Contact Us

An Overview of Enquire Within


An overview about how the Enquire Within interviewing software application implements the Repertory Grid interview an interviewing technique invented over 50 years ago by George Kelly, an American engineer who became a clinical psychologist.


Using Enquire Within and Repertory Grid

The Enquire Within software application puts Repertory Grid onto an interactive Windows-based platform. Repertory Grid is an interviewing technique invented over 50 years ago by George Kelly, an American engineer who became a clinical psychologist.

Kelly was annoyed with the lack of precision in psychological techniques and theories. By inventing Grid, he produced a technique with the following unique combination of features:

  • it is totally free from interviewer bias (which is why it can be put onto a computer platform)
  • it allows for precise statistical measurements of the perceptions of an individual person
  • it is person-centred rather than expert-centred – that is, he believed that most people can take responsibility for themselves, and that the role of the therapist or other expert is primarily to act as a skilled mirror

Kelly’s work attracted only a limited number of adherents. He was not a self-publicist; his own writing is almost incomprehensible; and in order to give Grid its full power it needed computational techniques that were simply not available at the time. Enquire Within is the first program to put the entire Grid process onto a computer platform. In doing so, we think we have made a qualitative difference to what can be done with it.

Some of the history can be found at Background and Theory. What follows is a description of the basics of Repertory Grid and Enquire Within.

  1. A session can be done about any subject of which the user has some experience.
  2. Repertory Grid/Enquire Within has been described as a powerful, empty procedure. It is pure structure, content-free. Therefore each session has to be configured to meet its particular purpose.
  3. All sessions have to have a purpose. Purposes can range from extractive (e.g. market research) to reflective (e.g. situations where I learned something). They can be concerned with subjective or objective data. All components of the session, and the choice of analysis, are related to the purpose.
  4. The components of a session are: elements, constructs, and qualifiers.
    • Elements are concrete examples of the domain you wish to explore: A political scientist doing his small-scale scenario planning exercise actually set up two sessions: one had Ministers as elements, and one used government departments.
    • Constructs are bipolar scales on which the user judges the elements. Constructs are elicited by taking the elements in groups of three and asking the question ‘What have any two of these have in common which makes them different from the third?’ This process is the basis for Grid’s claim to be free from interviewer bias, because the interviewer does not suggest the content. The fact that the constructs are bipolar reflects the fact that all evaluations are scalar: good only makes sense if we can contrast it with bad.
    • Qualifiers are questions starting with ‘... in terms of’. They focus the process of eliciting constructs, by referring back to the purpose. So the political scientist takes his Ministers in groups of three and asks how two are similar and different from the third ‘in terms of their effectiveness ... in terms of their policy skills ... in terms of their political objectives,’ etc. Qualifiers such as ‘in terms of their family life ... their cultural interests,’ etc. would move the session away from the purpose.
  5. The processes in a session are:
    • Eliciting elements: these can be provided in advance or created as the result of questions.
    • Eliciting constructs: this is done by the two-against-one process described above.
    • Laddering up and down: laddering up asks for the preferred pole of the construct, in term of the purpose, and the reason. Laddering up takes you towards ‘core constructs’ – the user’s deeply-held values and beliefs. Laddering down asks for more specific examples of how the poles of the constructs differ, thereby adding clarity to the meaning of abstract constructs.
    • Rating elements on constructs: when there are enough constructs (six as a minimum) the constructs are turned into a scale (usually five points) and the user rates each element on each construct. This produces a matrix which can then be analysed statistically.
    • Analysis: the statistical analysis searches for elements which are closely correlated. It re-sorts the Grid to place the elements in families of closely correlated elements, and draws ‘trees’ above them to indicate the level of correlation. It then does the same thing for the constructs, so you have a visual presentation of the relationships between the elements, and between the constructs.
    • Differentiation: this is the point at which the user starts to refine and clarify the picture so far. At the user’s choice, Enquire Within will present the closely correlated elements. It asks a question, whose basis is: given that you perceive these two elements as so similar, (a) do they really mean the same thing to you, or (b) is there more difference between them than has been captured so far? If the answer is (b), then Enquire Within asks the user for a new construct on which one element will be rated at one extreme and the other element at the other. The new construct is then used to rate all the elements and the Grid is re-calculated. The same process happens with the constructs, but with a slightly different range of choices: (a) are these constructs so similar that you want to amalgamate them into a new one, or (b) are they more different than they seem, in which case please supply a new element which will ‘split’ the construct as before, or (c) is this an important insight which you want to leave in place.
  6. The purpose of the differentiation process is to explore meanings, achieve clarity, and gain insights. For example, a lawyer studying miscarriages of justice could be presented with the fact that two elements – GUILDFORD FOUR and BIRMINGHAM SIX – are correlated at the 95% level; in other words, she sees them as very similar indeed. She can accept this, or she can say ‘No, there are more differences, I just haven’t expressed them yet,’ and she searches for some ways in which the two cases are different). A person examining her learning styles may find a very close correlation between two constructs – competent teacher - incompetent teacher and learning about the big picture - learning about procedures. In other words, when she was learning about the big picture she was much more likely to judge the teacher as competent, and when she was learning about procedures she was much more likely to judge the teacher as incompetent. She could not think of an example that would split these constructs; she had instead to accept that this was an important insight into her own prejudices which she had better think about - an uncomfortable but inescapable lesson, and one which she would have to work on.
  7. Experimenting with scenarios: once the analysis graph has been developed in whole or in part, the user can use it to experiment. For example, our political scientist into scenario planning could simply delete a Minister or two from the analysis and see what happened. He can add a new one and incorporate him into the analysis. He could amalgamate two departments, or introduce a new one. He could change someone’s political orientation (not for the first time). And so on.

