This chapter outlines the methods and procedures used to collect data, as well as how the collected data was analysed and also the limitations of the study.
2.1 Research design
A research design functions as glue that holds the research study together, structuring the study and fitting together major research components such as sampling and measures in answering the research questions. In research, qualitative design and the quantitative design form the major types of research designs. Qualitative research design gives a complete description of particular phenomena for which the researcher seeks an explanation. This design is based on the viewpoint that all research is ultimately linked to qualitative research. Here, the researcher is highly involved in the process, and the data collected takes the form of words and objects (Creswell, 2003). Quantitative research aims at explaining everything in terms of 0 and I. It classifies, counts and creates statistical models to give explanations about particular occurrences. Data collection tools such as questionnaires are used to collect numerical data. This research design results in precise measurements of objects under investigation (Creswell, 2003).
This study used the quantitative research design which enabled the researcher to collect data from a large sample of schools in order to enable the findings to be generalised. In studying leadership behaviour of secondary school Rectors, it was important to use a large sample since there are many secondary schools in Mauritius. The data collected on the leadership styles and teachers' participation in decision making needed to be statistically correlated to determine the impact of leadership on collaborative decision making. Thus, it was crucial to use the quantitative research design to yield data that can be manipulated statistically.
2.2 Population and sample
In research, a population refers to subjects whose attributes are of interest to the study. It consists of the elements that the study would like to investigate (Creswell, 2003). The secondary school population in the Republic of Mauritius is divided into five zones, according to the Ministry of Education and Human Resources. The schools are further categorised as State, Private Aided and Private Non-Aided. The total number of secondary schools in Mauritius is 182. For the purpose of this research, the schools in Zone 5 (Rodrigues) were not included, as well as the Private Non-Aided schools. For this study, the population was all secondary school teachers working in State and Private Aided schools of Mauritius. Under this population teachers were the respondents. This included all practicing teachers and excluded all trained but not practicing teachers.
The sample is a small proportion or subset of the population. The importance of samples in research is the impracticability of obtaining study data from all the elements of the population (Creswell, 2003). This present study used a sample which is 15 percent of the whole school population. The school population in this study is 156 and thus there were 23 schools that were sampled. This sample was divided among the zones on the basis of the number of schools in the zone.
This study used a stratified random sample to get to the respondent. In this sampling technique, the sample of schools is first divided into clusters or groups. Here the clusters were the four zones, whereby each zone constituted a cluster. The sample for each zone was based on the proportion of the schools in that zone relative to the total number of schools. Thus Zone 1, with 51 schools, had 8 schools making it to the sample; Zone 2, with 41 schools, had 6 schools; Zone 3, with 34 schools, had 5 schools and Zone 4, with 30 schools, had 4 schools. Table 2.1 illustrates the sample of schools chosen.
Table 2.1: Sample of schools per zone
The sample was further divided amongst State and Private-Aided schools, based on the proportion of each in the corresponding zone relative to the total number of schools. The State and Private-Aided schools were then categorised according to pupils' gender, namely Girls, Boys and Mixed. Simple random sampling, using a random number generator, was then used to select the 23 schools forming part of the sample.
Table 2.2: Types of schools in sample
the sample size of 100 teachers has been obtained from the relation where n = N 1+ Ne2, where n= sample size, N = population size, e = margin of error; an error of 10% has been considered here (Pasigpasigan, 2007).
Table 2.3: Sample of teachers
POPULATION OF TEACHERS
2.3 Investigative Techniques
The study used questionnaires for data collection. Questionnaires are quite cost- effective and they can capture a large sample, while offering a means of integrating present findings with former similar studies. They are the main investigation techniques used for large samples where quantitative data is required. The study employed this technique to collect data and this allowed for statistical manipulation and generalisation of the data to the whole population. Questionnaires allow anonymity in response, thereby enhancing data quality and reliability. The questionnaire was structured in order to have standard questions whose responses are predefined. This allowed for easy data analysis, even though new material that the researcher may not be aware of can be omitted.
The questionnaire used for the study incorporated both the Multifactor Leadership Questionnaire (MLQ) and the Teacher Involvement and Participation Scale Questionnaire. The questionnaire is given in Appendix A.
