Características de los estilos de Aprendizaje

McMillan and Schumacher (2006:315) postulate that qualitative research is inquiry in which researchers collect data in face-to-face situations by interacting with selected persons in their settings. These authors claim that qualitative research describes and analyses people’s individual and collective social actions, beliefs, thoughts and perceptions. In addition, they assert that in qualitative research the researcher interprets phenomena in terms of meanings that people assign to them. In this study qualitative research was used to “inductively explore phenomena and to provide ‘thick’ [that is, detailed] descriptions of phenomena” (Durrheim 1999:43).

Qualitative research designs emphasise gathering data on naturally occurring phenomena. McMillan and Schumacher (2006:26) classify qualitative research designs as interactive and non-interactive. Interactive methods consist of ethnography, phenomenology, case studies, grounded theory, and critical studies. Non-interactive methods comprise of concept analysis, and historical analysis.

For this investigation a phenomenological study was utilised. This approach aims to understand and interpret the meaning that participants give to their everyday lives (Fouché 2005:270). Johnson and Christensen (2008:48) maintain that the key element of a phenomenological study is that the researcher attempts to understand how people experience a phenomenon from their own perspectives. A phenomenological study describes the meanings of a lived experience (McMillan & Schumacher 2006:26) and the way in which the researcher puts the experiences together to develop a worldview (Marshall & Rossman 1995:82).

This qualitative component of the research entailed the researcher adopting dual roles. The first role was that of a participant observer in which the researcher created the role for the purpose of the study. The second role was that of the interviewer in which the researcher established a role with each person involved.

155

observer will possibly utilise the ethnographic approach; Strydom (2005:276) asserts that the phenomenological approach is important in participant observation, as the researcher endeavours to gain an in-depth insight into the manifestations of reality. I preferred the phenomenological approach since my research aimed to understand and interpret the meaning that participants give to their everyday lives and not an intact cultural or social group.

In ethnographic studies the interactions are quite widespread but the researcher is less intrusive in data collection. Whereas, in phenomenological studies the interaction is more intrusive, close and personal. In this research the researcher preferred interacting with the learners and educators, and observing and interviewing them. Thus, the phenomenological approach was more appropriate.

Johnson and Christensen (2008:212) affirm that qualitative observation involves observing all relevant phenomena and taking extensive field notes without specifying in advance exactly what is to be observed. Observation is mainly done by means of naturalistic methods of study, analysing conversations and interactions that researchers have with participants. Fouché (2005:270) asserts that researchers using this strategy of interpretive inquiry will mainly utilise participant observation and long interviews as methods of data collection.

Participant observation and interviews were used as the main strategies of inquiry in this study. The interview is the most prominent data collection tool in qualitative research (Punch 2009:144). It is an effective way of accessing people’s perceptions, meanings, definitions of situations, and constructions of reality. However, during the research process I created the research strategy best suited to my research, or even designed my whole research project around the strategy selected.

Strydom (2005:274-276) describes participant observation as a research procedure that is typical of the qualitative paradigm and that studies the natural and everyday set-up in a particular community or situation. McMillan and Schumacher (2006:347) state that participant observation enables the researcher to obtain people’s perceptions of events and processes expressed in their actions and expressed as feelings, thoughts and beliefs. Participant observation is a combination of particular data collecting strategies: limited participation, field observation, interviewing, and

156

strategies were used to corroborate data. Observation, limited participation and interviewing were utilised.

4.3.2.1 Sampling

Sampling involves decisions about which people, settings, events, behaviours and/or social processes to observe (Durrheim 1999:44). Although Mouton (2005:101) argues that bias samples are common errors in data sources and access to data, this is mostly related to heterogeneous populations. Most qualitative researchers prefer to select a purposive sample (Gorman & Clayton 2008:128). Hence, the sampling technique that was employed in this study was purposeful sampling [sometimes referred to as purposive, judgement or judgemental sampling].

Purposeful sampling is a type on non-probability sampling (Strydom & Delport 2005:328). According to McMillan and Schumacher (2006:125) non-probability sampling is the most common type of sampling in educational research. In non-probability sampling, the researcher uses subjects or participants who happen to be accessible or who may represent certain types of characteristics. This does not include any type of random selection from the population. In purposeful sampling, the researcher selects particular elements from the population that will be representative or informative about the topic of interest (McMillan & Schumacher 2006:126). In this case the elements that were selected were “information rich” cases. This was based on the researcher’s knowledge of the population.

