Introduction

The best known and widely used method which originated from PCP is the Repertory Grid Technique (cf. Neimeyer et al. 1990). Today Repertory Grid Technique is a heading of a variety of interview methods which are even used by many people working not within a PCP frame of reference. One of features of these methods is the structured (grid) form in which data are being collected and analysed. The results of an interview can be represented in matrix form and thus provide an ideal basis for computer based analysis.

The first suggestions to analyse psychological space using statistics came from George A. Kelly (1991) the founder of PCP himself. He devised a nonparametric form of factor analysis for grids. The next major step was done by Patrick Slater (1964, 1966, 1976) who developed computer programs for the Principal Components Analysis of Repertory Grids. Another major contribution was the adaptation of cluster analysis for use with grid data (cf. Jankowicz/Thomas 1982; Shaw 1980). Most of the computer programs available today still use these statistical procedures or derivatives of them.

Why Grid Software?

In the past there have been (sometimes quite heated) debates about the usefulness and adequacy of statistical analysis within the theoretical and methodological concept of PCP (cf. Bell 2000; Ravenette 2000). This is not the place to deal with all the arguments put forward in this debate nor to solve it. But at least a few points may help to understand the controversy:

People who are sceptical about the usefulness of statistics typically claim that numerical data which are the basics for statistical analysis can not make full use of the psychological meaning which is the background of the numbers which in the end are processed by the computer – and that a hermeneutical approach is necessary to catch this meaning. This is perfectly true – and should be kept in mind when grid software is used: A computer does not understand, it is only capable of identifying numerical characteristics and structures of data. So whatever software is used the result does not go beyond results which are meaningful in a numerical sense. Whether these results also have a psychological meaning remains always the question.

On the other side people who are sceptical about a hermeneutical approach claim that content analysis strategies often lack (interpreter) reliability and are limited by the cognitive capacities of interpreters to detect contingencies and structures hidden in the data. An argument that could and should be added (but seldom is): content analysis has to rely on words. But words may be misleading because there is no 1:1 relationship between words/expressions and constructs (cf. Kelly 1991).

So content analysis as well as statistical analysis have specific advantages and disadvantages. Where content analysis is important to identify psychological meaning statistical analysis may unearth structures which are not obvious or even obscured by words.

Which program?

Among the many characteristics of the programs which are currently available some are especially important for the decision which one to choose.

Operating System

First of all the program should run under the operating system one uses. Today all relevant programs will run under Windows. But some of them, the older ones which were programmed to work under the DOS operating system, do no make use of the advanced possibilities Windows as successor of DOS offers.

From the programs available today two (RepGridIV and GridSuite) can also be used under MacOS.

Elicitation/Analysis

One of the crucial differences between programs is their ability to elicit and/or analyse grids. This is not only a question of a program’s functions but also one of the underlying philosophy and the usability of the program.

A program which is focused on analysis provides a set of statistical routines, usually a specialised and sometimes modified selection of routines that can also be found in general statistics packages like SPSS. A program of this kind provides an input facility for data, a selection of calculations and output of results as tables and graphs. The typical user of such a program is a researcher who is familiar with the statistics and able to interpret the numerical results.

The central task of a program which is used for the elicitation of grids is completely different: it has to offer some kind of computer-client-interaction during which a repertory grid is developed. Where the analysis programs are judged on the basis of the statistics they provide, the elicitation programs are judged on the basis of the quality of the person-machine-interaction they enable.

In both cases there are extremes: From simple calculations to impressive menus of statistical options, from elicitations which offer just a form to fill to different flexible interview procedures which make full use of the visualisation and interaction possibilities of contemporary computers.

Most programs offer both, elicitation and analysis, but are clearly specialised in one area. The result of which is that some of them are only useful for research and should not be used with clients where others may be used with clients but can not be used for proper research. However there also programs which provide advanced elicitation and analysis options and/or can be combined with other software if special functions are needed.

