Emerald Group Publishing Limited
Copyright © 2000, MCB UP Limited
Technology and qualitative research
Technology can potentially be used at every stage of the qualitative market research process. For example, there are marketing information systems and literature databases for problem definition, databases for recruitment, the Internet as an interview medium, group support systems (GSS) for orchestrating aspects of a group interview and software to aid analysis and presentation of interpretations, to name the more obvious.
The Market Research Society (MRS) in the UK recently published a guide to software for the industry, Research Guide to Software; a Supplement of Research Magazine. The guide covers software in several categories:
CAPI (computer assisted personal interviewing);
CATI (computer assisted telephone interviewing);
end user tabs;
paper and scan;
The software in each category is profiled across a common set of features, which makes comparison easy, and the numbers of installations and users are given when available. The vendors were invited to indicate the specifications and functionality they were planning to release in the first quarter of 2000 (survey carried out in 1999) to ensure the survey findings had some relevance when printed.
However, the survey findings do not include qualitative research software. Of course, some of the software could be used in qualitative or hybrid qualitative-quantitative research. Nonetheless, there is nothing in the survey on interviewing on the Internet for qualitative researchers (see AQRP education programme below) nor any of the following qualitative tools:
group support systems;
qualitative text analysers;
In this issue we will examine software relevant to these three applications.
Group support systems
I have come across two qualitative research companies in the USA offering the use of group support systems (GSS). A number of researchers have also explored the value of such technological support to group moderation. Deal and Hodson (1997) and Sweeney et al. (1997) examined the characteristics and output of conventional groups and electronic focus groups and, in particular, the relative effectiveness of the two methods for eliciting information from participants. The electronic method they examined and have experimented with was the one location group (decision) support system. The features of a G(D)SS group discussion include:
an experienced moderator/facilitator and technical support;
group of respondents at the venue;
small PC buried into table for each respondent connected to a local area network;
moderator runs sections of the group discussion like traditional groups;
moderator can also invite individual responses to be keyed in - "parallel interviewing process" (Sweeney et al., 1996);
questions can be pre-programmed or fed in as the group progresses;
these responses can be shown anonymously on a central screen.
Deal and Hodson argue that allowing participants in focus groups to privately key in answers to some questions may:
encourage "hesitant" participants to speak more freely;
"moderate" the more dominant respondents;
reduce the problem of participants speaking at once;
encourage frank responses to sensitive issues.
Both papers looked at the productivity of GSS and conventional groups and the types of output. An earlier paper by Ruyter (1996) also demonstrated the productivity of "nominal grouping" in a non-GSS setting where group participants work on their own "brainstorming". Disadvantages of GSS include greater potential repetition and boredom and GSS failure to tap the more vivid emotional feelings (did respondents slip into a more serious mode because they were recording their answers?). Clearly, the technique needs to be used judiciously and may be useful for tasks other than those so far researched.
At present there is no guide to where such software can be obtained. Deal and Hodson note that "although several firms provide forms of GDSS, Ventana Corporation of Arizona seems to be a dominant provider". Deal and Hodson also list venues they are aware of where GDSS is available, including a mobile lab.
The only software I have come across in the UK is Perception, which is available from Questionmark (www.questionmark.com/perception/). This software was designed for tutoring and student feedback but could be used for group interviews. However, questions may need to be pre-programmed rather than "on the hoof". If you know of other software please let me know (email@example.com).
Computer tools to facilitate analysis of qualitative data have been discussed in marketing research circles from the early 1980s (Jones, 1981). Catterall and Maclaran (1998) wrote a very useful overview of what text analysers are about. They emphasise the difference between analysis and interpretation and make the point that if a research project mostly involves little analysis and relies heavily on interpretation, then such software may be of limited value. Evidence to date is that practitioners have resisted using such software and adoption has largely been by academics (Nancarrow and Barker, 1998)
A major software provider is Scolari (www.scolari .co.uk). Scolari seems to have cornered the text analysis market with products that include ATLAS.ti, Code-A-Text, The Ethnograph 5.0, HyperRESEARCH 2.0, QSR Nvivo and QSR NUD*IST 4.0 , and Wimax.
If you are considering purchasing such software then the following categorisation may be useful (see Catterall and Maclaran, 1998; Weitzman and Miles, 1995):
Text retrievers - specialise in finding all instances of words, phrases (or other character strings) in one or several data files.
Textbase managers - organise, sort and make sub-sets of text systematically with text search and retrieve facilities.