The researcher used the same five schools as in the quantitative section since this is part of the same project. The sample was drawn from the five English medium primary schools. Both male and female foundation phase educators were selected. However, this did not ensure maximum variation; since it is noticed that majority of the foundation phase educators in the five schools are females. In selecting the educator sample, the researcher also took into consideration the educators’ post levels. Thus, foundation phase HODs and foundation phase level one educators were selected.

Paton (2002) (in Strydom & Delport 2005:328) claims that there are no rules for sample size in qualitative inquiry. Qualitative inquirers view sampling processes as dynamic, ad hoc, and

157

2006:321). While there are statistical rules for probability sampling size, there are only guidelines for purposeful sample size. Dörnyei (2011:127) reinforces that a well-designed qualitative study usually requires a relatively small number of participants to yield the saturated and rich data that is needed to understand even subtle meanings in the phenomenon under focus. Thus, Dörnyei (2011:127) draws from his experience in stating that an initial sample size of six to ten might work well and using computer-aided data analysis we can increase the sample size to as many as thirty; although that would be barely manageable for a single researcher.

Contrarily, McMillan and Schumacher (2006:321) argue that purposeful samples can range from one to forty or more. The logic of sample size is related to the purpose, the collection strategy and the availability of information-rich cases. In qualitative research sampling occurs subsequent to establishing the circumstances of the study clearly. Thus, sampling is undertaken after the investigation has commenced (Strydom & Delport 2005:328). Furthermore, McMillan and Schumacher (2006:322) emphasise that the insights generated from qualitative inquiry depend more on the information richness of the cases and the analytical capabilities of the researcher than on sample size.

Against this background, sampling was conducted simultaneously as data was collected and continued until data saturation was reached.

4.3.2.2 The research technique

Data collection was conducted simultaneously when the quantitative component of the study was done. Participant observation was one of the two main data collecting strategies in the qualitative component (Refer to Appendix 14 for observational grid). The primary interest was observing groups of learners and individual learners during reading and writing lessons and the physical settings in which reading and writing took place. Data collection entailed writing accurate and systematic observational notes during observation and as soon as the session had ended. Loose notes or jottings were converted into field notes at the end of every day. Field notes contained a comprehensive account of the participants themselves, the actual discussions and communication, and the observer’s attitude, perceptions and feelings. Individual interviews were a corroborative data collection strategy. The form of interaction or role taken by the researcher

158

on the role of the observer much more than the role of the participant (Johnson & Christensen 2008:214). Classroom observation continued until data saturation was reached.

Since a phenomenological study was used, “the typical technique was for the researcher to conduct long interviews with the informants directed towards understanding their perspectives on their everyday experiences with the phenomenon” (McMillan & Schumacher 2006:26). By conducting individual interviews, the researcher attempted to understand the world from the participant’s point of view and unfolded the meaning of the educators’ experiences. I preferred individual interviews to focus groups so that each educator could, in private, express his or her feelings qualitatively.

Johnson and Christensen (2008:207) emphasise that qualitative interviews can be used to obtain in-depth information about a participant’s thoughts, beliefs, knowledge, reasoning, motivations, and feelings about a topic. The semi-structured one-to-one interview was used. Researchers use semi-structured interviews in order to gain a detailed picture of a participant’s beliefs about, or perceptions or accounts of, a particular topic (Greeff 2005:296). With the semi-structured interviews, the researcher will have a set of predetermined questions on an interview schedule, while still having the freedom to digress and probe for more information (Mackey & Gass 2008:173). Greeff (2005:296) points out that an interview schedule or guide is a questionnaire written to guide interviews. Although there were a set of prepared guiding questions [interview guide or schedule] and prompts, the format was open-ended and the interviewees were encouraged to elaborate on issues raised in an exploratory manner. Hence, the interview schedule merely assisted the researcher in directing the interview instead of being restricted to the predetermined questions.

