Experimenting with fiction: bibliotherapy

Hello, my name is Rocío Riestra-Camacho. I am a PhD candidate at the University of Oviedo and in this video presentation, you will be able to learn about the preliminary results of a reading experiment I conducted in collaboration with Emily Troscianko at The Oxford Research Centre in the Humanities (TORCH), from September to November 2019, and which was finished at Oviedo in March 2020. If you prefer the written version, you will find it below, but I’d strongly encourage you to watch the video as well!



What was the hypothesis?

This reading experiment evaluated the potential of fiction in the context of anorexia bibliotherapy. In it, participants read two novels to see whether their attitudes towards food, exercising, and the body improved. These were Breathe, Annie, Breathe and Coming Up for Air, two young adult sports novels written by Miranda Kenneally in 2014 and 2017, respectively. The hypothesis was that reading sports fiction with a reading guide would reduce participants’ vulnerability to an eating disorder more than in the case of a group receiving no guidance. Changes in vulnerability were measured using before-and-after questionnaires.

In a nutshell, the reading guides consisted of pop-up messages placed at the margin of the books, encouraging readers to adopt the healthy mindset which the protagonists of the novels display. For example, a pop-up message might suggest that readers emulate the characters’ self-care behaviours of taking rest periods from exercise, which is something that can be challenging to do for someone over-concerned with losing weight through exercise.

What was the procedure?

Initially, participants completed a health questionnaire, which contained questions from the EAT-26, a standardized psychometric test to measure vulnerability to eating disorders, and some tailored questions I included to evaluate concerns about body image, eating, and exercise as directly targeted by the texts and the reading guide. Then, participants read the two novels, and were randomly allocated to either the experimental group (reading the novels together with the reading guides) or to the control group (reading the novels as published). Afterwards, participants repeated the questionnaire and took part in a 15-minute Skype interview to discuss their participation in the study, which I conducted and transcribed. The aim was to evaluate differences between the pre- and post-questionnaires, to see whether participants’ health perceptions had changed after reading these novels with or without guidance, and thus to evaluate the bibliotherapeutic effect of the guides.

Preliminary findings

65 participants completed the experiment. None of them had an active diagnosis of an eating disorder. Initial analysis provided evidence that partially supports the hypothesis: on average the scoring in the experimental group showed a trend towards decreasing (though the trend was not statistically significant), indicating that readers who received guidance may have slightly improved in their health perceptions. However, and contrary to the initial hypothesis, the average scoring of the control group showed a tendency to increase (also not reaching statistical significance). This opens up the possibility that the novels, without guidance, may have an iatrogenic (i.e. unintendedly harmful) effect on participants’ vulnerability as regards their bodily health. In this connection, you may find it of interest to read about what Emily Troscianko found in a previous survey study on reading habits and mental health, with a focus on eating disorders.

Digging further into the data, an analysis of variance (ANOVA) performed on specific questionnaire variables revealed interesting results. Regarding the EAT-26, which specifically measures oral control, dieting and bulimia, and food preoccupation, scoring increased in both groups, although more markedly in the control group. Again, none of these were statistically significant results, however. The tailored questionnaire measured perceptions and behaviours about food (guilty eating, anxiety over food control, eating inflexibility, food anxiety, calorie control, excuses and dieting, and dieting saliency), exercise (weight-loss oriented exercise, performance oriented exercise, and exercise inflexibility) and the body (body perception and gender stereotypes). The ANOVA revealed a statistically significant result only in the gender stereotypes variable: specifically, it showed a statistically significant decrease in beliefs about gender stereotypes about the body in the experimental group. This result could feasibly speak of the bibliotherapeutic value of the reading guides, at least regarding stereotypical beliefs regarding gender presentation (e.g. the belief that skinniness is linked to femininity). Finally, the data were analysed in more depth by James Carney.

As for the qualitative data generated by the interviews: a first reading of the scripts yielded an unexpected effect. In particular, of 49 interviewees, 22 noted that reading the novels prompted them to exercise or at least encouraged them to seriously consider doing so. This may appear like a self-evident result, but it was neither intended nor specifically tested for. Upon later reflection, it seems to add evidence to some narrative persuasion studies in health, which present similar results. One difference, however, is the higher level of ecological validity attained by this reading experiment. Asking readers to read two novels in full—rather than shorter texts created ad hoc for laboratory testing, as is the case in narrative persuasion experiments—is closer to reproducing actual experiences of readers “in the wild”. In fact, to the best of my knowledge, this is the first experiment in cognitive literary studies or the health humanities which has evaluated the effects of reading not only one but two novels in full, and has done so using a control group to test the effects of a textual variable.

Another exciting result stemming from the interviews is that some participants in the experimental condition reported acquiring a more relaxed stance towards eating and exercising than previously, which was one of the aims of the reading guides, while those in the control group did not. This would partially account for the bibliotherapeutic/iatrogenic results suggested by evidence from the questionnaires, although closer critical discourse analysis of the transcripts is still needed.

Conclusion: new avenues of research

There is an obvious need for research in the humanities to become increasingly interdisciplinary, and researchers with humanities backgrounds to take up a role in the ethics of scientific and cultural production sooner rather than later. The benefits of engaging with literature have been claimed for centuries, as a quick search into the vast field of “bibliotherapy” will yield. And yet, we may benefit from dropping long-established assumptions and starting to conduct reading experiments to find out whether it really does have beneficial effects, or harmful ones. Our assertions remain just fictions otherwise.


Rocío Riestra-Camacho is currently completing her PhD thesis in Gender Studies at the Department of English of the University of Oviedo in Spain, in combination with a part-time degree in Psychology at the National Distance Education University.

Thank you very much for your interest in reading about this work. For correspondence, you can email me at riestrarocio@uniovi.es, and Emily Troscianko at emily.troscianko@humanities.ox.ac.uk.