Import/Export Options

Again there is a wide range from software which has no import or export option at all besides typing in the data and having a preformatted printout to software which is able to handle and convert all sorts of data. Whether these options are important will depend on the specific use. But if it is possible that data must be edited or analysed in different ways later it is vital to choose a program which does at least not hinder this. Then the most risky approach is to use a program with limited functions and no export options.

Range of Functions

The range of functions which the available programs offer is wide and it differs considerably between programs: from programs which concentrate on a few functions to programs which aim to offer an almost complete set of tools. The main point for a decision here is which functions are needed or may be needed in the future (s. export).

Usability

There is no clear relationship between range of functions and usability: there are programs which offer very limited functions and are still difficult to use, others with numerous functions which can be handled without problems – and vice versa.

Again it depends what usability means. One important question here is: usable for whom? What may be a nice program for the researcher may be a completely confusing program for the client. What may be a useful program in a therapeutical context may be wrong in a business consultancy context.

Another aspect is the frequency of use. If one is using a program continuously over months for example for a thesis some disadvantages (e.g. irritating menus, poor documentation and help system) may be overcome by day to day practice and routine but not if the program is only used shortly a few times per year.

Overview

To give an overview over the relevant software turns out to be quite difficult. Just to name a few factors:

Age: INGRID by Slater was one of the first and most influential programs, and it’s routines are still being used in the programs today. And there will be still users of the original program as it was adapted for the personal computer, especially those who programmed some extra functions themselves. But for those who are searching for grid software today this is only of historical interest (cf Sewell et al. 1992).

Availability: Especially focussed on programs for elicitation of grids Bell and Fransella point out that “the programs that are currently available have a short shelf-life in that they are unlikely to be updated once their creators have retired” (2004, p. 38). But this is also true for some programs developed for analysis. Because in most cases grid software was developed and programmed by single people development was limited by time constraints and the necessary specialist competencies in different fields (constructivist psychology, mathematics/statistics, programming, layout/design). These problems increased with the development of programming languages and changes of operating systems. So many programs were not developed further and simply disappeared.

Usability: Some of the early programs were developed for special research purposes of the authors of these programs and often by themselves. So early programs typically had no documentation and all sorts of bugs. This was and is still perfectly all right for researchers with special needs and some tolerance for frustration but not for a larger group of users. So there may be lots of special programs which can only be handled by their developers.

Quality: In many cases it is not enough to list the functions of a program. Where one program just offers a command line input another one guides the user through a menu assisted by context sensitive help. Where one program offers a fixed and barely readable graph another one offers several views of the graph, filter, zoom, and screen shot facilities. Etc.
To describe these differences in detail is not possible. The best advice is to try out the programs one is interested in for functions, usability and quality – which is possible at least with trial versions of most of the currently available programs. Practical testing will show whether a program not only has certain functions but also whether one is able to find them and use them without nervous breakdown.

Control of information: Companies which use and develop grid software do not publish details about their tools for obvious reasons. The same is true for some developers who use their software in business consulting. In these cases the internal mechanics of the software are kept secret, sometimes even the data format.

Language: There are programs in languages other than English (e.g. German). To find and to describe them is difficult and not of prime interest here as long as no English version exists.

Multitude and heterogeneity of functions: Even for one of the programs a description and discussion of all main functions would fill quite a number of pages. These would add up considerably if special uses of these functions and unique functions of this program should also be described and reviewed in a way which would be understandable for someone not familiar with this program or grid software at all.

So what is possible here is an overview over currently available ‘matured’ programs with a reasonable range of functions which are (at least) in English.

If there should be others (e.g. research or specialised versions or programs in other languages) which could be of interest for a wider user group we would like to consider a review of these at a later time.

A description and review of the currently available software has to use criteria. These criteria will not be identical with those the creators of the software or experienced users of the software would like to be used. So the descriptions and reviews started here will profit from comments and additional information from creators and users - and will help people who are interested in a specific program to connect.

