PXGS6102DataAnalysisQualitative

What is Qualitative Data?

"..... detailed descriptions of situations, events, people, interactions, observed behaviours, direct quotations from people about their experiences, attitudes, beliefs and thoughts and excerpts or entire passages from documents, correspondence, records, and case histories" (M.Q. Patton, 1990, Qualitative evaluation and research methods. Thousand Oaks: Sage Publications. p.22). Usually quantitative data will be in the form of words and narratives, but may include visual images, video recordings or audio-recordings.

Introduction

Until the late 70s and early 80s, there was a lack of guidelines on how to analyse qualitative data unlike quantitative data where the methods of analysis were well established. What was available were general statements that when analysing qualitative one should avoid bias but not on the details of actual analysis. It was with the publication of books by Guba & Lincoln (1981) and Spradley, (1981) that it became clearer how qualitative data is to be analysed. However, However, some theoreticians continue to argue that analysis of qualitative data is an art rather than a science and hence can be rather intuitive.

Issues in Analysis of Qualitative Data

W. James Potter (1996), An analysis of thinking and research about qualitative methods, NJersey: Lawrence Erlbaum Associates; presents FOUR issues about data analysis that needs to be understood:

1) Researcher knows what to expect or does not know what to expect

One extreme: you have very clear goal what data you need which may be

based on theory/theories and you select data accordingly, i.e. expectations

predetermined

Other extreme: you do not know what to expect and as you gather data

you begin to focus and become selective, i.e. expectations emerge

2) Researcher describes or does not describe the process of data

analysis

One extreme: you describe the steps of data analysis

Other extreme: you do not describe the steps of data analysis

3) Researcher goes beyond or does not go beyond the data

One extreme: you only describe the data literally

Other extreme: you construct patterns and make inferences

4) Reseacher attempts to generalise or does not generalise

One extreme: you only report data or patterns about your observed

subjects during the time and place they were observed and do not

generalise

Other extreme: you conclude from the data and use the broad conclusions

to generalise

What is Qualitative Data Analysis?

"The mass of words generated by interviews or observational data needs to be described and summarised. The question may require the researchers to seek relationships between various themes that have been identified, or to relate behaviour or ideas to biographical characteristics of respondents such as age or gender. Implications for policy or practice may be derived from the data or interpretation sought of puzzling findings from previous studies. Ultimately theory could be developed and tested using advanced analytical techniques"

(Anne Lacey and Donna Luff, 2001, Qualitative Data Analysis. London: Trent Focus. p.3).

Transcription

Data that is recorded (audio, video or handwritten fieldnotes) have to be transcribed. Do not forget to include the "Hmm..., the Ah... or Well...er...what is mean" in audio and video recordings as they are indicative of emotions expressed, pause and non-verbal cues. Similarly, gestures should be recorded as they imply some non-verbal information. Give some examples of gestures that may be interpreted as information? Remember, transcribing takes 4 times the length of the interview (i.e. 15 mins of interview takes 1 hour to transcribe).

Organisation of Data

Organise your data so that you can easily retrieve them.

Each interview conducted, give a number or code (eg. Int1, Int2) - may need to give pseudonyms (eg. Mike - not his real name)

Fieldnotes: break up into sections identified by date or context (eg. section 1: children playing games (24.08.04 : sandpit)

Make sure you have a system that will help you identify the actual person with the pseudonym you gave

How do you Conceptualise your Data?

The method you use to analyse your data depends on how you conceptualise your data. Do you intend to answer a specific question? OR Do you intend to generate new understandings of a particular phenomenon? Do you notice the differences between these two approaches?

If your intention is to answer a specific question, then the approach you will most likely adopt is called FRAMEWORK ANALYIS

If your intention is to generate new understandinfs, then the approach you will most likely adopt is called GROUNDED THEORY

FRAMEWORK ANALYSIS

This approach is more aligned to applied research, where the aims of your research is to obtain specific information that tells you something about the outcome of the study or you need to make recommendations about a particular policy or action; usually this information needs to be made available within a short timescale.

This method of analysis is inductive but it allows for a prior determination of categories or themes. A prior means that the themes are established in the beginning, but it must be remembered that they may be changed and modified.

The steps for this method of analysis is as follows:

Familiarisation: transcribe data

Identify a Thematic Framework: Themes are categories in which data are grouped. For example, the theme 'fear of being wrong' could be expressed in observations made of students: Refusal to attempt answering questions, do not raise their hands to answer even simple questions, do not even try to answer when called upon by the teacher. The thematic approach is developed and refined as you go along analysing.

Coding: Here, you apply the thematic framework you have developed to the data. You have to use numerical or textual codes to idnentify specific pieces of data which correspond to the different themes. To code your data you can use the following techniques:

= Charting: use the headings from the thematic framework to

create charts of your data so that you can easily read across the

whole dataset.

GROUNDED THEORY

The Grounded theory approach allows you to generate social theory from data systematically. It is a way of thinking about data and it is conceptualised. This approach is inductive, in that the resulting theory "emerges" from the data through a process of rigorous and structured analysis. How is this approach different from other data analysis approaches?. The grounded theory approach aims at coming out with a theory as the final output of research; while other forms of qualitative analysis may 'stop' at the level of description and interpretation. The Grounded theory approach aims at theory development and has been also called 'analytic induction'.

A gounded theory consists of 'plausible relationships' (Strauss & Corbin, 1998) among sets of concepts, which are directly developed from data analysis. Theory, in this sense, provides a set of testable propositions that help us understand our social world more clearly, rather than absolute 'truths'. This approach starts with a clear, but often broad, research question which identifies the general area to be studied.

Steps:

open coding

identify emergent concepts

conceptual coding (use emergent concepts)

refinement of conceptual coding schemes

clustering of concepts to form analytical categories

searching for core categories

core categories lead to identification of core theory

"Process of constant comparison" = concepts or categories emerging from one stage of data analysis are compared with concepts or categories emerging from the next stage of data analysis. This constant comparing of concepts or categories is aimed at forming the emerging theory. This procedure is continued until "theoretical saturation" is reached; i.e. no new significant categories or concepts are emerging. One should also remember that the procedure of grounded theory analysis is NOT LINEAR rather it is cumulative and often the researcher REVISITS the data when new concepts emerge.