3.1 Introduction
This chapter describes the research design and methodology that was employed in this study. This is done under the following headings: research design, locale of the study, sample selection and sample size, research instruments, piloting, data collection techniques, and data analysis. The research methods used in this study included library research and a field survey. Library research involved review of documents such as educational policy documents, government reports, research findings and relevant publications on education. The field survey was conducted as described in other sections of this chapter.
3.2 Research design
According Robson(2002) and Chandran (2004) research design refers to an arrangement of conditions for collection and analysis of data in a systematic relationship with the purpose of the research such that research questions are turned into a project. A research design has the following components: purpose(s), theory, research questions, methods and sampling strategy. Design is therefore a plan or strategy for conducting the research.
This study adopted the exploratory type of research design although aspects of observation and descriptive design were also incorporated. Robson (2002) argues that exploratory design is suitable in seeking new insights, asking questions, assessing phenomena in a new light and generating ideas and hypotheses for future research. According to Chandran (2004), exploratory design is suitable in studies that seek to understand people, events and situations. The design provides new insights and discovery of new ideas to the researcher. Although the design has limitations such as lack of formulation and testing of hypothesis and it is only investigative in nature, the researcher adopted it because it has certain critical advantages. The advantages of the design include its strength in leading to formulation of research hypothesis for further research and its stimulation of interest and encouragement of seeking to understand and gain new insights instead of testing research-related statements. The design further promotes in-depth seeking for answers and explanations of events and situations as they take place without looking for causal links. It encourages drawing together various pieces of information and increases investigative power of the researcher. This study therefore used exploratory design because of the forgoing reasons. It focused on questions that clarify contemporary challenges in teaching and learning processes. This endeavoured to shed light on how PTE can be made more responsive to the challenges of the 21st century. Three techniques were used for data collection namely questionnaires, interviews and focus group discussions.
3.3 Locale of the study
This study was based in public teacher training colleges randomly selected in each province of the Republic of Kenya. The republic of Kenya is in East Africa on both sides of the equator between latitudes 4ï‚° N and 4ï‚° S, and longitudes 34ï‚° E and 41ï‚° E. To the north it is bordered by Ethiopia, to the northwest by the Republic of Southern Sudan, to the west by Uganda, to the south by Tanzania and to the East by Somalia. At the time the data collection was carried out, Kenya was divided into eight administrative provinces namely: Central, Coast, Eastern, Nairobi, North Eastern, Nyanza, Rift Valley, and Western. Each province was further divided into districts, divisions, locations, and sub-locations.
Public teacher training colleges in Kenya were spread all over the eight provinces except Nairobi province. Nyanza had two colleges, Eastern province had four, Western province had two while Rift valley province had four. One college was located in Coast province, four were in central province while one college was located in North eastern province. There was no public TTC in Nairobi province. Other institutions involved in this study included the primary school and the divisional educational office closest to the selected teacher training college in each of the selected provinces. KIE which is located in the city of Nairobi was also involved in this study. It is located along Desai road, off Murang'a road.
3.4 Target population
The target population for this study was 18 public primary teacher training colleges (PTTCs). According to Chandran (2004), population in research refers to human or non-human items under investigation. The key population involved in this study were the PTTCs under whose docket is placed the task of training primary school teachers in the country. These institutions are expected to prepare teachers who implement primary school curriculum in the country. The PTTCs in this study were represented by heads of departments (HODs), lecturers and teacher trainees who constituted the targeted population. The population of heads of department was 144 while that of lecturers was 827 and the population of teacher trainees was 15789. These figures were based on January 2009 establishment and enrolment in the 18 public teacher-training colleges in Kenya which were distributed in the country as shown in table 3.1.
Table 3.1 Primary teacher training colleges in Kenya
Province
Number of PTTCs per Province
North Eastern
1
Central
4
Nairobi
0
Coast
1
Rift Valley
4
Western
2
Eastern
4
Nyanza
2
Total
18
Although the core population for this study was the teacher training colleges, a few other stakeholders in PTE were included to enrich the findings. These were quality assurance and standards officers (QASOs), curriculum developers at KIE and practising primary school teachers. There were six specialists who were responsible for curriculum development of PTE at K.I.E. The entire population of these specialists was involved in the study as the number was small. The population of QASOs was 612 while that of practising teachers was 196,000.