The interview schedule or guide was constructed from the literature on the topic. The questions were carefully formulated, sequenced and based on the research questions, aims and objectives of the study. The interviews were conducted at the schools.

4.3.2.3 The pilot study

A pilot study was conducted before the main investigation commenced. In qualitative research the pilot study is usually informal, and a few respondents possessing the same characteristics as

159

(Strydom & Delport 2005:331). The pilot study in qualitative research allowed the researcher to test certain questions. By testing the nature of the questions in an interview schedule in the pilot study, the qualitative researcher was able to make modifications with a view to quality interviewing during the main investigation (Refer to Appendix 11).

4.3.2.4 Analysis and interpretation of data

McMillan and Schumacher (2006:364) emphasise that there is no set of standard procedures for qualitative data analysis or for keeping track of analytical strategies. However, these authors claim that qualitative data analysis is primarily an inductive process of organising data into categories and identifying patterns among the categories.

In this study the researcher utilised an eclectic approach to qualitative data analysis. Firstly, predetermined categories were used from the following sources:

 The research question and sub-questions.

 The research instrument [the interview guide].

 Concepts and themes from the literature.

 Prior knowledge of the researcher or personal experience as an educator and a researcher.

For data analysis, this study used the topics embedded in the interview schedule. The pre- established categories provided a frame for the next phase of data analysis.

Secondly, each category was divided into subcategories as the data were analysed. The subcategories were provisionally applied and thereafter refined.

Thirdly, each participant’s typed comments were cut according to the above categories, manually coded and stapled.

Fourthly, relationships among the categories were made by discovering patterns in the data. In this study the researcher used triangulation, which is the cross-validation among data sources and data collection strategies (McMillan & Schumacher 2006:374). To find regularities in the data,

160

observation, and the survey {quantitative}] to see whether the same patterns kept recurring. Fifthly, in seeking for patterns in the data, the researcher also searched for discrepant and negative evidence that modified or refuted a pattern. Finally, the developed patterns and themes were used to report the experiences of the participants.

4.3.2.5 Validity and reliability

McKay (2006:11) emphasises two qualities that are essential for sound research. They are validity and reliability. She stresses that these terms are defined differently in qualitative research because there is far less control and structure in qualitative research.

 Validity

Validity in qualitative research is the degree to which the interpretations have mutual meanings between the participants and the researcher (McMillan & Schumacher 2006:324). Thus, the researcher and participants agree on the description or composition of events, and especially the meanings of these events.

In qualitative research, internal validity refers to credibility and external validity refers to transferability (McKay 2006:13; De Vos 2005:346). Lincoln and Guba (1985) (in Marshall and Rossman 1995:143; Dörnyei 2011:57) propose four alternative constructs that more accurately reflect the assumptions of the qualitative paradigm. They are credibility, transferability, dependability and confirmability.

De Vos (2005:346) states that the goal of credibility is to demonstrate that the inquiry was conducted in such a manner that the participant was accurately identified and described. In this study, credibility, or truth value of the study, was enhanced by continuing data collection over a long period of time to ensure that the participants had become used to the researcher and were behaving naturally.

Credibility or internal validity was achieved by carefully recording the interviews, transcribing the interviews verbatim, analysing all the data gathered, and presenting the data in a fair and unbiased manner. Strategies that were used to enhance internal validity

161

participant reality], verbatim accounts [literal statements of participants were obtained], multi-method strategies [data was collected using in-depth interviews and observation], mechanical recordings of data [interviews were recorded using a voice recorder], and participant review [educators were asked to review the researcher’s synthesis of interviews for accuracy].

Transferability or internal validity refers to the degree to which the findings of the qualitative study can be applied to other contexts (McKay 2006:13). De Vos (2005:346) affirms that the transferability of a qualitative study to other settings may be problematic. However the degree of transferability depends to a large extent on the similarity of the learning contexts being examined.

