• Tidak ada hasil yang ditemukan

Chapter 1 Introduction

3.4 Research process

3.4.3 Qualitative phase

81

Internal consistency (Cronbach’s Coefficient Alpha) was computed on items in the questionnaire as a reliability estimate to ensure that all items grouped together on an instrument were measuring the same construct consistently. If an instrument had high internal consistency, then if in-service teachers strongly agreed on one item, it was expected that they would also strongly agree on other items measuring the same construct.

Internal consistencies (like Cronbach’s Alpha) of 0.7 or higher are considered adequate (Barclay et al., 1995). For this pilot test the Cronbach’s Coefficient Alpha obtained was 0.87, indicating good internal consistency.

3.4.2.4 Validity and reliability 3.4.2.4.1 Validity

To assess the content validity of the questionnaires, a panel of local experts familiar with ICT in teacher education programs has been asked to review the questionnaires and provide feedback on content relevance and clarity.

3.4.2.4.2 Reliability

The UTAUT and TPACK sections of the finalised questionnaire were independently tested with Cronbach’s Coefficient Alpha. The UTAUT section, comprising 24 items, achieved a reliability score of 0.902, which was consistent with prior studies (Avci & Askar, 2012;

Baltaci-Goktalay & Ozdilek, 2010). The TPACK section, comprising 18 items achieved a reliability score of 0.875 and was consistent with other TPACK studies (Chai, Koh, & Tsai, 2011; Schmidt et al., 2009). The UTAUT and TPACK constructs have been tested separately for reliability before conducting EFA and the results are shown in Chapter Five.

82 3.4.3.1 Participant selection

In a sequential explanatory mixed-methods design, individuals participating in the qualitative phase should also have participated in the initial, quantitative phase (Creswell

& Plano Clark, 2007). Because the qualitative phase uses the quantitative results, only individuals that contributed to the quantitative phase are suitable to take part in the qualitative follow-up. In this study, participants were asked to indicate on the survey questionnaire if they would be interested in participating in a semi-structured interview on use of Web 2.0 in teaching and learning. The participants were informed that the interviews would last between 20 and 30 minutes, and that if selected, they would be contacted via email.

3.4.3.2 Target population

Purposeful sampling was used to select participants for the semi-structured interviews. The logic and power of purposeful sampling lies in selecting information-rich cases for in-depth study (Patton, 1990). Information-rich cases are those from which one can learn a great deal about issues of central importance to the purpose of the research (Patton, 1990). Due to the nature of the sequential design of this study, the selection of the participants for the second, qualitative phase depended on the results from the first, quantitative phase. More details on the final sample is provided in Chapter 5 Section 5.2.

3.4.3.3 Semi-structured interviews

In a semi-structured interview, the researcher–interviewer uses pre-prepared questions that are used to guide the interview process. However, the researcher–interviewer is free to explore responses on a deeper level and in addition has the opportunity to add questions when he or she deems it to be necessary – for example if he or she finds a question or series of questions not thought about before but seeming to be relevant to ask at a specific point in time during the interview.

The semi-structured interview promotes flexibility, as it allows one to move beyond the initial pre-determined questions and as a result helps one to capture personal experiences outside the realm of the pre-determined ones (Creswell, 2013), and as a result contributes

83

to being a very productive data gathering tool The purpose of using in-depth semi- structured interviews with in-service teachers was to understand their attitudes and beliefs regarding use of Web 2.0 tools in their teaching strategies. The interview protocol included 10 to 15 open-ended questions and was pilot tested. The content of the protocol questions was based on the results of the statistical tests obtained from the quantitative phase. The interviews were tape-recorded, and word processed. Respondents were given the opportunity to review and, if necessary, modify the contents of their word-processed interview.

3.4.3.4 Qualitative data analysis

The researcher transcribed the data from the interviews. The qualitative data was examined using content analysis, to “identify, code and categorise the primary patterns in the data” (Patton, 1990, p. 381). For this phase of the analysis, inductive analysis was used, meaning that “the patterns, themes and categories of analysis come from the data”

(Patton, 1990, p. 390). The steps in qualitative analysis included: (1) preliminary assessment of the data by reading through the transcripts and writing memos; (2) coding the data by segmenting and clssifying the text; (3) using codes to build up themes by bringing together similar codes; (4) linking and interrelating themes; and (5) building a narrative (Creswell, 2013). A visual data display was created to show the developing conceptual framework of the factors and relationships in the data (Miles & Huberman, 1994).

3.4.3.5 Establishing credibility

The criteria for passing judgement on a qualitative research are different from quantitative study. Within qualitative research, the researchers must scrutinise themselves and the participants to deal with issues relating to reliability and validity (Creswell, 2013). In the 1980s, Guba and Lincoln replaced the terms reliability and validity with the notion of trustworthiness. Fundamentally, trustworthiness relates to what exrtent a study accomplishes what it is intended to do (Merriam, 1998). In qualitative research, the researcher looks for believability, based on consistency, insight, and trustworthiness (Lincoln & Guba, 1985) through a process of verification instead of traditional validity and

84

reliability measures. To validate the findings, determine the credibility of the information and whether it matches reality (Merriam, 1988), four procedures were used in the qualitative phase of the study: (1) triangulation by converging diverse sources of information; (2) participant checking by receiving the feedback from the participants on the accuracy of the identified categories and themes; (3) provision of rich, thick description to express the findings; and (4) external audit requesting a person not involved in the study to review the qualitative study and report back (Creswell, 2003).