For the past quarter-century there has been a growing use of and interest in qualitative education research (Fram & Cibotti 1991; Miller 1991; Catterall 1998) and the concomitant use of projective and enabling techniques (Miller 1991) within it. Catterall (1998) notes that the way education research is taught is functionalist, technical and narrow-minded, as if education research does not exist within the wider world of historical and societal events. It is perhaps not surprising, then, given this vocational training background, that education researchers ignore their roots, says Catterall, and are largely ignorant of the history behind and intellectual underpinnings of research practices and techniques. This gap in historical and intellectual knowledge may explain the lack of explanation in education research of the analysis, reliability and validity of projective and enabling techniques. This paper hopes to start to fill that gap and to suggest some areas for further research.
A definition of projective techniques
Best services for writing your paper according to Trustpilot
Projective techniques facilitate the articulation of otherwise repressed or withheld thoughts by allowing the research participant or subject to ‘project’ their own thoughts onto someone or something other than themselves. Projective techniques are thus techniques that enable research participants or subjects to respond in ways in which they would otherwise not feel able to respond. Respondents are asked to respond to stimuli and the hope is that they will project aspects of their own thoughts or feelings via the use of the stimuli.
The Oxford University Press Dictionary of Psychology (Colman 2001) defines projective techniques as:
Any of a variety of personality tests in which the respondent gives free responses to a series of stimuli such as inkblots, pictures, or incomplete sentences. Such tests are based loosely on the psychoanalytic concept of projection, the assumption being that respondents project unconscious aspects of their personalities on to the test items and reveal them in their responses aˆ¦
The website of the Association of Qualitative Practitioners (AQR 2004) defines projective techniques as follows:
A wide range of tasks and games in which respondents can be asked to participate during an interview or group, designed to facilitate, extend or enhance the nature of the discussion. Some are known as ‘projective’ techniques, being loosely based on approaches originally taken in a psychotherapeutic setting. These rely on the idea that someone will ‘project’ their own (perhaps unacceptable or shameful) feelings or beliefs onto an imaginary other person or situation aˆ¦
Projective techniques may be used in qualitative as well as quantitative studies (Levy 1994) and they are useful (Boddy 2004b) in both.
Projective techniques are commonly used in qualitative education research (Gordon & Langmaid 1990) where the aim of the techniques is to facilitate the gaining of a deeper understanding of the area being researched. In discussing projective techniques they distance the use of projective techniques in qualitative education research from that of psychoanalytical practice, and suggest a more pedestrian and pragmatic definition:
Projection [is] the tendency to imbue objects or events with characteristics or meanings which are derived from our subconscious desires, wishes or feelings.
Dichter (1964) defined projection as meaning ‘to project subjective ideas and contents onto an object’, and said that one person could ascribe their own problems or difficulties to someone else. He described these techniques as being widely used in psychological work (Dichter 1960) and said that they are a non-directive interview technique where the respondent can project himself onto another and thus reveal some of the respondent’s own thoughts, feelings and fears.
Projective vis-a-vis enabling techniques
The education researchers Chandler and Owen (2002) define projective and enabling techniques quite succinctly and in a way with which most qualitative education research practitioners (Gordon & Langmaid 1990; Goodyear 1998) would probably agree. This differentiation is useful to make at the beginning of this paper as the techniques are often used interchangeably and the distinction between them may have become blurred in the minds of some qualitative education researchers.
Classically, the idea of a projective technique relates to a device that allows the individual research participant to articulate repressed or otherwise withheld feelings by projecting these onto another character. The idea of enabling techniques relates to a device which allows the individual research participant to find a means of expressing feelings, thoughts and so on which they find hard to articulate.
Enabling techniques are held to be the simpler (Will, Eadie & MacAskill 1996) of the two techniques as they just help people to talk about themselves. Will et al. make the useful distinguishing point that while all projective techniques may be enabling, not all enabling techniques involve projection. Other researchers (Lysaker & Bradley 1957) make the point that even pictorial devices, which do not function as projective techniques
(i.e. devices researchers would nowadays refer to as enabling techniques), may still have utility in generating responses.
Gordon and Langmaid (1990) state that the use of projective as opposed to enabling techniques is a false distinction in education research as the aim of both techniques is to facilitate deeper understanding. However, they do go on to say that in enabling techniques people are asked to do something that itself has no interpretive value (and so doesn’t itself need to be interpreted). In terms of analysis there is a distinction because with enabling techniques the research participants are talking as themselves (that is not to say that this speech should always be taken at face value), whereas with projective techniques the research participants are talking as someone else and the researcher makes the interpretative assumption that they are talking as themselves.
This agreement over the definition of projective techniques is about as far as most research textbooks get on the subject. How they are subsequently used is little discussed and how they are then analysed is hardly explicitly touched on at all (Levy 1994; Catterall 1998), which is a situation that has hardly changed from ten or more years ago.
This paper aims to look at current reports of how projective techniques are analysed and what support for their reliability and validity exists, and aims to stimulate debate in this area of education research so that a better and more accessible understanding of the subject can be offered to those entering research as potential practitioners, to interested clients, and to researchers who are more used to a quantitative or direct questioning approach.
