Blog post by Prof. G. Neil Martin

All horror, said Dario Argento, is psychology. I am a psychologist. I am also a horror fan. In 2017, these two factors dovetailed. Making my usual annual August soujourn to Frightfest, the UK’s premier horror film festival, I began wondering: Why do people go to this sort of thing and what effect - what reliable effect - does watching fictional horror have? What does the psychological literature say, if anything? And why are there never festivals devoted to romantic comedy?

The cogitation led to what became the first review of the psychological effects of, and responses to, horror. It even started with a parenthetical title - (Why) do you like scary movies? - which suggests either a stylistic leitmotif or plummeting appetite for creative punctuation (Martin, 2019). The review analysed all the major, typical aspects of horror that researchers have studied empirically - personality (empathy, Big Five, Dark Triad, sensation-seeking), sex, the associations between mental health and horror (principally, negative), the behavioural effects of watching horror, how children and older adults respond to frightening TV/film fiction, the importance of sound to the horror experience, and how the brain responded to watching and hearing horror.

It is fair to say, there was not much research to review even when you included books - and those, while generally diverting, were hardly ever empirical and when they did stray into the empirical you sometimes came away with the certain impression that the straying into the empirical for most of them was ill-advised. No one wants yet another misinterpretation of a questionable, small sample fMRI study. The paper included 193 references - around 86 of those references were specifically related to empirical (or quasi-empirical in the case of case studies) studies of horror and horror-related, fear visual stimuli (including fictional and factual horror, and film, TV and video-game stimuli). Perhaps not so much a field, more of an erratically maintained lawn with some alarming topiary.

The review concluded with a re-establishment of two of the field’s most well-established cliches: that boys and men were more likely than girls and women to enjoy horror and seek it out, and that horror fans were low in empathy and high on some aspects of sensation-seeking. But behind the cliches, there was always the niggling doubt about how robust these conclusions were.

In 2005, for example, Hoffner and Levine published a meta-analysis of studies examining the correlates of the enjoyment of fear-based fictional visual stimuli. They found, amongst other things, that enjoyment of fright and violence was negatively correlated with empathy - the lower the empathy, the greater the enjoyment. However, the strongest effects, and these were found only in two studies, were found when stimuli were horror film but specifically horror involving torture. When these studies were removed, the relationship disappeared. The finding highlights, perhaps inadvertently, one of the unspoken limitations in much of horror research which may cloud any serious conclusions about replicable findings: what we measure is very much dependent on the stimuli presented and the specific film chosen, and these might be some distance removed from the overall genre of horror.

A challenge to the accepted view of horror fans was made by Scrivner (2024) in his paper which asked people for their views of horror fans and correlated new measures of social desirability with horror fans’ own self-perceptions. In the study, people regarded horror fans as less kind but this finding interacted with the rater’s own favoured genre: people whose favourite genre was horror regarded horror fans as kinder than did those whose favourite genre was action, drama or “other” although the paper reports a one-way ANOVA while the analysis suggests the inclusion of two variables. Horror fans were regarded as significantly less empathetic than were comedy or drama fans, and action fans were also regarded as less empathetic than fans of those two genres. Similar results were found for compassion. So much for stereotypes: others think horror fans are a little odd.

A follow up study of 244 individuals asked participants to indicate how much they enjoyed the following sub-genres of horror: gore/splatter, monster, paranormal, psychological, and slasher. Measures of empathy and coldheartedness were administered. Overall, gore/splatter was enjoyed less than slasher, paranormal, psychological, and monster subgenres, and the slasher subgenre was enjoyed less than paranormal, monster, and psychological subgenres. Enjoyment of all was positively correlated with empathy measures and negatively correlated with coldheartedness. Sex and age were significant moderators. Older people liked horror less and men liked horror more. When both variables (sex, age) were controlled for, lower coldheartedness predicted horror enjoyment as did higher cognitive empathy (but not affective empathy). Horror fans across all subgenres were just as likely to donate money to other participants in a subsequent laboratory exercise. As an empathy proxy, this appeared to demonstrate that horror fans are just as kind as fans of other genres of film.

Collectively, the Scrivner studies suggest that while the perception of horror fans is negative the reality is more positive. In the subgenre study, participants were recruited from a general pool. If the results are replicable, they should also apply to those horror fans who attend horror film festivals - the hardcore devotees of the genre. This would be the true test of the hypothesis that those most likely to seek out horror are the least empathetic and compassionate.

