CAPÍTULO III. ESTADOS DE BIENESTAR
3.2. Los Estados de bienestar en un mundo complejo
The quantitative and qualitative data generated from the questionnaire and interviews were organised according to the each quantitative and qualitative process of data analysis as explained in the following sub-sections:
4.8.1 Statistical analysis
A total number of 421 questionnaires were received. They are organised based on numbering for identification code from number 1 to 421. I used SPSS to organise and analyse the data from the questionnaire. At the initial stage I coded the responses according to number. Each of the five answers to the statements in Likert scale questions in Section A and C was given a number; ‘strongly agree’ was coded as 5 and ‘strongly disagree’ was coded as 1. For Section B, is about the teachers’ reactions towards the different types of SEN, semantic differential scale which consisted of seven items including adjectives like anxious-relaxed, worried-self-
assured and negative- positive. The items were coded from 1 to 7. Regarding Part 2- Demographic and teaching information, “YES” was coded as 1 and “NO” was coded as 2.
With regard to the missing data such as unanswered statement or a statement which has more than one answer, this was marked as missing when the data was fed into SPSS (coded as 99). As there were not many missing answers by the same participant, no questionnaire was excluded from the study. Similarly, no statements were omitted from the analysis as there were not many questions left unanswered. After the initial coding, the data were checked to identify if any errors occurred and to ensure that the data was entered accurately. This was done
through generating tables of frequencies for all statements and checking the values in the table.
The demographic and teaching information were then analysed by SPSS to get the general ideas of the participants followed by the individual variables. Frequency distribution analysis was run for each questionnaire statement to check how many respondents have answered. Frequency tables were produced which provide the number of participants and the percentage to each of the categories for the variable. These frequency tables helped in gaining an
understanding of the overall distribution of the responses at the initial stage of the analysis. After that descriptive statistics (mean, median, mode, variance and standard deviation) were calculated to summarise patterns in the responses of the preschool teachers in the sample as well as in inferential statistics where appropriate. Finally, in the last section three open-ended items were qualitatively analysed using content analysis.
In summary, Table 9 presents the data analysis as follow:
Table 9. Data analysis
Research Questions Statistical procedures
Purpose What is the attitude of the preschool
teachers towards the introduction of IE in Malaysian government preschools? Table -frequency, percentage and group
To report the characteristics of the sample based on the demographic data: Gender, age, teaching experience, location, race, academic qualification, SEN training received
Table -mean, SD
To report the mean scores measuring:
Cognitive, behaviour and Affective
Content Analysis
To analyse text by coding based on the following open ended questions:
What extra things which could make the participants’
responses more positive towards children with SEN (specifically with learning difficulties and emotional and behavioural difficulties)?
What changes to be considered in the classroom, school
environment and community before children with SEN are included in the mainstream classes?
What factors that influence the preschool teachers towards the introduction of IE in Malaysian government preschools?
t- test To investigate the difference between the mean scores of the teachers on the two affective scales; learning difficulties and emotional and behavioural difficulties
One way MANOVA
To test for differences in the cognitive, affective and behavioural components between group identified in terms of: gender, age, teaching experience, location , races SEN training received, academic qualification
Pearson correlations
To test for correlations between mean scores of the cognitive, affective and behavioural scales. To what extent do these factors
affect the preschool teachers’ attitudes towards IE?
Thematic analysis Semi- structured interview
To search for themes among codes for meaningful patterns.
4.8.2 Transcription
I used NVivo software to transcribe the audio recordings of the interviews in full. I adopted a non-verbatim approach to the transcription (not include all speech phenomena such as hesitation) since such information would not have increased my ability to answer the research questions. However, for the purpose of reading the data I included annotations such as pause indicated by three continuous dots (…) and punctuation where possible. As mentioned earlier, all of the transcriptions were done in Bahasa Malaysia and only one transcription was in English Language.
4.8.3 Data coding
I employed the general inductive approach for the data analysis where multiple readings and interpretations of the raw data were carried out. According to Thomas (2006), this process
allows researchers to make decision on the basis of the research questions about what is more important and less important in the data.
In the open-ended question, three items were analysed using thematic analysis (Braun and Clarke, 2006). I used Nvivo11 software to carry out the coding process with the initial
categories as the primary nodes. I coded the data for each item and created sub-nodes where there appeared to be useful sub-themes emerging within the data.
As in the interview, all the 18 interviews were transcribed. For the interview transcriptions, I also employed thematic analysis to help me to organise the large amount of qualitative data. The first phase was familiarising myself with data. During the transcription process, note-keeping on initial thoughts on the data helped to make sense of the data and identify the key issues and themes. (Wolcott, 1994) mentioned that a closer look and relook of the interview was made afterwards in order to develop a comprehensive picture of the content. Therefore, during this phase, I transcribed the data followed by reading and re-reading the data. I also noted down some initial ideas taken from the data.
In the second stage, with the research questions in mind, I began the thematic coding in which all the information from each teacher interview was classified into main categories (nodes). Then I classified the main categories into sub-categories. The third phase was searching for themes. In this process I collated codes into potential themes and gathered all the data which were relevant to each potential theme. This process involved coding any interesting features of the data in a systematic way across the entire data set. Through NVivo I was able to highlight related comments and placed them in the suitable sub-categories. Kvale and Brinkmann (2009) commented that this process of organisation allowed for convenient handling of the data which also provided easy identification and comparisons for different patterns.
The fourth phase was consisted of reviewing themes where I checked all themes in relation to the coded extracts and the entire data set. Here I generated a thematic map of the analysis. The fifth phase was defining and naming the themes where the ongoing analysis was done to refine the specifics of each theme in order to generate clear definitions and the name of each theme. Finally, the sixth phase was producing the report where the selections of compelling extracts relating to the research questions and literature review would be done to produce the report of the analysis.