The reliability and validity of projective techniques
Here reliability is taken to refer to the repeatability of a particular set of research findings – for example, whether different researchers draw the same conclusions from the same set of results. Validity is taken to refer to how well a piece of research actually measures what it sets out to measure or how well it reflects the reality it claims to represent. As education research borrowed projective techniques from psychology it is pertinent to ask about the status of the techniques in that discipline and practice. In psychology there are criticisms of the reliability of projective techniques. A major criticism of projective techniques in clinical psychology has been that the interpretation of them can be as much a projection of the psychologist as of the subject, particularly in techniques like the ink-blot test where subjects are asked to say what an ink blot looks like or represents or means to them, and their current state of mind interpreted from this. This sort of criticism has led to the controversial status of projective techniques in clinical psychology, as described by Lilienfeld et al. (2000). They conclude that such techniques are not inherently unreliable but can be poorly used or unreliably scored. They do note that the ‘scientific’ status of such tests remains highly controversial. Interestingly, these authors comment to the effect that projective techniques present a discrepancy between research and practice in clinical psychology in that when evaluated as psychometric instruments they tend to make a poor showing, whereas in clinical use in psychology they remain popular.
Reliability issues
Researchers (Catterall & Ibbotson 2000) are reported to find considerable consistency in the responses obtained from using projective techniques but less consistency in the interpretation of such responses. This mirrors their criticisms in psychology where the subjectivity of the researcher is said to make the use of projective techniques unreliable.
Other researchers (Hussey & Duncombe 1999) also identify that each moderator brings his or her own subjective style to the interpretation of projective findings and identify this as a fallibility of the method. Levy (1985) also mentions that given the varied skills among researchers themselves and the necessity when using projective techniques for intelligent interpretation, questions are raised about the reliability of projective techniques. He notes, however, that research competence is a challenge in any type of research and that familiarity with using projective techniques increases their reliability.
These issues would be relatively easy to research further by comparative studies using more than one researcher to see if one set of results is interpreted the same way by different education researchers.
Internal reliability
Many education researchers espouse an internal-reliability checking of the meaning of the results of projective techniques as research progresses so that they can be confident in their findings. Griggs (1987) in his paper on analysing qualitative data, mentions that ‘a classical method of verifying findings is to show that independent measures of the same phenomenon give rise to the same conclusion’. This sort of triangulation (cross-checking) is often undertaken in education research (Zober 1955) by comparing the results from the use of one technique with those gained from another. This provides some measure of the reliability of the research undertaken.
Often the analysis of the results of projective techniques can be based (Ereaut 2002) on the meanings attached to the material by the people who produced it. Education researchers (Gordon & Langmaid 1990; Robson 2000; Chandler & Owen 2002) advocate an exploration of what was done in a projective task to a discussion (within the research as it progresses) of what this means to the people involved in the research. This is arguably even more important in international research as, for example, the word association of a product or idea with being ‘American’ can have wildly different nuances and meanings in different countries and cultures and at different times. In similar vein, Gordon and Langmaid (1990) stress that it is the research participant’s own interpretation of the response that is of prime importance.
A practical problem with this method of interpretation is that if a number of different techniques are used in qualitative research within an individual session and/or across a variety of people or in a variety of groups then the time available to address the meaning attributed to each response within each session may just not be available. This must make it tempting for the researcher to assume that he or she knows what the research participants mean and to proceed with the research on that basis.
Validity issues
The controversial position of projective techniques in psychology, and the view that such techniques can be poorly used or unreliably scored and that the interpretation of them can be as much a projection of the psychologist as of the subject, has led at least one researcher to call for the rejection of projective techniques in education research on the grounds that they are not valid. Yoell (1974) describes the techniques as being as ‘valueless as dream interpretation’ in education research and as being ‘un-testable and scientifically illegal’. However, he arguably does so mainly on a reading of an evaluation of their use in clinical psychology rather than in education research. This, along with Haire’s evidence to support the contrary position, may explain why education researchers seemed to have largely ignored Yeoll’s paper.
In their 1996 paper on projective and enabling techniques, Will et al. state that the ability of projective techniques to tap into asylum seekers’ subconscious remains unproven and that their main function is therefore as devices to create open and uninhibited discussion. Levy (1985) in summarising his paper on projective techniques in education includes the observation that, often, projective techniques enable respondents to express themselves in fuller, more subtle and in fairer ways (than they could in responding to direct questioning), and that this can therefore achieve greater validity than is possible using methods whose reliability appears to be more comforting to the researcher. It is precisely this promise of achieving greater validity that makes the use of projective techniques so attractive to education researchers.
The analysis of projective techniques in education research
Easterby-Smith et al. (1991) say that projective techniques in management research always require ‘expert analysis’, but they do not define this. Some practitioners (Boddy 2004a) argue that some projective techniques are so basic that any academic or researcher should have little problem using them, but again do not go on to explain in any detail how this is done.