Code and retrieve programs - assist with dividing text into segments by theme or category and assigning codes to these. All text segments with the same code attached can be retrieved for examination.
Code-based theory builders - have all the capabilities of code and retrieve programs and also allow graphic links to be made between codes, for example, taste, texture and colour may be sub-categories of a category called reasons for drinking brand. Theory builders permit the testing of hypothesised links between categories in the data.
Conceptual network builders - systematically build graphic networks and permit the testing of sophisticated semantic relationships between codes.
Some software can now also store and category visual data.
One major objection to text analysers such as the above is the loss of audio and visual information (Pike and Thompson, 1995). Glen (1997) also refers to the motivational properties of revisiting recordings:
...invigorate the nerve endings; to rekindle the buzz of the interviews. It is essential to remind oneself of the ways in which respondents expressed themselves and of the tone of different parts of the discussion (flat, enthusiastic, chatty, hostile etc.), the silences
Visual cues in interpersonal communications can be crucial to understanding the true message. Revisiting transcripts and written-up accompanying notes (summaries and interpretations) is limiting. So is revisiting audio-tapes, but less so. However, in some disciplines - psychology, anthropology, ethology and human factor studies - researchers are using a computer-based system for the collection, analysis, presentation and management of visual as well as textual data.
A project using such software (Observer™ Video-Pro: see www.noldus.com) for instance, might be characterised as follows:
During the group discussion or in the field
video-record groups, depths or "real" behaviour say in store (the latter can be time sampled);
or use pen cameras (private eye), hand held or fixed, concealed for unobtrusive recording or accompanied shopping or accompanied consumption;
mark or code interesting behaviour remotely on the video as it happens.
Analysis (after transferring analogue video to digital if necessary)
revisit and analyse videos of qualitative interviews or behaviour afterwards on PC or small portable computer;
Observer Video Pro allows you to have several boxes on screen (the live action and parallel coding in another box, transcription in another box and interpretation notes box - all are synchronised with the video action;
quantify if appropriate;
using codes or key words video action can be found in a fraction of a second.
easily edit video data for supporting presentations (minutes not hours).
My personal experience is that this software makes analysis of qualitative data more interesting (motivating), thorough and user-friendly. I can also foresee its use for training new qualitative interviewers (showing the best of the best moderators).
Objections to the use of computers for analysis are well documented and include time and cost, as well as the fears that the programs may drive the analysis process and the analysis becomes decontextualised (see Catterall and Maclaran, 1998; Nancarrow and Barker, 1998).
Nonetheless, there are constantly new developments which may serve to overcome some of these barriers. Please let me know of your experiences using any form of technology in qualitative research and perhaps we can put together a regular review of qualitative technology.
Clive NancarrowBristol Business School, UWEclive.firstname.lastname@example.org
Catterall, M. and Maclaran, P. (1998), "Using computer software for the analysis of qualitative market research data", Journal of the Market Research Society, Vol. 40 No. 3, pp. 207-22.
Deal, K. and Hodson, T. (1997), "Electronic and conventional focus groups: comparisons and relative merits", Canadian Journal of Marketing Research, Vol. 16, pp. 61-71.
de Ruyter, K. (1996), "Focus versus nominal group interviews: a comparative analysis, interviews", Marketing Intelligence & Planning, Vol. 14 No. 6, pp 44-50.
Glen, R. (1997), "Analysis and interpretation in qualitative research: a researcher's perspective", in Butterfield, L., Excellence in Advertising, Butterworth-Heinneman.
Jones, S. (1981), "Listening to complexity: analysing qualitative marketing research data", Journal of the Market Research Society, Vol. 23 No. 1, pp. 26-39.
Nancarrow, C. and Barker, A. (1998), "Plug and play", AQRP Trends Day, London.
Pike, R. and Thompson, L. (1995), "The technological fix in qualitative research - let's be honest in what we are trying to achieve", Proceedings of ESOMAR seminar, Looking through the Kaleidoscope: that is the qualitative mission?, Paris, December.
Sweeney, J.C., Soutar, J.N., Hausknecht, D.R., Dallin, R.F. and Johnson, L.W. (1997), "Collecting information from groups: a comparison of two methods", Journal of the Market Research Society, Vol. 39 No. 2, pp. 397-411.
Weitzman, E.A. and Miles, M.B (1995), Computer Programs for Data Analysis: A Software Sourcebook, Sage Publications, Newbury Park, CA.