3.5 Sample selection and sample size
The field research was carried out in seven teacher-training colleges, which were sampled from the 18 colleges available in Kenya. The sample was limited to one college per province to ensure that views from all the provinces were incorporated. One college was randomly sampled from all the PTTCs available in each province. Private colleges were left out on the basis that there are many mushrooming colleges and their standards cannot be ascertained. The researcher aimed at collecting views and opinions regarding teacher education in relation to responsiveness to contemporary challenges. The research targeted 258 respondents as follows:
Table 3.2 Number of respondents
Target respondents/subjects
Number of
respondents
PTTC heads of departments (8 per PTTC, 1 PTTC in each of the 7 provinces)*
56
PTTC lecturers(8 lecturers per PTTC, 1 PTTC in each of the 7 provinces)
56
Practising teachers(5 per school, 1 school in each province)
35
Quality assurance and standards officers (3 per PTTC locality, )
21
PTE curriculum developers at KIE (all the 6 PTE specialists)
6
PTTC lesson observations (2 per PTTC)
14
Teacher trainees (a group of 10 per PTTC)
70
Total number of respondents
258
*Data was collected in 7 provinces, 1 PTTC per province. Nairobi province was not represented because there was no public PTTC. There were 8 HODs representing each of the 8 departments. The list of the PTTCs and primary schools sampled for this study is shown in appendix IX and X respectively.
Since this study focused on primary teacher education, teachers training colleges played the central role in data collection. The seven colleges selected from different provinces constituted 38.9% of all the public primary teacher training colleges in Kenya. According to Mugenda and Mugenda (1999) a sample of 10% of the target population is representative enough. This sample was therefore considered reasonable. In order to collect information from other significant stake holders, quality assurance and standards officers, curriculum developers and practising teachers were also included as respondents. The various respondents were selected as follows:
3.5.1 PTTC heads of departments.
Each teacher training college had eight departments namely languages, sciences, creative arts, professional studies, computer, social sciences, mathematics and physical health. Eight heads of departments for each of the 7 PTTCs selected were involved in the study. They were selected because they were responsible for training of primary school teachers in the various subject areas. It was expected that they were experts in their areas of specialisation and they were well versed with contemporary issues relevant to primary teacher education. In addition they were policy and curriculum implementers in their respective colleges.
3.5.2 PTTC lecturers
Eight PTTC lecturers in every college were selected using stratified random sampling method. A list of all lecturers in every subject area was acquired from the relevant heads of department. The first name on the list was sampled. If for whatever reason the sampled lecturer was either not available or was the head of department, the next lecturer on the list was sampled. This sample was important because PTTC lecturers are directly responsible for the delivery of the PTE curriculum and the training of the primary school teachers.
3.5.3 Teacher trainees
10 teacher trainees were sampled for the focus group in each PTTC. These were volunteers from the final year class. Volunteers were preferred because they were likely to be better motivated to discuss freely in contrast to randomly selected members of focus group. According to Robson (2002), a focus group should ideally comprise of eight to twelve members. This study chose a mean of the two figures.
3.5.4 Quality assurance and standards officers (QASOs)
The study targeted three QASOs from the divisional education office closest to the selected teacher training college in the province. Snowball sampling, which uses one sample to lead to the next, was used. The researcher identified the PTTC, which was then used as the basis of identifying the divisional education office from which QASOs were randomly sampled. A total of 21 such officers, three from each province, were therefore sampled. Quality assurance and standards officers are responsible for ensuring that the quality and relevance of primary education is maintained in schools within their jurisdiction. They were involved in this study because they are considered knowledgeable in determining the extent to which primary education is responsive to contemporary issues. This is essential because implementation of successful primary education is a direct outcome of primary teacher education. In addition they are familiar with the challenges encountered in the teaching and learning process in primary schools in Kenya.