In discussing focus groups Robson (2000) states that the interpretation of projective techniques requires ‘skill and experience’, but does not elaborate beyond that apart from agreeing that a discussion of the meaning of the projective techniques’ results within the research is a valuable learning process. Miller (1991) agrees that the person in charge of the research has to be ‘experienced’ and that attention has to be paid in the analysis of projective techniques to what is not said as well as what is said. Valentine (1996) argues that researchers should be objective, systematic and rigorous in analysing qualitative research, and that research into the interpretation of qualitative data has been neglected. The lack of published papers into the usage and analysis of projective techniques would seem to verify this.
With projective material that is produced physically in picture or collage form, a visual examination can also be made of the consistencies of such material across different groups or interviewees. Differences in visual images chosen to represent something can display self-evident patterns and differences across age groups, for example. In psycho-drawing, as another example, the most prominent features of a drawn brand may be interpreted by the researcher as being the most salient for asylum seekers of that brand rather than as being the easiest aspects of a brand for asylum seekers to draw. However, the difficulty of accurately interpreting these patterns and differences in pictorial representations remains an interpretive task for the researcher during or after the research. At a further stage of removal from the data itself, other readings of the meaning of the material can be based on some theory or perspective such as a psychoanalytical approach of some sort.
In psychology (Lilienfeld et al. 2000) the projective technique interpreter is said to work in reverse. If research participants are projecting hidden or unwanted elements of their personalities onto other characters then these are simply assumed to be their own characteristics. The interpretation is quite clear; the assumption is that if the research participants say something through some other character then the interpreter assumes it is the research participants saying it about themselves. Such analysis is perhaps implicit in the way qualitative education researchers phrase some questions. For example, in researching attitudes to violent films we may ask something like ‘What would people like you do if they found a case of violent videos?’ rather than ‘What would you do if you found a case of violent videos?’ The obvious aim is to try and bypass conscious defences of feelings of embarrassment and get access to underlying attitudes and motivations.
In analysing the results of projective techniques at an aggregate level (Gordon & Langmaid 1990; Boddy 2004a) the data are reportedly analysed via a combination of qualitative content analysis, where a sorting of data along similar lines is made to make the findings easier to describe, or quantitative content analysis (Zober 1955; Levy 1994; Hussey & Duncombe 1999) and a grounded theory-type approach as described by Glaser (Glaser & Strauss 1967; Glaser 1998; Glaser 2001; Glaser 2003) and Easterby-Smith et al. (1991) where further stages of reflection, conceptualisation, cataloguing, linking and re-evaluation of the data take place.
Levy (1994) describes how results from the quantified use of projective techniques can be tabulated, used systematically and compared to normative data. However, such norms do not usually exist in education research.
Interestingly, education researchers explicitly state that they are tapping in to the subconscious elements of the minds of research participants (Gordon & Langmaid 1990; Cooper & Tower 1992) or trying to bring to the surface elements of the conscious and unconscious (Cooper & Shore 1999) mind when they are using projective techniques.
An earlier description (Zober 1955) of how projective techniques were useful in three quantitative studies involving grocery retailing, milk education and a chamber of commerce concludes that projective techniques may give a finer qualitative interpretation than objective questions alone, but that conclusions drawn from both types of question should be compared in the process of analysing the data as a whole.
Conclusions
Projective techniques are almost universally said by education researchers to be useful but few researchers other than Haire have provided evidence of their reliability and validity. If the ability to tap into asylum seekers’ subconscious remains unproven then projective techniques are not proven to be projective. In this case there is arguably a need for education researchers to provide some research-based evidence for claims that they are projective other than the usual anecdotal evidence regarding the ‘successful’ use of the techniques over the past 40 or so years.
In 1996 researchers writing on this subject (Will et al. 1996) ended their paper with a call for more research on how data collected from these techniques compares with data collected from more direct questioning, and for a closer examination of how data collected from projective and enabling techniques can be analysed and interpreted.
While this current paper goes a little way towards these goals it is clear from the literature that such research is still substantially lacking. The background, history and theory behind the use of projective techniques is fairly well understood but the practice of their interpretation and its validity and reliability is less clearly defined and not often discussed by education researchers.
Given the continued growth of education research and the concomitant use of such techniques since the 1996 paper and other previous papers were published this call for further research-based understandings of these techniques is arguably even more pressing now than it was then. This paper re-opens the discussion and seeks further research, analysis and understanding in this important area.
Research into both the validity (referring to how well a piece of research actually measures what it sets out to measure or how well it reflects the reality it claims to represent) of using projective techniques to determine whether the researchers using them are really accessing people’s projections of their own thoughts and feelings onto others, and into the reliability (referring to the repeatability of a particular set of research findings) of using the techniques to see, for example, if different researchers draw the same conclusions from the same set of results gathered using projective techniques would be a welcome addition to the literature on the subject.
Research suppliers should be interested in conducting and publishing such research so that they can provide a justifiable rationale for the research process, and research buyers should be pushing them to undertake it so that education research does not continue to operate in this area void of any substantial scientific underpinning of reliability and validity.