In order for the readers to determine the transferability of this study, the researcher provided a complete description of the participants and the context of the research so that readers can determine to what extent the findings might be applicable to other contexts. This implies that transferability, or applicability of the results to other contexts, was achieved by reporting “thick descriptions”. “Thick descriptions” refer to the process of using multiple perspectives to explain the insights gleaned from the study (Mackey & Gass 2008:180). Interpretative validity refers to the degree to which the participants’ viewpoints, thoughts, feelings, intentions, and experiences are accurately understood by the qualitative researcher (Johnson & Christensen 2008:277). Participant feedback was used in order to achieve interpretative validity. Participant feedback refers to discussion of the researcher’s conclusions with the participants (Johnson & Christensen 2008:277). When writing the research report, interpretive validity was achieved by using many low-inference descriptors. These are helpful because the reader can experience the participant’s actual language, dialect and personal meanings. A verbatim is the lowest inference descriptor of all because the participant’s exact words are provided in direct quotations (Johnson & Christensen 2008:278). Direct quotations were used when writing the research report. This ensured interpretative validity.

162

The reliability of qualitative research depends on what is termed dependability (McKay 2006:14). McKay (2006:14) defines dependability as the degree to which the results reported in the study can be trusted or are reliable. Dependability, or consistency of the findings, was achieved since the interviews were recorded. Electronically recorded data helped the researcher to make use of all interpretive cues in order to draw inferences and evaluate the dependability of the inferences that had been drawn (Mackey & Gass 2008:81). Additionally, in reporting the qualitative study, the researcher provided a rich description of the participants involved in the study, the context for the study, and, most importantly, all the steps that were taken to carry out the study.

Confirmability, or neutrality of the findings, was achieved since full details of the data on which claims or interpretations are based were made available. This will allow other researchers to be able to examine the data and confirm, modify, or reject my interpretations. 4.4 ETHICAL CONSIDERATIONS

Empirical research in education inevitably carries ethical issues, because it involves collecting data from people, and about people (Punch 2009:49). Generally ethics involves beliefs about what is right or wrong, proper or improper, and good or bad (McMillan & Schumacher 2006:142). Ethics is a set of moral principles which is suggested by an individual or group, is subsequently widely accepted, which offers rules and behavioural expectations about the most correct conduct towards experimental subjects and respondents, employers, sponsors, other researchers, assistants, and students (Strydom 2005:57).

Qualitative research is more likely to be personally intrusive than quantitative research. This is because in qualitative research, more specifically interactive qualitative methods, the researcher uses face-to-face techniques to collect data from people in their natural settings. Thus, ethical guidelines include policies regarding informed consent, deception, confidentiality, anonymity, privacy, and caring (McMillan & Schumacher 2006:333).

163

Obtaining informed consent implies that all possible or adequate information on the goal of the investigation, the procedures which will be followed during the investigation, the possible advantages, disadvantages and dangers to which respondents may be exposed, as well as the credibility of the researcher, be rendered to potential subjects or their legal representatives (Williams, Tutty & Grinnel (1995) as cited in Strydom 2005:59). Strydom (2005:59) stresses the voluntary nature of participation in research. He also emphasises that participant must be made aware that they would be at liberty to withdraw from the investigation at any time. Rosnow and Rosenthal (1999:59) describe autonomy as “independence”. They state that in the context of research ethics, autonomy refers specifically to a prospective subject’s right as well as ability “to choose” whether to participate in the study, to continue in the study, or to opt out of the research process at any time when the participants needs to.

For this study, the researcher obtained permission from the KwaZulu-Natal Department of Education [In order to maintain anonymity, the letter received from the KZN Department of Education is not attached – Refer to Appendix 2], the principals [In order to maintain anonymity, the letters received from the school principals are not included – Refer to Appendix 3], participating educators [Refer to Appendices 8, 10 and 13], and the parents of isiZulu-speaking learners from the five schools [Refer to Appendices 4, 6, and 12], and the school governing bodies. Durrheim and Wassenaar (1999:66) argue that obtaining consent from participants is not merely the signing of a consent form. This requires that participants receive a full, non-technical and clear explanation of the tasks expected of them so that they can make an informed choice to participate voluntarily in the research. Hence, in gaining permission, the researcher gave participants the assurances of confidentiality and anonymity and described the purpose of the study.

Both questionnaires [in English and isiZulu] consisted of covering letters in English and isiZulu. The covering letters also assured the subjects of confidentiality and anonymity and explained the purpose of the study. All the educators who were interviewed were asked to sign forms that