3.5.5 Practising teachers
Snowball sampling was again used to select practising primary school teachers. The PTTC selected formed the basis of selecting the primary school which would be involved in the study. The researcher identified the primary school nearest to the selected PTTC. Purposive sampling was used to select five teachers with the most years of teaching experience. One such teacher was selected from each level in upper primary section; that is from class four to class eight. The choice of teachers with most years of experience was based on the following assumptions: first, they were more familiar with the issues that have confronted primary education over the years. Secondly, they were more likely to provide strategies they may have used to respond to challenges they faced over the years. Finally, after years of teaching, they may have reflected on effectiveness of teacher preparation with regard to responsiveness to contemporary issues. Consequently this study sought to obtain feedback from them on all these issues.
3.5.6 PTE curriculum developers at KIE.
There were six curriculum specialists at K.I.E who were responsible for curriculum development of primary teacher education. The entire population of six staff formed the sample for this study. This sample was regarded as important because it had a special role in determining how PTE was implemented in Kenya. They were responsible for determining the content of PTE curriculum as well as appropriate pedagogical approaches to deliver the curriculum. KIE is also mandated to monitor emerging issues and incorporate them in the curriculum.
3.6 Research instruments
This study utilized four data collection instruments namely questionnaire, interview schedule, focus group discussion and observation checklist. The objectives of the study as well as the research questions were used to guide the formulation of items in each of the instruments.
3.6.1 Questionnaire
This instrument was used to collect data from PTTC heads of departments, lecturers, QASOs and practising primary school teachers (see appendix II, III and IV). The instruments were considered suitable because the respondents are literate and conversant with educational issues addressed in this study. In addition the instrument was appropriate because the questions could be answered at the respondents' convenience given that they are busy due to their nature of work. In order to address all the areas under investigation in this study, the questionnaires had five sections based on the objectives of the study as well as research questions.
3.6.2 Interview schedule:
This tool was used to collect data from PTE curriculum developers at KIE (see Appendix V). The interview schedule enabled the researcher to get detailed responses from the respondents. It also considered appropriate because it enabled the researcher to probe further for clarifications and explanations where necessary. The interview questions were designed in such a way that explored the various areas covered by the research questions. An interview typically involves the researcher asking questions and getting answers from the respondent. According to Robson (2002), interviews are widely used in social research and the common types include structured, semi-structured and unstructured interviews. Selection of the type depends on the "depth' of response sought. Structured formats have fixed questions in pre-decided order and standardised wording where responses are selected from a list of alternatives. Semi-structured interviews use much more flexibility in terms of order of questions and responses. Unstructured "in-depth" interview was adopted in this study as it allowed the respondents to say whatever they liked on the broad topic thus giving insights on diverse pertinent issues.
3.6.3 Observation checklist:
The researcher observed PTTC lessons in order to examine the instructional methods as well as teaching and learning materials used by PTTC lecturers. Other areas of observation included teacher-student interaction, linkage of content to contemporary issues and involvement of pupils in the learning process. The observation checklist is found in appendix VII. Observation was used as a supportive or supplementary method to collect data that compliment or set in perspective data obtained using the other instruments.
Observation method entails watching what is happening, recording this, describing, analysing and interpreting what has been observed. Robson (2002) depicts two extremes of observation namely participant observation and structured observation. The former is essentially qualitative style while the latter is quantitative style. He proposes a third approach, unobtrusive observation which is non-participatory, non-reactive and is mainly unstructured and informal. This study chose observation for its directness which allowed the researcher to look at and listen to the respondents instead of asking them for their personal views, feelings or attitudes. Observation also avoids discrepancies between what people claim to do and what they actually do. Data from observation was used to compare, contrast, and supplement information from other techniques. It was also used to set in perspective data obtained by other means. The major disadvantage of observation approach is that it is time consuming and may also introduce some artificiality due to the presence and influence of the observer. Unobtrusive observation was used as it is the most appropriate approach to exploratory research (Robson, 2002).
3.6.4 Focus groups
According to Robson (2002) focus group refers to a group interview on a specific topic. It is an open ended group discussion guided by the researcher. He argues that focus groups are easy to carry out and have the advantages of being efficient in generating substantial amounts of data. The method is relatively inexpensive and flexible and participants who may be reluctant to be interviewed on their own or even fill a questionnaire are likely to participate. This study used focus groups to collect data from teacher trainees (see appendix VI). This method was considered appropriate because of its potential to generate detailed and relevant responses to pertinent questions in the study. It also had the advantage of providing data more quickly and at lower cost than interviewing individuals separately. It was particularly useful here because groups of teacher trainees could be assembled together on short notice.