Which begs the question, why has the literature reported a different finding historically? Even if we begin with the basic steps, we encounter an obvious challenge: a plethora of media and presentation methods and stimuli. Studies have used films, television shows and video games. They have presented, usually, clips of these stimuli when the study is genuinely empirical and directed. Occasionally, participants watch an entire film. These are studies in which the effects of the horror stimuli are examined directly, and there is normally an experimental control and some other form of control. Many studies, however, are correlational- they examine whether enjoyment and seeking out of horror is correlated with specific personality types. Correlational research can provide answers to a prescriptive and limited set of questions but the questionnaire and questions administered can come with some embedded presumptions and assumptions. One such assumption is that horror is unitary, a thing, a genre with very specific criteria.

But, of course, it isn’t. If you examine the most financially successful horror films of the past 50 years - those that have garnered the highest, absolute box office returns - you detect not unity but diversity. The list includes IT and IT2, The Nun, The Conjuring, The Exorcist, Hannibal, The Blair Witch Project, The Silence of The Lambs, Us, Parasite, Halloween and two Resident Evil films. They are all clearly horror films by the broadest of criteria, but they are different kinds of horror film. They are the same family, but your grandmother isn’t your daughter and your father isn’t your uncle. Platts and Clasen’s (2017) analysis of trends in horror film between 2006 and 2016 found that the most common subtype of horror was supernatural horror. That 1931’s Dracula, Donovan’s Brain, Rosemary’s Baby, Night Of The Living Dead, Dracula AD 1972, The Texas Chainsaw Massacre, The Exorcist, A Nightmare On Elm Street, Henry Portrait of a Serial Killer, The Silence Of The Lambs, Annabel, Alien, The Omen, Halloween, 28 Days Later, Hostel, Get Out and Paranormal Activity are all classed a horror perhaps illustrates, in a very few titles, the potential problem with the umbrella term, “horror”. Research has assumed that this is an identifiable genre (defensible) and that the films they have selected are exemplars of that genre (less defensible). Research has also been largely unclear about what it is actually looking for. Kiss et al. (2024) have recently reported a variety of interactions between personality and psychological variables, and excitement generated and enjoyment evoked by watching horror clips, and note that the latter two are, perhaps understandably, not synonymous. Participants watched clips of 20-30 secs duration from five subgenres of horror which the researchers describe as gore, killer/slasher, monster, paranormal, and psychological. The films were not identified in the paper.

This representation of horror’s diversity, and its fractal nature as a genre, may be the hidden moderator behind some of the area’s historically most well-known, if perhaps unsound, findings. It is not a new observation - Hoffner and Levine almost made it explicit in their review - but it is one explanation amongst several for why low empathy may not be the replicable correlate of horror enjoyment that it has stereotypically been considered to be. When Hoffner and Levine removed the horror involving victimisation from their analysis, the relationship between empathy and enjoyment dropped to almost zero. An examination of the types of films researched in these studies historically exemplifies how varied and unhomogeneous they are - Nightmares, Friday the 13th, Night of the Living Dead, Jaws, Boogens, Videodrome, Candyman, Needful Things, Leprechaun, Pet Semetary, The Omen (2006) and others, have all been included as stimuli. And it is impossible to ignore how anchored some of these films are to the times in which they were used. Films rated X or 18 decades ago may now be passed as 15 or even PG suggesting that historical research may be time-bound. Which begs another variant of the question: what is horrifying about a horror film if it is judged to be less horrifying a decade or more after its release? Perhaps we need gradations of horror.

And if the research suggests that people respond to specific films or specific sub-genres of film, where does this leave research? Horror research shares many of the same issues and challenges as two other seemingly unrelated but actually quite cognate research areas-comedy and olfaction. Co-incidentally, each has at least one journal exclusively dedicated to their study (Horror Studies, Comedy Studies/Humor, and Chemical Senses). Few, if any, empirical reports are published in Horror Studies; it is occupied mainly with aperçus and observations about specific films or genres opinion masquerading occasionally turgidly as academic discourse. It publishes little research of the kind undertaken by the Recreational Fear Lab. The subject of all three fields involves a stimulus designed to evoke emotion, positive or negative. Some observers have also drawn more explicit comparisons between comedy and horror but these comparisons are often over-exaggerated, often revealing failures in syllogistic reasoning (my cat has four legs, my dog has four legs, therefore my cat is a dog).

