95
question with ‘an arsenal of methods that have no overlapping weaknesses in addition to their complementary strengths” (p. 248).
4.5.6 Drawbacks of the qualitative approach
Focus groups tend to lead the facilitator in a specific direction that can weaken the trustworthiness of the findings. Discussions can be side-tracked or dominated by a few vocal individuals. The research allowed the participants to talk with each other, ask questions, clear doubts and express opinions. Focus groups generate important information. Yet, such information cannot be attributed to a whole population, but rather concerns a small group of individuals. Therefore, this aspect was handled in the quantitative phase. Another challenge faced by the researcher was to assemble all members of a focus group at a given time. The participants were asked with whom they were more at ease in a discussion. Then the researcher asked the group members in what range of time it would be possible to meet. The researcher had to remind everyone of the meeting 15 minutes before the time scheduled.
Participants may guard their views because of fears of divergence with others.
Morgan (1988) added that the researcher has therefore less control on the collected data. The researcher planned the focus group interviews to overcome and moderate these limitations.
But still as the circumstances of each situation differs from case to case, the researcher cannot forecast precisely how those participate will construe the factors that impact on their choice to engage in an online professional development programme. The researcher is totally unable to make a rigidly structured experiment or know beforehand what to control, alter or leave out (Baskerville & Wood-Harper, 1992).
96
quantitative research design tolerates variation in data treatment, in terms of statistical analyses, comparative analyses and repeatability of data collection so as to ensure reliability.
But this approach is able to offer any explanation outside the descriptive level. Quantitative methods are limited in explaining teachers’ beliefs about online professional development in depth. The quantitative approach is employed to find cause and effect or the link among variables mainly to substantiate hypothesis or theory (Creswell, 2002; Teddlie & Tashakkori, 2012; Feilzer, 2010). A questionnaire was developed from the outcomes of the focus group.
The questionnaire was piloted. The required amendments were made. For collecting data from the respondents a questionnaire was used. The quantitative data results assisted the researcher in the interpretation of qualitative findings such as testing emerging theory, generalising qualitative findings to other samples and validating specific sets of questionnaire items (Creswell, 2003). The questionnaire was administered to a randomly chosen sample.
4.6.1 Research context and sampling
The target population was secondary school teachers in Mauritius. The study sites were two Confessional secondary schools in Mauritius. The schools were chosen for the same reasons mentioned in Section 4.4 (qualitative phase). For this phase of the study the researcher carried out a random sampling procedure to select respondents (teachers) from two secondary schools. In quantitative research, random sampling is frequently used. It has different variations such as stratified random sampling, quota random sampling and systematic random sampling.
The two secondary schools selected were named ‘Maya secondary school’ and ‘Sam Secondary School’ (fictitious names). The student population in one school is of mixed ability, while in the other school the students are high achievers. Data was collected from 75 teachers from two secondary schools, excluding those who took part in phase one. The second phase occurred in both secondary schools using a questionnaire which was developed from results of phase one. The overall number of teachers in both secondary schools was 90 including those (15) who participated in phase one.
Hence, a sample size of 75 respondents was selected from two secondary schools including the school chosen in phase one. Hard copies of the questionnaire were hand
97
delivered to the rectors of the schools by the researcher, and teachers in the two schools were invited to fill in the questionnaire. The surveys were collected within one week. Sixty-five teachers gave back the completed questionnaire with a response rate of 87%. Each questionnaire was numbered before analysis.
4.6.2 Instruments used in data collection (questionnaire)
To answer the four research questions (the first regarding teachers’ beliefs about online professional development, the second the extent to which facilitating and inhibiting factors influence teachers’ beliefs about online professional development, the third how these beliefs are constructed and enabled and fourth why teachers have such beliefs about online professional development), the entire themes and sub-themes were placed into one of the seven categories a) teachers’ beliefs about educational systems; b) teachers’ beliefs about online professional development activities and content; c) facilitating conditions for online professional development; d) inhibiting factors of online teacher training; e) performance expectancy; f) teachers’ experiences with teacher-training courses; and g) social influence.
These were used to develop the survey questionnaire (see Appendix E). Some words or groups of words from the results of the focus group interviews were used to formulate questions for the questionnaire, thus developing a measurement tool for the quantitative phase.
The survey was a cross-sectional one, where data collection occurs at one point in time only. The questionnaire consisted of questions regarding biography, scaled, closed- ended and multiple response type questions. The questionnaire was divided into nine sections. Most importantly, the questionnaire had 10 to 19 items for each of the six factors The sections were: biographical information, perceptions of teacher professional development, beliefs of secondary school teachers of educational system, online professional development activities and content, enabling factors for online professional development, inhibiting factors of online professional development, performance expectancy, teachers’
experiences with teacher-training courses and social influence. The questionnaire, together with the consent form was administered to 75 respondents from the two secondary schools.
To improve reliability of quantitative data the researcher used clear conceptualisation of constructs that defined the constructs to eliminate confusion. The researcher also used
98
precise levels of measurement, that is, a combination of nominal and ordinal levels of measurements. The researcher used a pilot test before conducting the main study.