Bogdan & Biklin (1998) depict focus groups as a good method to provide information on how groups of people think or feel about a particular topic and give greater insight into why certain opinions are held. The groups are ideal for planning and design of new programs as well as provide a means of evaluating existing programs. The researcher can interact directly with respondents and they allow clarification, follow-up questions as well as probing. The researcher acquires more information by observing non-verbal responses and this complements or contradicts verbal responses. Data gathered from focus groups use respondents' own words and the method is flexible and can be utilized with large variety of topics and settings. Finally the results are easy to understand than complex statistical analysis of survey data.
There are however some disadvantages of using focus groups. The researcher has reduced control over the group of respondents and has less ability to regulate the information that will be produced. The groups produce relatively cluttered data making its analysis more difficult. The outcome may also be influenced by certain dominant individuals who dominate the conversations while shy members may be hesitant to participate. Focus groups cannot give valid information about individuals or tell how things have changed over time. According to Krueger (1988), "focus group session should include around five or six questions" and should always include ten people or less.
3.7 Piloting of research instruments
The instruments were improved with the guidance of the university supervisors. The instruments were pre-tested before the data collection exercise in order to determine and enhance their validity and reliability. This was done in two teacher-training colleges, which were not part of the study sample. The findings of the pilot study were used to refine the data collection tools and procedures.
3.7.1 Reliability
According to Wiersma (1980), content reliability refers to the degree to which a particular measuring procedure gives equivalent results over a number of repeated trials. The test-retest or coefficient stability method was used to determine the degree to which the same results received from the questionnaires could be obtained with repeated measure of accuracy in order to determine the reliability of the instrument. First the developed questionnaires were given to twenty subjects identical to the ones sampled for the study. The answered questionnaires were coded and analysed manually. Secondly, the same questionnaires were administered to the same group after a period of two weeks and analysed as the previous ones. Thirdly a comparison of the two sets of results was made using Spearman's product moment formula. The correlation coefficient was computed to show the magnitude of the relationship between the two results. The calculated Spearman's correlation coefficient was 0.74. The relationship of the two results was deemed to be sufficiently high because the magnitude of the coefficient was closer to one. According to Wiersma (1980), a correlation coefficient of 0.5 or more indicates sufficient reliability of the instrument.
3.7.2 Validity
According to Wiersma (1980) and Annabel (1992) content validity refers to the extent to which the contents of an instrument measure what they are supposed to measure. In this study triangulation was used to determine the validity of the items in the data collection instruments. Triangulation refers to using diverse methods and processes of collecting and analysing data to enhance credibility and rigour of research (Robson, 2002). There are four aspects of triangulation, namely: use of more than one method of data collection such as observation, interview and documents (data triangulation), use of more than one observer in the study (observer triangulation) combining qualitative and quantitative approaches (methodological triangulation) and using multiple theories or perspectives (theory triangulation). This study used all these forms of triangulation in order to enhance content validity of the instruments.
3.8 Data collection
The data collection was preceded by familiarization visit to each of the sampled teacher training colleges, schools, KIE and divisional education offices. During these visits, relevant authorities were requested to allow the researcher to carry out research in their institutions. The purpose of the study and the significance of the data collection exercise were carefully explained to the target samples. Consent of the respondents was also sought. Sampled respondents were notified and arrangements were mutually made for the data collection exercise. The researcher and his two assistants administered the research instruments.
3.8.1 Questionnaires
A cover letter briefing the respondents about the purpose of the research accompanied each questionnaire. The researcher discussed with each respondent and set an agreed time frame for collection of the completed questionnaires. After the agreed period of time elapsed, the researcher made follow up by reminding the respondents who had not returned the completed questionnaire. This was done through telephone conversation, or actual visit depending on the location of the respondents. A few questionnaires were distributed through e-mail. Respondents who had good access to the internet and who preferred this option were requested to fill the online version of the questionnaire and return to the researcher through his e-mail account. This was an advantageous option to the researcher as it enabled the transfer of the data straight to the data management software (SPSS). This saved a lot of time which would otherwise have been used to receive the questionnaire, code and key in the data.
