Chapter Resources

Tip: Click on each chapter heading to expand and view the extract from that chapter. Click again to collapse.

Chapter 4: Qualitative Research Design

At the early stage of a project, you need to know what approaches are supported by software? What advances in method does software offer? And what are the traps and temptations to which you should be alerted?


Early in a project, important tasks are organizational. Taking the steps to a good research design as described in this chapter, you will make a lot of records of your thinking, reading, scoping of the project, and planning. Research proposals, grant applications, and literature reviews are the intended outputs. The inputs can be messy and confusing.

Qualitative software is designed to help you handle messy inputs, so it will assist with these early records. Novice researchers often are misled into thinking they have to have "real" data before they can start using software. But there is no need to start by storing research design records separately from later records of interviews or field research. If you do so, it will be much harder to access these together throughout the project.

Start by learning how your software will store records and allow you to see them separately—in folders, sets, or groups. If you set up a project carefully, you will be able to access your plans and reviews alongside the other data records you will create when you commence interviewing or field research. So the contribution of software at the research design stage is as a reliable (but not rigid) container for plans, early considerations, and topics.

You need to learn now

  • what your software can do and how to do it,
  • how to start a project, and
  • how to manage it—saving, backing up, and transporting it.

If you are planning to combine qualitative and quantitative modes of analysis, it is important to build into your research design consideration of the ways you will "mix" these. This will involve moving data between software packages… Good mixed method research does not merely juxtapose two projects but also integrates them. To do this, you will need to plan, from the start of your research design, for the appropriate data and staging of analysis.


Computer software cannot design your project, but it can assist greatly in the data management tasks at this stage.

Once you learn software skills, you will find that starting early in software has great advantages. And more important, starting early in software does not disadvantage you. Software tools are now far more fluid than those first developed for qualitative research. A good software package will allow you to create a project and then later change practically all aspects of it as your ideas about the data and analysis grow.

At the early stage of design, you can store drafts and estimates of project stages, and using the tools taught in later tutorials, you will be able to link them and shape the ideas that inform your design. As you work the ideas and issues, you will be able to see more clearly what design decisions must be made or how, for example, you can design the sample of your study to encounter the range of discovered issues.

In the early stages of research design, the computer offers storage for documents and ideas—and the ability to link them by coding the relevant passages of documents and the relevant ideas so that all the relevant material can be retrieved later. The research design can be informed and directed by systematic storage of early explorations of the topic, serious reflection on the range of options for approaching it, and informed decision making.


  1. Software is not a method. Having chosen the software you will use, ensure that all your research moves are directed by your design and your method. Always be concerned if you are doing something just because your software can do it.
  2. Starting in software gives great advantages later—as long as you remain flexible. Where you start will not be where you go: This is built into the method. If you start your project with software, start flexibly. Use software tools for storing your changing definitions of concepts, finely coding data about them, ordering them, and exploring their relationships. Any qualitative software package will allow the memos or coding categories to be changed at any time, reordered at any time, combined, or deleted as the data direct your understanding of them.
  3. Avoid computer-assisted preemptive analysis. Software invites detailed elaboration of ideas and concepts from early data.
  4. Setting up a project is not analysis. You must also start and continue reflecting on what you do. Making and designing a project should involve processes of recording and logging your thinking about your research design. The computer is not necessary to do this, of course, but it will help you clarify the choices you have and the decisions to be made. As you prepare designs and time estimates, edit them to reflect on, change, and manage them. As you store your early ideas, describe and write about them.
  5. Beware of flexibility as well as rigidity! Don't allow the fun of setting up a project and moving parts of it around distract you from the tasks of creating a research design (at least not for too long).

Chapter 5: Making Data

As qualitative data records are created, any researcher is challenged to manage their complexity and richness well and responsibly. Software is obviously an aid for certain tasks. Qualitative packages specialize in ways of creating, importing, handling, and managing data records on the computer.