2.4.1 Multifactor Leadership Questionnaire
This is a data collection instrument that measures leadership styles. Respondents score their responses on the Multifactor Leadership Questionnaire. Bass and Avolio (2000) crafted the MLQ from their full range of leadership models. Various tests have confirmed that MLQ is a good predictor of leader performance and behaviour. The MLQ scores leadership behaviours on a scale from transactional to transformational styles of leadership. This instrument was used to ask teachers in Mauritius to give their opinion on the leadership styles used by their Rectors. The adapted MLQ instrument used for this study excluded the measures for leadership outcomes since this would be measured by the T.I.P.S questionnaire. Thus, instead of the usual 45 questions in the MLQ, the study used 21 questions. Out of the 45 items, the questionnaire measures five transformational leadership variables, three transactional variables, one variable of non- transactional leadership and three outcome variables. The latest version of the MLQ questionnaire comes in two forms: the self rater and rater form (other people rate an individual) (Bass and Avolio, 2006). This study used the rater form where the teachers will rate the Rectors in terms of their leadership styles.
Bass (2005) advances that transformational leadership is based on five variables, as measured by the MLQ: Idealised Attributes, Idealised Behaviours, Individualised Consideration, Intellectual Stimulation and Inspirational Motivation. The transactional leadership variables or components are: Contingent Reward, Management by Exception (Passive) and Management by Exception (Active). The outcome components measured are: Satisfaction with the leader, Extra effort by Associates and Individual, Group, and Organisational Effectiveness. Laissez-Faire is the non-transactional component (Bass and Avolio, 2006). The MLQ is based on the 5 point Likert scale of 0 = Never to 4 = Frequently, if not always. The MLQ has been applied extensively in laboratory and field studies of passive / avoidant, transactional and transformational leadership styles (Bass and Avolio, 1994). According to Bass and Avolio (2006), leadership styles are not mutually exclusive since one person can alter his/her leadership styles to suit different circumstances. Appendix B contains a description of all the components in the MLQ.
Bass and Avolio (2000) advance that the MLQ has a reliability of 0.74 to 0.91 for each leadership components that it measures. The tool has been used for hundreds of studies throughout the world, both academic and commercial (Bass and Avolio, 2006). Lowe et al. (1996) support the validity of MLQ through their study of over 3 000 raters. It meets the stringent validity measures in research (Trochim, 2005). Appendix C contains Part B of the questionnaire and their variables within the MLQ framework.
2.4.2 Teacher Involvement and Participation Scale (T.I.P.S)
We started development by reviewing the literature on shared decision making. Through the work of Conley and Bacharach (1990), Sirotnik and Clark (1988), David (1989), and others, we discovered that implementation of shared decision making occurs across eight dimensions:
Goals/Vision/Mission: the degree to which teachers are involved in framing the goals and mission of the school.
Facilitating Procedures and Structures: the degree to which teachers have adequate time, reduced teaching loads, waivers from contracts and regulations, and changed schedules to permit collegial work to occur.
Curriculum/Instruction: the degree to which teachers participate in determining the school program, curriculum goals, textbook selection, educational materials, and classroom pedagogy.
Budgeting: the degree to which teachers participate in matters related to designing and implementing the school budget.
Staffing: the degree to which teachers are involved with the administration in making decisions such as recruiting, interviewing, hiring, and assigning staff.
Staff Development: the degree to which teachers can design and implement staff development activities that meet their own needs.
Operations: the degree to which teachers are involved in managing the building (its use, improvement, and maintenance).
Standards: the degree to which teachers share in setting standards for their own performance and for student performance and discipline.
Items on the TIPS 2 Instrument are rated on a Likert Scale from one to five indicating level of teacher participation in each decision during the past year. (The scale ranges from "Strongly Disagree" to "Strongly agree."). TIPS has demonstrated high reliability and validity for the instrument as a whole and in each of the dimensions.
2.5 Data Collection
The sample used in this study comprises secondary school teachers. The questionnaire was distributed to the respondents during a two-week period and the respondents were given a further three days to fill in the questionnaires. The respondents were issued hard copies of the questionnaires in their schools and, for those who were comfortable to respond immediately, the questionnaires were filled and collected. Otherwise a reminder was sent in two days and arrangements made when the questionnaire would be collected. Email questionnaires were also issued to respondents who preferred to respond via email. These strategies were put in place to ensure that data was obtained in good time. Data was collected from the respective schools at prearranged times.
