Understanding story completion
What is story completion? Story completion, involves participants telling a story, or stories, in response to a pre-determined ‘stem’ or ‘cue’ they have been given. The story stem consists of an opening sentence, or several opening sentences, created by the researcher, and usually presents a hypothetical scenario involving one or several characters. The participant is then asked to continue or complete the story, either unconstrained or following some guidelines. Most qualitative story completion to date involves written story-telling (but see here for a paper that combines written story-telling and visual methods).
Story completion offers a radically different way of collecting data from human participants for qualitative research, which has typically involved asking people to speak or write about what they think, feel, believe, know and do (as in an interview, focus group, qualitative survey or diary study). This is one of the reasons why we are so passionate about the use of story completion as a qualitative method, because it expands the possibilities for qualitative research, and the sorts of information we can get access to.
“David has decided to start removing his body hair….”
This stem generated great data! Stems do not have to be that short and simple – for some projects, it might be necessary to use a longer, more detailed stem to generate the type of responses you seek. Here’s an example a two-sentence stem, from Irmgard’s research on weight loss motivations:
“Catherine has decided that she needs to lose weight. Full of enthusiasm, and in order to prevent her from changing her mind, she is telling her friends in the pub about her plans and motivations.”
Stems are typically followed by completion instructions, which can be as simple as asking ‘what happens next?’ or they might specify a time frame for the story, and/or particular things the story should address.
These two stems are both examples of third-person stems – the participants are invited to write about the characters David or Catherine, rather than imagine themselves as the character in the story (that would be a first-person completion). Third-person stories are the most common type of story completion, but some researchers have used first-person completions, particularly earlier in the history of story completion research. Nowadays first-person completions are considered somewhat problematic, because they require the researcher to make the interpretative leap that the stories represent how the participant thinks and feels about something, without asking them directly how they think and feel. First-person completions reflect the legacy of story completion as a clinical assessment tool – so let’s explore that history briefly.
A brief history of story completion as a clinical assessment tool and a quantitative method
Story completion has a relatively long history – it was first developed as a projective technique (see Rabin and Zlotogorski (1981) in our reading list). If you’re a psychologist, you will probably be familiar with the most famous projective technique, the Rorschach ink blot test. Projective techniques were developed for use in clinical practice and assessment by psychoanalytically informed clinicians and researchers. The assumption was that ambiguous stimuli like ink blots and incomplete stories provided a way of accessing patient’s unconscious beliefs and motivations – things they weren’t consciously aware of or things they were uncomfortable speaking about directly. Because of the stimulus was ambiguous patients were assumed to ‘project’ (hence projective technique) their unconscious thoughts and feelings into the stimulus to make sense of it. There have been some fairly dubious uses of projectives over the years (e.g. to ‘diagnose’ homosexuality and lesbianism in the middle of the last century) and some critique of what they access. But they are still valued in child developmental research on attachment. When story completion is used as a projective technique, the focus is on the psychological meanings presumed to sit behind (or ‘underneath’) the stories, rather than the stories per se.
Story completion research on child development and attachment has predominantly been quantitative. In this research, the stories are coded by multiple researchers using a structured codebook, the level of ‘agreement’ between coders is checked to ensure the coding is reliable and then the coded stories are transformed into numerical data for the purposes of statistical analysis. Again, the focus is on the psychological meanings presumed to sit behind the stories and are presumed to explain why the participant responded to the story stem in the way that they did, rather than on the rich, narrative detail of the stories.
The first uses of story completion in qualitative and mixed methods research are best thought of as ‘qualitative-ish’ rather than fully qualitative enactments of story completion. One particularly controversial early story completion study was conducted by the feminist psychologist Matina Horner – this study has fuelled scepticism about story completion as a qualitative technique for some. Horner used story completion to research women and men’s feelings about success and her psychoanalytically-informed interpretation of her data concluded that women have a ‘fear of success’. Her findings were seen by some as essentially blaming women for their lack of achievement in the workplace, rather than centring the sexist and misogynist social context. The generalisability of her findings was also disputed. For some, story completion became associated with ‘over interpretation’ of findings and there has been little use of the story completion method until very recently – see Clarke et al. (2019) for more discussion of this history.
The paper that inspired us to use story completion was published in 1995 by two feminist psychologists – Celia Kitzinger and Deborah Powell – and it was crucial to the development of story completion as a (fully) qualitative research technique…
The Kitzinger and Powell study | reimagining story completion as a qualitative method
Kitzinger and Powell used the story completion method to explore meaning making around infidelity in heterosexual relationships. Both their conceptualisation of story completion and their study design (in particular, their use of a two-stem comparative design) have been hugely influential. They wrenched story completion free of its psychoanalytic moorings and argued that story completion, like most research methods and techniques, could be used within different theoretical frameworks; story completion wasn’t a necessarily psychoanalytic or essentialist method. Researchers were not obliged to interpret stories as reflecting psychological meanings (but they could still do this). Instead stories could also be interpreted as reflecting social meanings – the meanings available to participants in the wider social contexts. Furthermore, they argued that if different participant groups made sense of the same scenario in different ways, this might tell us something about their social positioning, and the different social meanings available to them.
They also saw the potential for using story completion in a comparative design – something typically associated with quantitative rather than qualitative research. By varying some details of the story stem (e.g. by changing the sex/gender of the central character) it was possible to explore any differences in how participants made sense of a woman and a man in the same scenario. They also chose to compare the responses of female and male participants to the same stem. Hence, there design had two (intersecting) levels of comparison:
- Comparing responses to female and male characters in the same scenario
- Comparing the responses of female and male participants to the same scenario
In keeping with their revisioning of story completion as a theoretically flexible technique they argued that these differences (if present in a data-set) could reflect either psychological differences between women and men or social differences (e.g. differences in socialisation).
Their stems retained the ambiguity associated with projectives, but deliberate ambiguity wasn’t used to tap into unconscious feelings and motivations. Instead, it was used to explore the assumptions participants made in responding to the stems and to permit a diversity of responses. Their stem design also taught us that for story completion to work well, there has to be some tension or provocation in the scenario presented or the potential for a diversity of responses. Scenarios with very predictable or obvious outcomes can fall rather flat, with short, thin stems and very similar narrative features and resolutions. This latter point highlights another important feature of story completions – they are stories (or are ideally stories, some participants may ‘refuse’ the task and not write a story but respond to the stem as if it was a question)! Within Western (and non-Western) cultures stories have particular structures and features – very simply, stories often have a beginning, middle and an end (or resolution), endings can sometimes have a moral dimension, and stories often have a protagonist or central character as well as other secondary or peripheral characters. What this means is that, analytically it can be useful to focus on some of the storied dimensions of the data. This could be as simple as considering patterning in how the central character is depicted or as all-encompassing as using the ‘story mapping’ outlined by Virginia and Victoria in their book Successful Qualitative Research and developed in our analysis of the male body hair data.
Story completion as a qualitative technique
Excitingly, there is growing interest in story completion as a qualitative technique (and even a post-qualitative or post-humanist technique). We have been working hard to explore, develop and ‘promote’ the technique over the last few years, so we’re thrilled to see this burgeoning enthusiasm for story completion. There is a small but building body of literature with authors across different fields exploring what story completion can offer – see our resources page. We want to emphasise that story completion is still new, there is still a lot we have to learn about what does and doesn’t work in story completion research, and we invite you to join us on this journey of discovery!