Qualitative Data Analysis

Similarly there is no single approach to understanding qualitative data. It is not sufficient to say that the common feature of qualitative data is that they deal with meaningful action because as positivists use words as data and claim that their data is meaningful, providing information about human data. So this isn’t a unique claim about qualitative data. What we can say about qualitative data is that it is data produced in the making, whether that be data traces produced in the course of social interaction or whether that be data generated in the course of research with the purp with the purpose of data collection.

We tend to refer to qualitative data as being naturalistic emerging from the real world context of human behaviour and communication through which this data is then produced. Maxwell and Chmiel in their chapter in Flicks The Sage Handbook of Qualitative Data Analysis discuss the need to theorise qualitative data analysis, arguing that the way in which qualitative data is understood has implications for the ways in which researchers analyse their data. Delamont 2016 highlights the importance of analysis in qualitative research in chapter 11 of her book Fieldwork in Educational Settings: Methods Pitfalls and Perspectives she provides some examples from qualitative studies in education which stand out because of their contribution to conceptual ideas.

This chapter is particularly useful for illustrating how researchers can approach qualitative data theoretically. In Getting Smart: Feminist Research and Pedagogy within the Postmodern Patti Lather in 1991 refers to qualitative data analysis as a black hole meaning that the details of data analysis are unknown similarly St. Pierre and Jackson observe how qualitative data analysis is often taught in terms of coding leaving unknown how to theoretically approach the analysis of data.

If we think further about the black hole of qualitative data analysis as St. Pierre and Jackson go on to explain, it is easier to teach qualitative data analysis as coding than it is to teaching thinking with theory. So the result according to St. Pierre and Jackson is that this results in the reduction of qualitative data to brute data that is waiting to be coded and categorized by equally brute words. One of the problems with this approach is that it supports a positivist quasi-statistical and analytic practice.

The data exists out there in reality representing a reality. Practices such as blind coding by multiple coders with no knowledge of the context of the research in an effort to achieve inter-rater reliability can be seen as examples that support a problematic approach to qualitative data that fits more with the positive approach. Koro-Ljungberg 2016 makes a similar point about qualitative data, that it becomes reduced to lists or collection of quotes. However an interpretivist epistemology underpins qualitative research and so therefore we need to think theoretically about approaching the the the data we are working with in order to achieve a contextual understanding of social practices experiences and interactions.

St Pierre and Jackson go on to refer to the vacuum cleaner approach to qualitative data collection, this is where we are inclined to sweep up all the data from our studies the result is that we gather up a great amount of data in terms of quantity and all the data we have collected has equal status in terms of its worthiness for an analysis. The problem with this approach is that it pays little attention to the quality of the data that we c… collect we take for granted that data we have collected from interviews or observations for example are important.

But St. Pierre and Jackson argue that instead we need theory to help us determine what is data and what is good good data. Koro-Ljungberg 2016 observed that researchers might be overwhelmed by the quantity of data and approach the data as best they can that is through a brute coding approach that we’ve already that has already been mentioned by St. Pierre and Jackson. Researchers needing to approach a large quantity of data may do so mechanically and consequently they may not ask critical questions about the data.

To engage with qualitative data critically means as Koro-Ljungberg discusses experiencing uncomfortable messy data and engaging with the resistance of data to reveal meaningful insights. Approaching qualitative data mechanically might be expedient given time frames and external pressures to produce reports as well as make the the data gen generalisable. Something that you shouldn’t really do with qualitative research but something that may be demanded of you by external agencies.

08:18 - So now we’ve considered the importance of theory to data analysis we will pay some attention to the process of qualitative data analysis.

08:31 - Basit identifies data analysis as the most difficult and most crucial aspect of qualitative research Basitt’s statement echoes concerns expressed across the social sciences about the challenges researchers face in approaching qualitative data.

08:53 - Miles,1979 describes qualitative data as an attractive nuisance. On the one hand it is attractive because qualitative data is in depth, rich and drawn from the real world and represents the authentic voice of participants. However it is also in nuisance with the process of analysis being challenging. Not only is this due to the time needed to work through potentially large volumes of data but as Miles also argues the methods of analysis are not well formulated.

09:43 - Following on from Miles Anselm Strauss makes a similar point arguing that we have a very long way to go yet in understanding how we do qualitative analysis and how to impr improve analysis. Since miles was writing in 1979 and Strauss made this observation in 1988 has our understanding of qualitative data analysis improved? Maxwell and Miller writing some 20 years after Anselm Strauss in 2008 would argue that this claim remains true and that the development of a theory of qualitative analysis needs to come needs to continue.

We now turn to coding as this illustrates the practical application of qualitative data analysis. We can start by asking the question of whether coding is an art or a science or whether it can be reasonably considered a combination of the s two. Basit offers us something to consider by claiming that coding is a dynamic intuitive and creative process of inductive reasoning thinking and theorising. and that’s uh from the 2000 that’s from 2003 page 143. As Basit argues coding and analysis are not synonymous though coding is a crucial aspect of analysis.