There are advantages to using a questionnaire. It is more practical. A large quantity of data can be collected from a large sample size in a short period of time. It is also a cost- effective way of collecting data. Anonymity encourages honesty and openness from the respondents. The way of asking questions is more standardised and the presence of the researcher is not necessary. The questions are asked in the same order. Respondents can answer the questionnaire at their own convenience.
The researcher used the Likert Scale as it is the most appropriate scale to measure beliefs. The researcher chose a 5-point Likert Scale as it would be easier for the interviewer to read the complete list of the descriptors (Dawes, 2008). It also increased the response rate and the response quality (Buttle, 1996). Using strongly agreed to strongly disagreed to rate a phenomenon like beliefs in this study reduces the weakness of the quantitative approach which is mainly concerned with statistical data, where most social phenomena, such as health, organisational performance and education, do not obviously produce numerical data (Muijs, 2010).
The sub-themes from the qualitative analysis guided the formulation of the statements in the questionnaire. The sub-themes were grouped under the following headings of the instrument: beliefs of secondary school teachers of educational system (Part 3); online professional development activities and content (Part 4); enabling factors for online professional development (Part 5); inhibiting factors of online professional development (Part 6); performance expectancy (Part 7); teachers’ experiences with teacher-training courses (Part 8) and social influence (Part 9). A matrix table showing the link between the qualitative findings and each of the survey items is provided (see Appendix I).
4.6.3 Validity and reliability
Validity is the extent to which a concept is accurately measured in a study while reliability measures the accuracy of the instrument. After the survey had been developed from the qualitative findings, validity and reliability tests were performed. The questionnaire was sent to four participants of the focus group, to ensure that the statements were in line
99
with what they said in their interview. Groves, et al., (2004) pointed out that all survey should meet these three different standards:
a) “content standards (i.e., whether the questions are asking about the right things);
b) cognitive standards (i.e., whether the respondents understand the questions consistently; whether they have the information required to answer them; and are willing and able to formulate answers to the questions); and
c) usability standards (i.e., whether the respondents and interviewers can complete the questionnaire easily and as they were intended to).” (p. 241).
Five options are available for survey designers to evaluate the draft questionnaire to be able to ensure content, cognitive, and usability standard (Groves, et al., 2004). Reliability of the questionnaire was determined with coefficient alpha (Cronbach, 1984) values. The reliability of items in a questionnaire is acceptable if the alpha is within .70 and .99, but a value of .60 is accepted in social science (Ghazali, 2008). More details about the reliability of the questionnaire are provided in Section 6.2.
4.6.4 Pilot for phase 2
A pilot study was carried out to identify any unforeseen problems with the questionnaire. The latter was refined after the pilot study for its structure, relevancy, validity and reliability. The pretesting sample consisted of five Confessional secondary school teachers. The five respondents were questioned according to the subsequent questions which came from Fink (2003) with regard to the instrument usability: (a) Whether the instructions were obvious for completing the questionnaire; (b) Which questions were not clear, if any;
(c) Did they know in what way to add their responses; (d) Were the response choices mutually limited; (e) Were the response choices complete; (f) whether their privacy were respected and secure; and (g) Did they have any proposals regarding the addition or removal of questions, clearing up of instructions or enhancements in questionnaire layout.
They were also requested to specify the amount of time it took them to complete the questionnaire. The responses of the respondents were as follows: (a) the instructions were clear; (b) there were no confusing questions; (c) the way to indicate responses was understood; (d) the response choices were mutually exclusive; (e) the response choices were
100
complete; (f) they felt that their privacy had been preserved and respected; and (g) they had no responses to this question. The respondents were requested to supply feedback about the questions and also to specify the time taken to fill in the questionnaire. The open-ended question was removed as none of the participants had answered that question. Upon successful accomplishment of the pilot study, the survey was distributed to the pretesting sample consisting of secondary school teachers (n= 4).
4.6.5 Limitations of quantitative approach
In quantitative approach the questionnaire may lead to limited outcomes as the respondents have limited response options. Thus, the results cannot always represent the actual situation. As quantitative researchers act as outsiders, they tend to miss more details when studying a phenomenon (Mays & Pope, 1995). Quantitative researchers relied too much on procedures (Jogulu & Pansiri, 2011), so some quality information is sacrificed for standardisation (Stenbacka, 2001). Moreover, as extensive statistical analysis is required in quantitative research, a lot of time is needed to perform data analysis.
4.6.5.1 Limitations of questionnaire
The questionnaire was administered to a random sample of 75 teachers of two Confessional secondary schools, which are not representative of all secondary schools, and not even the Confessional secondary schools in Mauritius. Moreover, the questionnaire measured the constructs found within the qualitative data set and those derived from the conceptual framework. The questions could be interpreted differently. Also, there was no way of knowing how truthful the respondents were with their responses as the participants took part in the study of their own free will. There was a risk that they were not honest while answering the questionnaire. To minimise the risk, a letter was included to describe the questionnaire and that it would be used for academic purposes and also that the respondents would be anonymous.
There was also a chance that the questionnaire would be neglected; therefore, the questionnaires were sent to the rectors of the two secondary schools to stress their importance and enlist their involvement. There was also the possibility that the participants perceived the questionnaire as time- and energy-consuming; hence, resulting in just ticked
101
boxes or even left unanswered. Thus, the researcher does not have any control over the responses. To reduce that risk, a 5-point Likert scale was used.