My job requires me to do many things at work, using a variety of skills and talents (intellectual and non-intellectual ideas).”
(4 items: Q 3, 4, 10, 12)
Job stability scale: “The way my job provides for steady employment.”
“The way my job provides for a secure future.”
“If I do a good job the department/ministry will recognise my effort and my job is safe.
“Compared with my employer, other employers fire workers more readily in times of difficult.
47 (4 items: Q 6, 7, 13, 14)
Employee commitment scale: “I have no intention to leave my employer but I may do so if an opportunity opens.
“My employer offers better salary and other privileges so I cannot leave my job.”
(2 items: Q 8, 16)
Job satisfaction scale: “The overall joy I get from my job as a whole.”
“I like my job better than the average worker”
(2 items: Q 5, 11
Turnover intentions: “I intend to remain in my profession but to leave my current employer for a better one within these 2 years.”
“I am considering taking another job the soonest I find it.”
“I have no intention to leave my current employer at all.”
(3 items: Q 15, 17, 18) Distribution of questions in terms of model 1 to 3 X1: Salary
1. The amount of pay for the work I do.
2. My salary compared with other employees in similar position in other government institutions other than the civil service.
3. My salary compares well with that of other workers on the same position X2: Job Enrichment
48
1. The challenging nature of my job (tasks & responsibilities).
2. The involvement and treatment by colleagues and supervisor that I get.
3. I can make autonomous decisions and suggestions at work and have them implemented.
4. My job requires me to do many things at work, using a variety of skills and talents (intellectual and non-intellectual ideas).
X3: Job Stability
1. The way my job provides for steady employment.
2. The way my job provides for a secure future.
3. If I do a good job the department/ministry will recognise my effort and my job is safe.
4. Compared with my employer, other employers fire workers more readily in times of difficult.
Y1: Job Satisfaction
1. The overall joy I get from my job as a whole.
2. I like my job better than the average worker.
Y2: Employee Commitment
1. My employer offers better salary and other privileges so I cannot leave my job.
2. I have no intention to leave my employer but I may do so if an opportunity opens.
Y3: Turnover
1. I intend to remain in my profession but to leave my current employer for a better one within these 2 years.
2. I am considering taking another job the soonest I find it.
3. I have no intention to leave my current employer at all.
Validity and Reliability
49
Validity of the questionnaire was determined through the content validity. Content validity is actually a way of trying to determine whether the instrument really addresses the issues at hand. It therefore looks at the relationship between the content in the instrument (questionnaire) and what the researcher is trying to find out. This validation can be done by experts or friends who would look at the questionnaire thoroughly. If they do not understand or find other problems with it, then the researcher modifies the questionnaire. The questionnaire for this study was validated in two ways: expert review and peer review. It was submitted to a few scholars who looked at it. Also some peers went through it and gave feedback and suggestions on how to improve the questionnaire. Changes were made to that effect. However there were a few changes because the questionnaire was adapted from already tried and tested instrument.
Reliability, the measure for internal consistency of the instrument was also obtained for both the first set and the second set of the data. The first set, n=68 acted as a pilot study.
It obtained a reliability of Cronbach Alpha of .735 for all the 18 items. The researcher did not test individual questions. For the final data with n=116 the Cronbach Alpha was .703 with the 18 items. A possible explanation in the drop of the Alpha could be because there were a lot of missing data in the final collection. The means of the results were also obtained but will be reported in chapter four in order to avoid duplication.
Research Processes
The researcher went through a number of processes to come up with this paper. The figure below shows the whole process that the researcher went through.
50
Design of Final Report Conclusions and Suggestions
Analysis of Data
Administration of the Questionnaire
Development of Website for Administering the Questionnaire Development of the Questionnaire
Development of Framework of the Study Establishing of Research Questions and Hypotheses
Discussion of Literature Identifying Research Subject
Figure 3.2. Research Process
51