Almost all software designed for use by qualitative researchers will handle text. (Most handle "rich text" and/or word processor and other formats.) Importantly, programs have different ways of including or linking to nontext records such as audiotapes or videotapes. If you wish to use such records as primary data, without reducing them to representations, explore software that will allow you to retain them whole and to code "streaming" tape.

All qualitative software will also have some ways to store your reflections on the data, your early ideas as they happen, ensuring that these impressions will not be lost and that you can revisit your account of your data making as the project grows.

Explore your software to get a feel for the processes of making data records and handling them on the computer.

Your software should support all the following processes, which you need to learn:

How to create and edit documents in your qualitative software program, instead of in a word processor

How to import or link to all data files so they can be coded and analyzed in your project

How to store information about the data records and cases they represent (e.g., demographic data)

How to handle nontext records (e.g., photos, videos) so they can be integrated with text

How to store your reflections in memos

How to add annotations and edit

How to link between documents or parts of data records


The list of software functions above includes much that could not be done before computers. When documents were on paper, storing them and marking them up was awkward and sometimes very time-consuming, but it could be done. On the other hand, changing them, finely annotating them and accessing those comments, storing information about them, and linking that data with statistical analysis were often practically impossible.

Methods change with technology, and these lists offer the first glimpse of the effects of software on qualitative method. You can do much more with your data once they are on the computer, and, of course, you can do it with much more data.

This does not mean you should!


Four cautions apply here, and they all concern researchers' tendencies to try to fit research to computer programs rather than making the programs work for their projects. Data types, the volume of data, and the data's heterogeneity should be driven by the research goals and method, not by the computer program. Be very careful not to skew your project to what the computer seems to want. If your program won't handle the sort of data your project requires, devise a new strategy using the program's tools (and tell other researchers about it) or move to another program.

  1. Don't create bulk data records just because the computer can handle them. Most qualitative projects are hindered by large volumes of data, and some are destroyed by data bulk. In the processes of designing your study and fitting your research question to a method and ways of making data, you will have good reasons to predict the volume of data to be created. Always avoid the temptation to make the project impressive by expanding the scale of data.
  2. Be discriminating and thoughtful in deciding what qualifies as data. There is security in the computer's ability to store and access all the peripheral and often unexpected material that comes your way—background information, unrelated observations, available documentation, and so forth. But your research design should inform your selection. Never refuse possibly relevant data, but don't assume it must be immediately included in your project.
  3. Beware of tidying up your data for the computer. We argued earlier that qualitative research rarely thrives on homogeneous data. Your software does not need homogeneity and, indeed, can do less with homogeneous data than with varied data sources and types. Any computer program is better than a human brain and far better than a filing cabinet at managing complexity. And qualitative projects almost always require complex data.

Chapter 6: Coding

In Chapter 4, we introduced the ways of gathering the preliminary material that lead to research topic and question and using the computer to handle and manage data records as data are made. Now we go on to coding. The computer can take much of the clerical burden from each of the modes of coding—but it leaves you with the task of interpretation.


All software packages support coding, but they support it in different ways. Many have a choice of ways of coding. In Chapter 5, we discussed management of data records by the "descriptive coding." This is usually done by import of attributes and their values for the relevant documents or cases. Learn to do this routinely, before you start topic and analytic coding.

Software will support the following processes, which you need to learn:

How to do topic coding by making and defining categories and managing them as they develop

How to do topic coding automatically, by section or by text search

How to do analytic coding, which includes discovering and developing categories, linking them to memos, and using the computer to support the discovery and exploration of themes

How to review the data coded at these categories, online, or away from the computer, whether text or visual data (such as photographs or video segments). Most software will also support rethinking these data segments, recoding them, and developing the concepts they represent or illustrate.


You don't need qualitative software to code. .. Before software, coding was done with pens, index cards, and files. Some instructors insist that students code by hand first, to learn how to think when coding without the distraction of the software... So why use specially designed software? For the researcher, there are three critically important differences:

Computers code easily and swiftly, so coding with software is much faster and more efficient than coding on paper. Your interpretation can be far more easily and immediately stored as coded data.

Computers can store far more information than paper files can. Effectively, good software has no limit to the number of coding categories you can make or the amount of data you can code at them.