A preliminary overview of the software which will be presented and discussed here in the time to come:

	OS	Elicitation	Analysis	Author - Email/Website
EnquireWithin	Win	X	X	Mayes,John/Stewart,Valerie http://www.enquirewithin.co.nz/
GridCor	Win		X	Comejo,José Manuel/Feixas,Guillem http://www.terapiacognitiva.net/record/gridcor.htm
GridStat/ GridScal	DOS		X	Bell,Richard http://www.pcp-net.de/info/ index.html
Idiogrid	Win	X	X	Grice,James W. http://www.idiogrid.com/
Flexigrid	DOS	X	X	Tschudi, Finn finn [dot] tschudipsykologi [dot] uio [dot] no
Rep IV	Win/ MacOS	X	X	Shaw,Mildred/Gaines,Brain http://repgrid.com/repgrid/
WebGrid	Online	X	X	Shaw,Mildred/Gaines,Brain http://repgrid.com/WebGrid/
GridSuite	Win/ MacOS	X	X	Fromm,Martin/Bacher,Andreas http://www.gridsuite.de

More information is available at: http://www.pcp-net.de/info/comp-prog.html

The grouping of programs used here is, as stated before, a tentative one. The sequence starts with EnquireWithin the author of which, John Mayes, stresses the focus of this program on elicitation. Although cluster analysis is performed this is not the main characteristic of this program.

The three following ones are on the other side focussed on analysis where the number of available functions increases within this sequence. Idiogrid also includes (several) basic functions to elicit grids but the strengths of this program are clearly in the analysis section.

The third group is made up of programs with a more balanced offer of elicitation and analysis options. Although they already contain several options for analysis they fall short of what GridStat and especially Idiogrid offer. In this group according to functions Flexigrid and RepGrid are comparable in many respects – where RepGrid up to the latest version was only available for MacOs and Flexigrid only for DOS – but with some differences in usability due to the different operating systems. GridSuite historically is the latest in this series, offers most of the options of Flexigrid and RepGrid and includes some additional functions.

References

Bell, R. (2000): Why do statistics with Repertory Grids? In: Scheer, J. W. (ed.): The Person in Society. Challenges to a Constructivist Theory, Gießen (Psychosozial), p. 124-133.

Fransella, F./Bell, R./Bannister, D. (2004): A Manual for Repertory Grid Technique. 2. ed. Chichester (Wiley).

Jankowicz, D./Thomas, L. F. (1982): An Algorithm for the Cluster Analysis of Repertory Grids in Human Resource Development. In: Personell Review, No. 4 (Vol. 11), p. 15-22.

Kelly, G. A. (1991): The Psychology of Personal Constructs. Volume One - A Theory of Personality. London/New York (Routledge).

Neimeyer, R. A./Baker, K. D./Neimeyer, G. J. (1990): The Current Status of Personal Construct Theory: Some Scientometric Data. In: Neimeyer, G. J./Neimeyer, R. A. (Eds.): Advances in Personal Construct Psychology Vol. 1, Greenwich(Conn.)/London (Jai), pp. 3-22

Ravenette, T. (2000): Do we need grids and stats? - a contrary view. In: Scheer, J. (ed.): The Person in Society, Gießen (Psychosozial), p. 134-141.

Sewell, K. W./Adams-Webber, J./Mitterer, J. et al. (1992): Computerized Repertory Grid Technique: Review of the Literature. In: International Journal of Personal Construct Psychology, p. 1-23.

Shaw, M.L.G. (1980) On Becoming a Personal Scientist. London: Academic Press.

Slater, P. (1964): The Principal Components of a Repertory Grid. London (Vincent Andrew).

Slater, P. (1976): The Measurement of Intrapersonal Space by Grid Technique. Vol 1: Explorations of Intrapersonal Space. London (Wiley).

Slater, P. (1977): The Measurement of Intrapersonal Space by Grid Technique. Vol. 2: Dimensions of Interpersonal Space. London (Wiley).