RECOUP/Data management and analysis

Session overview

The basic presumption behind this session is that it is best to think about analysis before starting data collection, since failure to do so can leave a researcher with several major problems. Indeed, a common difficulty faced by new researchers is that they return from data collection and have little idea of what to do next; and when they do decide on a plan for analysis, they discover that they do not have all the material they need or have too much material which is of little relevance to the project.

Time:

If you want to address all three objectives outlined below, we suggest a whole day in total for this session. If you focus on the first two objectives and restrict yourself to general principles, you could manage with a long half-day.

Objectives of the Session:

The purpose of this session is to discuss three main topics:

• ‘Basic housekeeping’: how qualitative data should be stored, managed and distributed
• The essentials of qualitative data analysis
• The strengths and challenges of using a computer-assisted qualitative data analysis software (CAQDAS) package to help code and access qualitative data.

Approaches to the Session

How you deal with the first objective depends on the kind of workshop you are running:

• If this is the beginning of a project, or if you want to run this session after a few interviews in such a project have been carried out, then you might want to use this opportunity to go into the details of your project and begin to take decisions. Or,
• If this is a general training session, and your groups includes solo researchers as well as people likely to work in teams, you might want to discuss the general principles and encourage people to share their ideas of how they will resolve basic management issues in future large projects, PhD or other degree projects etc.

The second objective is crucial: we would expect you to include a session of this kind in any course.

How you deal with the third objective again depends both on your participants and on local constraints:

• If you already have purchased or have access to a specific CAQDAS program, then you might move directly to orientation with respect to that program; alternatively, you might want to reserve most of this material to a time nearer when the analysis will actually take place (because people rapidly forget how to carry out basic tasks in a new program, unless they get quite a lot of hands-on practice immediately).
• If your group expects to use a CAQDAS package, but has not yet decided which one, or has not yet got access to the package, you might restrict your discussion to a general overview, and offer some advice on choosing a package for those who have yet to make their minds up.
• If there is little prospect of people having access to a CAQDAS package in the near future, or if you do not feel that it is appropriate to use such a package, then you might be better to focus only on the first two objectives. Managing a small project without using a CAQDAS package may make most sense; but the larger the project, the more difficult data management without a package becomes.

Management of Data

Time: 30-60 minutes

Preparation: This depends on whether you are training a group of researchers in an existing project or not. If so, you should have start with your own ideas about managing the data you collect. But you’ll certainly need a flip-chart, black- or white-board or computer plus projector for the brainstorm sessions.

Process:

1. Small groups (2-3 people in each) to come up with three issues that need to be considered when thinking about the management of data (10 minutes)
2. Feed-back session in which issues are listed (5 minutes)
3. Full-group discussion of possible solutions to the more important issues (from 15 minutes to 45 minutes).

If you have a project which needs to set basic ground-rules with respect to these data management issues, you might like to prepare a handout that can be given out at this point.

Notes for facilitators: BJOERN: this would need to be in a box or something

Main issues we suggest should be considered here are:

• naming files, so they are easily retrieved and can be grouped if necessary
• how and where to save files, in electronic and paper copies
• when to anonymise personal data in files
• how to ensure security and privacy
• when and where to carry out back-ups (and how to name these)
• when to translate material into the project language (if necessary)

Translation is often a big issue for projects that have limited resources. There is a Handout which discusses some general principles. But in practice hard decisions may need to be taken about whether (or how much) to transcribe recorded interviews or focus group discussions; when (and how much) to translate; and who should translate, as well as what kinds of quality controls will be introduced to ensure the data are good enough for coding and eventual possible citation.

Link to HO_Issues of Translation

Basic Introduction to Qualitative Data Analysis

Time: Allow about 2-3 hours (half a day) for this part of the session. Many people find qualitative data analysis the hardest thing to understand, and the more they engage with practical examples, the better.

Preparation: You’ll need a flip-chart, black- or white-board or computer plus projector for the brainstorm sessions.

You might want to bring in a couple of books with indexes, to show people the similarities between a familiar exercise – looking something up in an index – and basic descriptive coding.

We have provided some photos of non-computer-based coding – what is often called ‘cut and paste’. This helps people to understand some basic principles of collecting data by individual and by theme or topic (reading across the sheets to understand the variation in responses to a particular question or around a particular theme, and reading down the sheet to see how an individual’s responses fit into a larger picture of their life). If you have an example of your own, it would be worth bringing it in and talking participants through what you did. Alternatively, you could create an example, using invented quotes, cut and pasted onto large sheets of paper or card.

You’ll also need a sample (perhaps one page) from an interview transcript, fieldnotes or other form of qualitative data, one copy for each participant and an electronic version that can be shown on a projector to help in the feed-back session.

In the PowerPoint we’ve provided a worked example from the PhD thesis of one of the authors of this manual. However, it would be useful to work with an example that you feel more comfortable with, in which case you can use this PowerPoint as a template.

Process:

You can use the PowerPoint as the basis for this session. We suggest that you break up the flow with a series of full-group brainstorms or small-group tasks. For example,

• Ask people what they see as the main differences and similarities between the analysis of quantitative and of qualitative data.
• Draw on the experience of the group in making sense of unfamiliar places – noticing, collecting and thinking are processes we all use.
• Use the sample extract as a basis for getting people to generate codes, and to think about the differences between descriptive and analytic codes.

Link to PPT_11_2_QDanalysis

Introduction to the use of computers to help with qualitative data analysis

Time: 45 minutes

Preparation: This depends on whether you are training a group of researchers in an existing project or not, whether you have the need, desire (and funds) to use a software package, and if so, whether you have already selected and purchased a package that you want everyone to be familiar with.

But you’ll certainly need a flip-chart, black- or white-board or computer plus projector for the brainstorm sessions.

Process:

If you want to run this just as a general introduction to CAQDAS (Computer Assisted Qualitative Data Analysis Software), we suggest you use the attached PowerPoint with worked examples from projects with which you are familiar.

Link to PPT_11_3_CAQDAS

There is a Handout: you could give this out at the end of this session

Link to HO_CAQDAS

Additional Resources:

There is a comprehensive web-page that covers many issues with respect to different versions of qualitative data analysis:

http://onlineqda.hud.ac.uk/Introduction/index.php

This site includes a summary statement comparing the various CAQDAS packages available in 2004, and links to other review pages:

http://onlineqda.hud.ac.uk/Which_software/index.php

We are not recommending any particular CAQDAS package: to our minds, each has its strengths and weaknesses. For specific reasons, Atlas.ti is the package that is being used in the RECOUP project. Some members of the team prefer one of the other packages on the market, such as HyperRESEARCH V.2.06, MAXqda, N6, NVivo, Qualrus

Each program offers a Tutorial to help you get started, but they are not always very easy to use in a group.

If you want to use NVivo 7, http://www.lynrichards.org/HQD_Tutorials.htm offers access to self-teaching tutorials and access to a global network of people using the package

If you want to use Atlas.ti then you might find the attached PowerPoint by Laura Jeffery useful. This has been used in workshops to intensively train people in the basics of the program (in 2 days) or teaching the program and developing the codes and doing some practice coding on a data-set collected by the participants (5 days).

Link to PPT_Atlas_Introduction

---

Singal, N., and Jeffery, R. (2008). Qualitative Research Skills Workshop: A Facilitator's Reference Manual, http://oer.educ.cam.ac.uk/wiki/RECOUP, Cambridge: RECOUP (Research Consortium on Educational Outcomes and Poverty, http://recoup.educ.cam.ac.uk/). CC BY-NC-SA 4.0. (original page)