Questions about story completion?

Have you got questions about story completion? We hope so! It’s an evolving method, and we encourage you to read widely and explored the approach… Here are a few questions we have encountered, and ‘answers’ to them.

What kind of research questions is story completion suitable for?

We think story completion is best suited to questions about social meanings and meaning-making around a topic. It’s less suited to questions about personal experiences and views because of the indirect mode of data collection. To use story completion to research personal meanings you would need to make the interpretative leap that the stories people write are a straightforward reflection of their thoughts and feelings about a topic.

How many completions do I need?

You know what we’re going to say, right? It depends! Of course it does… It always depends in qualitative research. It depends on the scope of the project (e.g. small student project or publish-able study), the number of stems, the number of stems each participant is asked to complete, and other less knowable in advance factors – such as the richness and complexity of the stories, and the diversity of the stories (if the stories are very diverse a larger sample may be needed to identify patterning in the data). There is no clear agreement in the published story completion literature with regard to sample size, with a wide range of sample sizes reported (from as few as 20 to more than 200 or even 1,000). Twenty to thirty rich and complex stories are likely to provide an appropriate data-set for a small one-stem-design student project. For a publish-able study expect to collect at least 100 to 200 completions for a one-stem-design project. We discuss this a bit further in this paper.

Isn’t comparison problematic in qualitative research?

Many qualitative researchers would say yes to this, but we don’t – well not for story completion at least. We don’t think it’s problematic for two main reasons: the ‘set of conditions’ for comparison is otherwise similar, but more importantly, because we’re exploring social-level meaning, not looking at group difference and therefore interpreting difference as revealing something fundamentally or essentially different about X or Y group. So, for instance, if there are differences told in stories about a male or a female protagonist, we’d be interested in what that says about gendered meaning-making.

What do I do with fantasy stories (when I haven’t sought them)?

Fantasy stories are not unusual. And sometimes they are effectively a ‘refusal’ of the task, whereas other times, they transpose the topic into the ‘fantasy’ domain… So, it depends on whether the fantasy content has relevance to your research question and focus. If it does, they do, and then you might incorporate them into your analysis (for instance see this paper).

How should I treat the data generated through story completion?

Through the history and use of story completion, data have been treated in different ways – as: 1) giving access to inner psychology, such as ‘the unconscious’ or ‘attachment’; 2) as giving access to the attitudes, values and perceptions of the story writers; and 3) as a way of accessing socially-available meanings about the topic, in the worlds of the story writers. This third way is how many qualitative scholars who have used the method, since Kitzinger and Powell’s ground-breaking paper outlining this approach. This treats the method as a way to gain access to the ‘out there in the world stuff’ rather than the ‘in there in people’s heads stuff’. Of course, many would also deconstruct a clear separation between the two! What is important is that you have an understanding of what you claim your data can give you access to, and a foundation for why that is.

Is story completion a good method for student projects?

Yes, yes, a thousand times yes! Our enthusiastic response tells you that we think story completion is a great method for time and resource limited student projects. Data collection can be very quick and efficient, especially when using online completion (e.g. using online survey software) or going into a setting armed with hard copies and lots of pens where there is a large group of your target population with time to participate in a study (e.g. students in a university lecture). The indirect mode of data collection and limited interaction between research and participant facilitates exploring sensitive topics that might raise tricky ethical questions if an inexperienced student proposed to address them more directly (e.g. asking people’s experience through an interview). Indeed, we first recommended the use of story completion in an undergraduate student project when the student wanted to research experiences of domestic violence – a topic we felt strongly wouldn’t (and shouldn’t) receive ethical approval in the context of the proposed study. Instead, the student successfully (and ethically!) used story completion to explore how people made sense of domestic violence in same-sex and different-sex relationships, identifying some interesting similarities and notable differences in sense-making.

Does story completion research have to be social constructionist?

Most of us have used story completion within constructionist frameworks, but it doesn’t have to be (visit our resources page). Qualitative story completion can be essentialist, it can focus on the psychological meanings presumed to lie behind the story and motivate the way it is told, but this requires the interpretative leap that personal feelings and motivations directly inform story writing, and so needs to be well thought through. Story completion can also be critical realist or contextualist – a focus on social meanings doesn’t require a constructionist orientation. For an example of a critical realist story completion study see this paper.