Where's the Substance?

Enquire Within is an information-rich process, with very little redundancy. The account above of its components and its processes may appear very heavy and complex. In fact when you do one – the tutorials guide you through and make it much easier than it appears – the concepts are introduced through practice and reflection. However it’s worth pointing out that, depending on your purpose, the part(s) of the process which are most rich in information will vary. You may not need to go all the way for some purposes. Examples of where useful data could be found:

  • Names/descriptions of elements, if the interviewee supplies them. We interviewed 200 doctors about why they chose their speciality, using the same protocol for eliciting elements - most preferred, least preferred, etc. On a sample of 200 doctors, that information is useful in itself.
  • Unrehearsed element sets. Most people have never formally asked themselves to name a set of situations in which they were trying to learn something. The process of asking oneself these questions is valuable.
  • Count of elements. The example of evaluating a course on psychological tests begins by asking how many tests the person could think of. Conclusions can be drawn from that, especially if as a result of training the number increases.
  • Count of constructs. The same example – evaluating a course on psychological tests – applies. A small number of constructs indicates a lack of reflection on the subject-matter.
  • Content analysis of constructs. The nature, and relative proportions, of construct families can be useful. For example, we interviewed 200 middle managers in an oil company in the mid-1980s. When we came to analyse the data I put a fiver on the table for the first set of interview records which mentioned customers. It went unclaimed.
  • Within the Differentiation process, sometimes all you need to do is compare two or three elements, or two or three constructs.

Basic Design of Enquire Within

The program has been designed to be as intuitive and flexible as possible. The user can prepare and store Resource Files in advance (i.e. elements, element creation questions, purposes, qualifiers, and complete sessions). Alternatively a session can be assembled on the screen.

The triadic comparison process for generating constructs has several options: the program will give you triads at random, or you can select your own. Alternatively you can call up a screen which allows you to pick up the elements and move them around in relation to one another.

Help systems are available from the menu, or context-sensitive help on the screen. Options for reconsidering and re-writing are available wherever you need them, and options for deleting sensitive information once it has yielded its value.

There is a ‘scratch pad’ incorporated for making notes etc.

The statistical matrix produced after rating can be output to other matrix analysis programs.


Related Resources

Suggestions

Search for results from the Enquire Within site.