Coding data—that is, the categories you create and the selections you wish to code at them—are stored by software as pointers to the coded segments, not as marked-up or cut-up extracts of text. This information is much more flexible, and much more easily altered, than are paper records.

Unlike the filing cabinet, qualitative software can easily take you from the coded segment to the context. Some software can show you "live" all material coded at a topic or concept, allowing you to rethink, revise coding, and "code on"... to further, new dimensions of the concept.

With specialized software tools, you can ask questions about patterns of coding that were literally impossible with paper records. Qualitative researchers usually need to go beyond their first coding. It's not enough to get in one place everything said about a problem. You are more likely to want to know, for example, when people were coded as saying they had this problem, what they said—anywhere in their interview—about their trust of this advice.


The greatest dangers of qualitative computing lie in the facility, capacity, and patience of the computer! Coding is essential to most methods, but it can become a trap if you are not aware of these risks.

  1. Never let coding become just a clerical duty; if it's qualitative coding, it has a purpose. Why did you create this category? What are you planning to do with it? Coding is not a substitute for interpretation but, rather, an expression of it.
  2. Keep coding in its place. It should not dominate any part of your research timetable. When you are unsure what to do next, or when the meaning is not "emerging" from your data, it is very easy to code some more.
  3. Don't overcode. Take care to avoid the dangers of "coding fetishism", .. a compulsive activity of researchers who feel they can't think about data unless it is coded. Coding compulsively can easily replace reflection and exploration of data.
  4. Don't allow coding to keep you from other interpretive processes. Whenever you are coding, aim to store changing ideas at the same time and to reach for other tools to write memos, revisit coded data, and ask about what your coding is uncovering.

Chapter 7: Abstracting

The primary difference between commercial database software and specialized qualitative software is that the latter is designed to help with the processes of analysis and abstraction. Software packages designed for qualitative research will store not only materials but also ideas, concepts, issues, questions, and theories.

The primary difference between using a computer system in qualitative research and using a manual system is that the computer gives a different sort of access, allowing for flexible growth in the webs of ideas the data are producing and enabling the researcher to manage those burgeoning ideas by storing them, defining them, accessing them, and writing about them.

The contribution software can make at the early stages of abstracting is considerable. All researchers are diffident about first ideas. If scribbled on stick-on labels, the ideas may be lost. The computer makes it easy for the researcher to store them, define them, write about and revisit them, and revise and review them as ideas build up.


Your computer software will offer ways to make, manage, and develop categories and to work with them from the early stages of your project. And all software offers some ways of asking questions about the relations between the categories you are reflecting on.

With your qualitative software, or other software for modelling, explore computer-based diagramming. This allows you to "play" with ideas in ways that are impossible with manual methods, including layering, labelled links, and live access to data.

Software will support the following processes, which you need to learn:

How to manage and move the categories for abstracting from your data

How to store definitions and how to describe and write memos about them and log changes

How to use search and query tools to explore the relations of categories

How to model your first ideas about the topic and the hunches growing from your exploration of the data


These are the areas where software has opened entirely new ways of working with qualitative data. As your project progresses, you will learn the uses of these tools to assist your exploration of the data, checking of your hunches, and developing of theories and reporting of patterns and themes.

Category management is much more possible when the researcher is using software. Find how, with your software, you can do the following:

Flexibly copy, move, and combine coding categories without losing coding

Link logging of project process

Search the text of documents

Expand search results to appropriate context

Automatically code the results so they can become the basis for another question

Make matrices that demonstrate patterns and allow the researcher to go to each cell to see what those people said about this issue

Search and query the data (e.g., software supports creating, and optionally saving, questions about patterns of coding)

Models and diagrams were always part of qualitative research, but computer-based modeling offers many advances, for example:

Showing connections you made by coding or linking

Allowing the researcher to open data items from within the model to explore them further


Because these new tools are so exciting, be careful to use them thoughtfully and flexibly.