3.8.2 Interviews
The researcher carried out the interviews in person. During the interview the researcher made brief notes with the permission of the respondents. Interviews can be classified in two categories namely personal and phone. This study mainly used personal interview as it allowed the researcher to collect data directly and personally from the respondents. Chandran (2004), argues that this approach has the advantage of being able to probe and allow for detailed descriptions and comprehensiveness as needed. Telephone interview is direct though not face to face. This approach was used sparingly and was restricted to people who were unavailable for personal interview. It had the advantage of efficiency in terms of time and reduced cost of travel. It however has limitation of cultural barriers and many people prefer face to face interaction. Respondents may also not want to answer questions over the telephone especially to a total stranger. The respondent may also not have sufficient time to think and give a comprehensive answer over the telephone. Probing by the interviewer is also not practical as it increases both time and the cost of the interview. Due to the foregoing reasons, this study adopted the unstructured, in-depth and face to face interview to facilitate the creation of rapport, comprehensiveness of answering questions and allow probing.
Where allowed, transcription of the interview was done but only included complete thoughts and useful information. Clarification for non-standard grammar or slang was sought and the appropriate meaning recorded. The researcher transcribed interviews immediately so as to resolve any ambiguity while the memory was still fresh. He then reviewed the notes and interview transcripts to refine, modify or add questions based on evolving topics. When important realizations were noted during interviews, they were written down immediately. After the interviews, the researcher read over the interview notes and wrote a summary of themes.
3.8.3 Observations
The researcher first met with the sampled PTTC lecturers whose lessons were to be observed in order to explain the need for the observation. An assurance was given that the observation was to be carried out professionally and confidentiality of the results maintained. The observation was both formal and informal in approach. The formal aspect included checking the occurrence or absence of listed items in the checklist while the informal aspect was less structured and allowed considerable freedom in what information was gathered and how it was recorded. It included note-taking of all observations that were found to be significant to the study.
3.8.4 Focus Group Discussions (FGDs)
The researcher was the moderator of the focus groups and directed the discussion and also made notes. His role was to keep discussions flowing and on track, guiding the discussions back from irrelevant topics and making transitions into another question. The assistant took a separate set of notes, operated the tape recorder whenever permission was granted and responded to any unexpected interruptions. All effort was made to ensure that notes were so complete that it could be used even if tape recording did not take place. The discussion was usually in this order: welcome, then overview of topic, ground rules and then first question. The overview was meant to provide an honest clarification about the purpose of the study and the importance of the topic of group discussion. Ground rules helped to guide the discussion and included guidelines such as: minimizing or eliminating side conversations, one person speaking at a time, not criticizing what others had to say, and treating other people's ideas with respect. The first question was in each case carefully selected in order to "break the ice" and encourage everyone to talk. The researcher made use of the "pause and probe" system in which he paused after a participant talked - to give other participants a chance to jump in. Probes were used to request for additional information. A suitable location was identified in which participants sat facing each other. The conclusion of the focus group was made and involved thanking the group for participating. The researcher in each case summarized what was said and asked if anything was missed out.
3.9 Data analysis
The data collected in this study was both qualitative and quantitative. The responses obtained were classified into the following categories:
Views regarding contemporary challenges
Indicators of responsiveness to contemporary challenges
Obstacles that hinder PTE from being responsive to contemporary challenges
Recommendations regarding enhancing responsiveness of PTE to contemporary challenges.
The data collected was recorded manually on data sheets. The number of observations and variables were found to be small hence the observation data was analysed on a computer using excel worksheets. The rest of the data went through three stages, namely Coding (transfer of data into coded sheets), typing (entering the data into computer) and editing (checking the data by comparing two independently typed data). The typing of data from paper questionnaires was done twice. The second time was done by a different person whose job specifically was to identify any possible mismatches between the original and second entries.
The notes made by the researcher during the interviews formed the data to be analysed. The notes were based on responses made in answer to specific questions by the researcher. Analysing interview data began by coding the notes into meaningful categories, which enabled the researcher to organize the large amounts of text and determine underlying patterns. Bogdan and Biklin (1998) suggest "first ordering interview transcripts and other information chronologically or by some other criteria". The researcher carefully conducted initial coding by generating several category codes as he read responses and labelling data that were related. He wrote notes, listing ideas and diagramming relationships that were noticed. Lastly, focused coding was used to remove, merge, or sub-divide coding categories and look for repeating ideas (same idea expressed by different respondents) and bigger themes that link codes. After coding categories were developed, a list that assigned each code an abbreviation and description was made. Matrices, concept maps, flow charts, and diagrams were, where applicable used to illustrate relationships and themes. These visual aids were used to enhance validation of themes or deliberation of other associations or explanations.