So we need to remember that coding can be atheoretical in that we use it as purely a mechanical exercise in categorising the data we have. So in this case coding could be limited to pure description about the data.

11:53 - Before going any further it is worth asking the question what is a code. And Saldaña in the coding manu in his coding manual for qualitative research offers the following definition of a code. A code is most often a word or short phrase that symbolically assigns a summative salient essence capturing and or evocative attribute for a portion of language-based or visual data, meaning that a code represents something and gives an indication of meaning providing the basis for analysis.

12:43 - Coding as Maxwell and Chmiel observed is a categorising strategy. Coding involves attaching labels and grouping segments of data into categories. The categories can then be compared with similarities identified and connections between segments and categories um they can then be analysed. The focus is on finding similarities and as Maxwell and Chmiel observe the contiguity-based approach to data that is finding the differences between the data is often not as strong as similarity approach.

Maxwell and Chmiel discuss types of coding categories starting with outlining organisational categories they describe these as abstract bins for sorting the data for further analysis this means that organisational codes don’t tell us anything about what is happening but they are purely or organisational, they operate as somewhere for further data to be placed to be located.

14:11 - Through the use of substantive codes researchers can approach what is happening and start to gain an understanding of the participant’s ex experiences. As Maxwell and Chmiel say substantive categories implicitly make some sort of claim about the phenomena being studied. That is they could be wrong rather than being conceptual boxes for holding data. Researchers may use apriori coding for substantive categories, this means codes that the researcher has already identified through existing knowledge about participants or a well-developed theory.

However categories may not always be identified through apriori coding. Primarily descriptive in that they describe beliefs or concepts held by the participants categories may use participants own words and are generally grounded in the data, so in other words they emerge from examining the data not through previously identified categories. So here we can see that researchers can use apriori coding so that’s codes that they’ve already identified before they really start analysing the data, they would then look at the data and apply an existing code or they can use a good grounded code that emerges from reading the data and identifying um codes that categorise that data.

15:56 - Finally theoretical categories are what researchers use to place the data into a theoretical framework. Derived from existing theory are generated in inductively from an examination of the data. So a similar approach they can use apriori coding or good grounded coding. And to highlight the importance of coding to qualitative research is um this quote from Anselm Strauss any researcher who wishes to become proficient at doing qualitative analysis must learn to code well and easily.

The excellence of the research rests in large part on the excellence of the c coding. And as we come to an end I’m just going to very briefly mention CAQDAS which is an acronym for computer assisted qualitative data analysis. So over the last the last few decades computer packages dedicated to the qualitative analysis have become increasingly popular and this is a consideration that you um this is this is consideration for your own qualitative data analysis.

This is a selection of the most popular computer packages designed to support qualitative data analysis. The decision to use a computer package to assist with your data analysis needs to be an informed one. You need to consider whether it will be worth you investing the time in learning how to use a package as it can take some time setting up your data within a database so that it is ready for analysis. As Basit observes it takes a considerable amount of time to learn how to set up a database on one or one of these packages.

Coding manually may be time consuming but setting up the coding system on a software package can also involve a lot of time, though it will save time in the long run if you then go on to analyse a lot of data. So if you have limited time available in the case of a dissertation deadline you may consider it too time consuming and prefer to to use um use more traditional methods of qualitative data and analysis, analysing by a by hand. Similarly if you don’t have large quantities of data then it might be a more efficient use of your time to process the analysis manually.

The real benefits of computer assisted qualitative analysis software are where the researchers are confronted with an overload of rich and intricate data and where their ability to process information in meaningful ways is seriously tested. And you’ll find that in García‐Horta and Guerra‐Ramos page 151 which is in the which is listed in the reference slide. What needs to be born in mind is that these tools offer assistance in qualitative data analysis they don’t do the analysis for the researcher.

García‐Horta and Guerra‐Ramos do discuss how they how they used NVIVO and um MaxQDA and found them helpful in avoiding data over load and could be used to assist with managing and organising data in a way that would be onerous if it was done manually.

20:14 - Similarly they found the packages helpful in facilitating information availability as the categorisation and coding system enabled efficient access to relevant documents and sources of data. However they do point to some limitations one of which is a temptation that technology invites us to treat the process of data analysis as a mechanical exercise rather than the dynamic intuitive and creative process of inductive reasoning thinking and theorising that is outlined by Basis, the quote that i’ve already referred to.

In short it is your choice whether a software per package is good for you.

21:13 - Trial periods are available through most packages to allow you to practice with the restricted of amount of the data so this provides you with an opportunity to try out the package and gain an insight into its features.

21:35 - And finally a couple of slides with the references to the sources that were mentioned in the course of of of this presentation. .