  1. Cataloguing and ordering categories can become a passion! Stop when the catalog is good enough for your purposes. Tools for managing, reviewing, and moving categories and for developing ideas about their relationships should be used as indicated by your research goals and design. If unplanned, those processes dominate, since categories can be changed at any time, reordered at any time, and combined or deleted as the data direct your understanding of them. The challenge is to use this flexibility to assist abstraction while avoiding the trap of constantly reworking a workable index system...
  2. Learn to conduct and interpret thoughtfully the powerful searches your software supports. When you come to use these tools in your own project, be careful always to prepare the question you are asking in plain language and then interpret the result accurately. ..
  3. Modeling with software offers attractive ways of "seeing" and showing your project. But don't assume you need qualitative software for this task. Qualitative software modeling tools offer the advantage of linking items to your data, but their output can be clumsy and difficult to visualize. A professional drawing package may provide a much more flexible tool.
  4. Use your model carefully, as an interpretive tool, with all the caution you bring to any other analysis stage. The risk is that you can "see" in a model a theme or pattern not reflected in your data. Always make the model true to the data.

Chapter 10: Writing It Up

Your writing will, of course, be done on a computer. Word processors have remade the tasks of composing, editing, and revising.


Your word processor will easily receive the data and reflections from your project if it has been created and managed in qualitative software. At any stage of your writing, you should use the ability to make reports of your data or the passages coded or discovered in your analysis.


Throughout this book we have encouraged you to record your project and your steps in analysis in a "trail." If appropriate, this can include the recording of your checks on coder consistency and sample diversity. This "log" can be efficiently maintained in your software, with links to the relevant data or categories described...

Your software's search tools have many tasks in this stage... Use them for locating relevant data or quotations, for checking the patterns you are discerning and the adequacy of your coverage of critical issues, or for finding exceptions to a generalization. Don't underestimate the uses of text search to check whether a "dominant" theme really is dominant! As you start the final reporting task, make good use of your software's data management and search tools to take stock of your memos and log trail writings so you can account for all the insights or concerns recorded during analysis...


  1. Be aware that with your data available online, using too much quoted material is easy. Patchworks of quotation do not substitute for good writing of your analysis.
  2. Be careful of the language you use as you interpret and discuss the results of software-supported search processes... Software searches are often very powerful and can certainly offer discoveries and checks not possible by manual methods, but they should always be reported for what they are—mechanical searches based on the text you provide in your project or the coding you do. (You are the weak link!)
  3. Schedule your time to include using your software well to provide material and check it out. Your analysis will be much stronger for this work. Readers and reviewers now expect that with software tools, hunches can be tested. Don't be tempted to dodge such tests and merely state your hunch and illustrate it with a juicy quotation—doing so will save time now but will cost time when you are required to resubmit your report.
  4. Expect that finishing will be harder when you know there is always another search or report you could do. Plan the processes needed, schedule them sensibly, evaluate, edit, strengthen evidence, check out patterns and explanations claimed—and finish!

Chapter 12: Getting Started

Choose your software early and become competent in computer use and all the relevant software before data collection begins. If you wait until you are making data, you will risk damaging or even losing files.

  1. If you are not confident with computers, seek help to ensure that you can manage the operating system and do good housekeeping of your files and backups. You need those skills for any research work, in any software.
  2. Ensure that you are competent in all the software you will be using, including your word-processing software (Can you format and edit well?).
  3. Now, your qualitative software. You need basic competence in it from the start of research design so that the project can benefit from software from the start. Assess your ability to self-teach. Good software will include self-teaching materials and will be supported by researchers who provide training.
  4. Use others' wisdom! Do not assume self-teaching is best: Seek colleagues or helpers who have the software skills and experience you want. Find what software support is available in your institution and use it. Go to the software maker's website to find workshops, consultants, or virtual courses. Seek out Internet discussion lists devoted to the software so you can pick up tips from other researchers and avoid their mistakes. Learning with others is usually far more productive and often faster and much more fun than learning alone.

Authors: Lyn Richards and Janice M. Morse

Pub Date: April 2012

Pages: 336

Learn more about this book