The researcher used the observation checklist to note the presence or absence of the listed items or behaviour. The observation was largely structured but there was flexibility in terms of taking note of any other observation that was found to be important to the study. For structured observation, a coding scheme was prepared and contained predetermined categories for recording what was observed. The scheme mainly noted whether certain items or behaviour were present or not. The researcher used tallies in the checklist which provided frequency data, both in absolute terms (how many times each item or behaviour was observed) and relative terms (the relative frequency of different items and behaviours). This data was then analysed and collated using percentages, frequency tables and charts. Unstructured items were processed by first giving codes to the initial set of materials obtained from the observation and then adding the researcher's comments or reflections. The researcher then went through the materials trying to identify similar patterns, themes, sequences or relationships and used them in focusing the next observation. A set of generalizations were then developed which covered the consistent ideas detected in the observation data.
The record made from focus group discussions formed the data which was categorized, coded and analysed. Analysis and reporting of data was descriptive and presented the meaning of data in contrast to a summary. According to (Krueger, 1988), "data can be examined and reported at three levels, namely, raw data, descriptive statements and interpretation". Raw data was a presentation of statements the way they were spoken by respondents. Descriptive statements summarized respondents' comments and provided illustrative examples using the raw data. Interpretation was used to build on the descriptive process by providing meaning of the data going beyond the summarizing of the data.
According to Chandran (2004), there are four main measures used to analyse statistical data. These include measures of central tendency, measures of dispersion, measures of correlation and measures of association. This study used measures of central tendency (arithmetic mean) and measures of association (chi square) to analyse statistical data. Measures of central tendency were used on univariate distribution and results were displayed in charts or frequency tables. Measures of association were used on bivariate distribution and included cross tabulations. Chi-square test was used to find out whether the views of the various categories of respondents were significantly different from each other. Where the number of frequency was less than five, the Fisher's exact test was used instead. The rationale for the choice of the two tests is that Chi-square test is inappropriate where expected values are below five in any of the cells of a contingency table. In contrast the Fisher's test can be used irrespective of the sample characteristics. It is difficult to calculate with large samples, but these are specifically the conditions where the chi-square test is most appropriate. The two tests were therefore used depending on the values in contingency tables.
Qualitative analysis was used to deal with words in contrast to quantitative analysis which deals with numbers. The field notes, focus group and interview summaries constituted the qualitative data and led to comments, interpretations and emerging hypotheses. Berkowitz (1996) identifies two basic forms of qualitative analysis, namely "intra-case analysis and cross-case analysis. A case in this study was a single respondent, a focus group sitting, or a teacher training college. Intra-case analysis was used to examine data from single PTTC, while cross-case analysis compared and contrasted several of them.
Miles and Huberman (1994) identify three major phases of qualitative data analysis: "data reduction, data display, and conclusion drawing and verification". The first phase of the analysis involved processing the mass of data to become organized and meaningfully condensed. Data reduction encompassed selecting and simplifying the data from field notes or transcriptions. It involved deciding which aspects of the data were to be emphasized or ignored depending on their usefulness in answering research questions. Data display provides "an organized, compressed assembly of information that permits conclusion drawing..." (Miles and Huberman's, 1994). Data displays, both in word and diagrammatic form, were used to extrapolate the data in order to discern patterns and interrelationships.
Finally conclusion drawing and verification entailed considering what the analysed data meant with respect to the research questions. Verification involved revisiting the data several times to cross-check and verify emergent conclusions. According to Miles and Huberman (1994), "The meanings emerging from the data have to be tested for their plausibility, their sturdiness, their 'conformability' - that is, their validity". Validity in this context had to do with confirming that the conclusions being made from the data were credible, sound and reasonable. A safeguard for validity such as using numerous sources and modes of evidence were built into the design of this study with the objective of coming up with credible results.