2.6 Data Analysis
The questionnaires used in the study yielded quantitative data which can be analysed statistically. The data collected was analysed by the use of statistical packages Microsoft Excel and SPSS (Statistical Package for Social Sciences). The analysis was done in the form of descriptive statistics, which enabled the basic description of the sample such as the number and gender of respondents, their length of service and qualifications. The descriptive data statistics included frequency, measures of central tendency (mean and mode), measures of variation and standard deviation. The study sought to investigate the impact of leadership behaviour on collaborative decision making. This was achieved through correlation statistical measures to determine the existence of a relationship.
2.7 Ethical Considerations
There is increasing awareness that the people on whom research is conducted have rights, and that researchers have responsibilities and obligations to these research subjects (Mc Neill and Chapman, 2005). The respondents in the study are human subjects and this required a number of measures to be put in place to ensure that the subjects were protected. The teachers are particularly busy at this time of the year (end of third term) and they do not have sufficient time to devote to surveys. The emphasis was laid on the importance of this study in order to motivate the respondents. The teachers were allowed to keep the questionnaires and respond at their own convenience. The main ethical issues that have been given due consideration during this research are given below.
Informed consent: inform all participants that they are subject to the research project and that they are free to refuse to participate.
â€¢ Deception: true information given to all participants about the purpose of the research.
â€¢ Confidentiality: identifying information about the subjects will not be made available to anyone.
â€¢ Anonymity: the identity of each participant will remain unknown.
â€¢ Risks: no subject will be harmed physically or emotionally.
For this current research, emphasis is laid on three of the above, namely informed consent, confidentiality and anonymity. Since teachers are involved in this research, the prior consent of their school Rectors has to be obtained through a letter detailing the exact nature and purpose of the research. However, many Rectors did not want their schools to be explicitly identified in the survey. The teachers who are selected are given the guarantee that their identity will not be revealed to anyone. Since the study involved a review of Rectors by the teachers, this may have caused the latter to respond favourably or feel unsecured to answer truthfully for fear that they can be identified, despite being given the assurance of complete confidentiality. Finally all those participating in the research have not been named anywhere in the report. Laying emphasis on the above ethical problems is of paramount importance because if the respondents do not trust the researcher, then the validity of the data collected will not reflect the reality.
Biases always find a way into research studies and bring detrimental effects to the study. The most subtle and harmful biases are those that the researcher is not aware of. The researcher and the respondents differ in many professional aspects, mainly in that the majority of respondents have not obtained a postgraduate degree while the researcher is in the process of acquiring one. This may have made the researcher to be insensitive to the needs and views of the respondents. To minimise this bias, the researcher worked closely with the respondents to understand their paradigm and readily offered clarifications on the purpose of the research. The contents of the data collection instruments were also explained to the respondents when necessary.
2.9 Assumptions of the study
The researcher assumed that the characteristics of composition and size of the sample accurately represent the population of the study. It was also assumed that the respondents had interacted long enough in the secondary schools to understand their leader's behaviour. Finally it was assumed that the teachers were also aware of the effects of their Rector's leadership style on their participation in decision making as teachers.
2.9.1 Limitations of the study
The methodology chosen in this study has limitations which may have adversely affected the outcome of the research. The study used the MLQ which is a standard questionnaire for measuring the modern school of thought for leadership behaviours. It has also used the T.I.P.S which is another standardised questionnaire.
Some respondents were not able to answer all the questions as they may not have been aware of or were not concerned enough with the happenings in their schools. This decreased the number of duly filled questionnaires used for data analysis. Also, the
standardised questionnaire did not capture other emergent views or new ideas on the questions that were asked, limiting the contextual richness of the findings.
The period of the school calendar year for this study is definitely unfavourable since all teachers are busy with school examinations. Many respondents may have answered at random due to the limiting time factor.
In research, larger samples are better since they reflect more accurately the characteristics of the population. This study used a small sample to allow the ease of data collection and analysis in order to meet the deadlines. The financial resources required to carry out such an extensive study covering the whole island were beyond the reach of the researcher. Furthermore, the Private non-aided schools were not included in this study since access was not granted in many cases. This also explains the size of the sample used by the researcher.