The issues and challenges mentioned above largely concern the methodological quality of studies and the procedures applied within them. Many older studies in the field are characterised by very small samples and very restricted samples (students, in the main). This is a common feature of nacent research areas in the 1980s and after in all three fields. The choice of horror film - when films are chosen - is seemingly arbitrary.

Horror research does not need anything as pompous as a manifesto, but perhaps it needs something like it. What would such a quasi-manifesto look like? Here is one, incomplete, inchoate attempt.

1. Define and defend the type and choice of horror film included in the research. It may include gore and violence as a defining characteristic; it may include tension and apprehension. It may even, ideally, include both and others as comparison and as a way of determining whether these sub-genres are distinguishable in terms of their putative effects.
2. Ask participants to determine whether the film to which they are exposed exemplifies horror and ask them why.
3. Justify the length of your stimuli if a film or film clip is being used. Define what is specific about the content clip and ensure that the clip only differs from others in this specific way, i.e., provide an internal control.
4. Recruit large samples and, ideally, include participants other than students
5. Ask specific questions about the films watched- ask participants to evaluate the characters, the sound, the music, the characteristics of the film’s perpetrator and victim, the use of colour and grading
6. Recruit samples that are not WEIRD (Western, Educated, Industrialised, Rich and Democratic)
7. Recruit theoretically justified diverse samples
8. Examine exogenous variables that can affect people’s responses to horror film. Does the temperature in the testing area affect this response? Does watching with one or more people (and one or more people of the same or different sex) moderate any response? Can ambient aroma in the environment in the testing area influence these responses? Can the lighting conditions under which films are watched affect the response? Does the size of the screen and the volume of the sound affect the response? And do these variables interact? Are there aspects of these variables that can enhance the experience/enjoyment of horror or, conversely, diminish?
9. Consider whether the sex difference in response to horror is mediated by disgust-sensitivity, a factor which the literature suggests may be one mediator.

Then there is the elephant in the room. Not whether the subject should be studied at all but whether it is possible at all to study empirically what we experience in what William Friedkin describes as “safe darkness”. At one level, this is an easy elephant to address. It is more than possible to study behavioural correlates of preference for, and enjoyment of, horror media. It is possible to investigate correlates of different types of horror media and sub-genres within these horror media (supernatural, gore, torture porn, psychothriller, found footage). That is horror’s easy consciousness problem.

The more difficult research question is behavioural - whether the effect of what people are watching is due to suspense, apprehension, fear, anxiety, disgust, haemophobia, revulsion towards violence, amongst other factors. And those are only some of the factors of interest. What participants might also be experiencing, and which we don’t measure, may be just as relevant. Are they looking at the actors, the attractiveness of the actors, the quality of the actors, listening to the quality of the dialogue, the quality of the sound etc etc. So, perhaps, rather than investigating horror, we should really investigate responses to characteristics of horror films- such as deaths, violence against others (seen or implied), the type of victim killed or injured, jump scares, or chases (which generate peril) specifically.

And while these characteristics are worth disentangling, there is also an obvious point to make that people who watch horror films want to watch horror films - so you are not investigating the effect of horror on people generally, but those who seek it out. And that is a very niche group, producing data with restricted utility from a potentially self-indulgent study.

If, as Warren Ellis wrote, “horror is, at heart, about being disturbed”, and not, to quote Virginia Woolf, “the strange human craving for the pleasure of feeling afraid” perhaps we are looking in the wrong place - not in the abattoir or the surgery, but somewhere more quotidian that is behaviourally measurable and explicable.

G. Neil Martin is Honorary Professor of Psychology at Regent’s University London, a Life Fellow of the RSA, and Associate Member of the Recreational Fear Lab at Aarhus University.

Reference List

Hoffner, C. A., & Levine, K. J. (2005). Enjoyment of mediated fright and violence: A meta-analysis. Media Psychology, 7(2), 207-237.

Kiss, B. L., Deak, A., Veszprémi, M. D., Blénessy, A., & Zsido, A. N. (2024). The role of excitement and enjoyment through subjective evaluation of horror film scenes. Scientific Reports, 14(1), 2987.

Martin, G. N. (2019). (Why) do you like scary movies? A review of the empirical research on psychological responses to horror films. Frontiers in Psychology, 10, 430538.

Platts, T. K., & Clasen, M. (2017). Scary Business: Horror at the North American Box Office, 2006–2016. Frames Cinema Journal, 11, 1-28.

Scrivner, C. (2024). Bleeding-heart horror fans: Enjoyment of horror media is not related to lower empathy or compassion. Journal of Media Psychology: Theories, Methods, and Applications, in press.