Research Breakdown: Behavioural Response to Threat

The amount of research that is out there on the topic of self-defence is limited, and a lot of it is not necessarily very accessible. I’ve spent a lot of time going through the existing papers as part of my own research project a while ago So I thought it might be interesting to break some of them down on this blog every once in a while. These are a lot more work than most other posts, so I’m not going to do this all that often.
This time I’m going to look at a paper from 2015 by Anderson and Cahill, titled “Behavioural Response to Threat (BRTT) as a key behaviour for sexual assault reduction intervention: A critical review“. It’s quite a large paper, mostly looking at other studies. Because this post is quite long, I’m including a bullet point overview, and you can read the full breakdown below.
The paper deals with sexual assault and focuses on the behaviour of the targets of such assaults. Be aware that this is the topic of this post.


  • The paper is a review paper and mostly looks at a variety of other research
  • The area of focus is sexual assault, specifically the effect that the responses of potential victims may or may not have on the outcome
  • The existing research on this topic is quite inconsistent in terms of methodology and terminology, which means it’s hard to make really valid conclusions
  • The authors are trying to introduce a term they came up with: “Behavioural Response to Threat” to make the terminology more consistent


  • According to the authors, the outcome of assaults is affected by the threat perception and the threat response of the potential victim
  • The authors conclude that based on the existing research assertive responses by the victim result in more favourable outcomes
  • They suggest that internal psychological factors may play a larger role in determining the responses than environmental factors
  • The environmental factor with the largest effect on the response of the target was the relationship with the perpetrator
  • Previous victimization was also found to have a correlation with risk assessment and threat response, therefore being a risk factor for more likely re-victimisation
  • Regarding interventions such as self-defence training, the authors state that they should work, but that there is some evidence against that, partly because of a major mistake with a reference they used.
  • Due to the scattered and limited nature of the existing research, more is needed before any firm conclusions can be drawn


  • The paper is valuable because it reviews a huge amount of existing research done in a variety of ways, making it a good starting point for getting into SD research
  • For a self-defence instructor, the focus on potential risk factors is really useful
  • Especially the focus on acquaintance assault – we really tend to overdo the stranger danger thing in SD – and how it can impact victim responses is great
  • Everything in the paper needs to be taken with a grain of salt, as there is little solid evidence for many of the statements made in it

Goals and nature of the paper

The paper seeks to provide a review of the existing literature on the nature and impact of the different responses of potential victims to the sexual assault threats. The goal of the paper is to introduce a term the authors developed for the potential responses of someone targeted for sexual assault. The term is “Behavioural Response to Threat”, or “BRTT”. They -correctly – point out that the existing terminology is inconsistent, and want to introduce the term to provide a better framework to be able to do things like compare results across studies. To this end they review a large chunk of the existing literature on the subject, looking at both qualitative and quantitative studies, describing some of the methodologies used and results obtained.
This is quite fascinating, as the nature of the subject limits truly empirical studies severely. The most common way studies are done is with questionnaires or interviews asking if women – like most existing research on self-defence, this paper focuses exclusively on women as potential victims – had been assaulted and about their responses. There are two commonly used questionnaires (“Victim Response Strategies” and “Response Questionnaire” if anyone is interested), although the authors point out that they are often modified for individual studies, making comparison difficult. There are also numerous well known issues with the reliability of self-reporting, particularly with sensitive subjects such as this one.

There are other ways research is conducted, often with hypothetical scenarios, using story vignettes, role-playing or even virtual reality. These are really interesting and innovative, and give some insightful results, though of course they come with their own caveats for reliability.

Different responses

What the research seems to suggest is that the responses of the target, in particular their ability to assess threats and the nature of their responses, have a significant impact on the outcome of the assault. It also suggests that assertive responses, both verbal and physical, are the most effective at deterring assaults. One study (Clay-Warner 2002) found that only a physical response predicted avoidance of a completed assault. The authors seem surprised by the relatively high (29% in a hypothetical scenario and 35% in epidemiological studies) percentage of women listing non-assertive responses or no responses. This should only really be surprising from a deeply academic point of view, as things like freeze reactions, socialised beliefs about the viability and consequences of resistance etc. impact this number. A caveat here is that the limitations of the existing research make it really difficult to assess details such as under which circumstances different responses may or may not be effective.

Predictors” of Responses

For these different BRTT’s, the paper also examines research regarding any predictors of them, i.e. what factor affect how people respond. There are some caveats here: While the authors frame it as “predictors”, upon reading I personal think it might be more accurate to say that these are correlates, I think there’s not quite enough empirical data to support a direct causation. This section is also written a little bit inconsistently, I suspect because the different studies they look at give different results. For example, at one point they mention that high relationship expectations are predictive of more diplomatic and non-forceful responses, while two paragraphs further they state that there are few predictors of non-forceful responses. Again, I think this is simply because they look at a wide range of studies and it’s difficult to synergise them consistently. If you read the paper, it’s just something to be aware of.
The way the authors model the factors is based on existing research (Nurius & Norris 1995), dividing them into background, interpersonal and intrapersonal factors.

For background the only thing that they really look at, and that seems to be a strong correlate is victimisation history. Some of the studies (Craford, Wright and Birchmeier 2008; Messman-Moore and Brown 2006) suggest that previous victimisation is linked to less threat awareness, lower response to risk – e.g. leaving a threatening scenario later – and less forceful responses. It is not clear from the presented data how the causation plays out, whether the effects are due to the previous victimisation or previous victimisation is more likely in people where the above factors are present.

For interpersonal factors, the most substantial effect on the response is the relationship of the potential victim to the perpetrator (Clay-Warner, 2002). This is hugely important, as acquaintance assaults are more likely than ones by strangers. Any tacticool “murderdeathdefence” techniques we teach might go straight out the window in such a situation, and we need to be able to take this into account in our training. The severity and exact nature of the assault may also play a role, but the evidence is still inconclusive. An existing relationship, as well as high expectations about that relationship, and concerns about it correlate with less forceful and less effective behavioural responses. This holds up even in some of the research with hypothetical scenarios which is remarkable.

The authors suggest, based largely on one study (Fisher et al, 2007) that intrapersonal (i.e. internal psychological) factors play a greater role than external environmental ones in determining the responses a target will be able to utilise. This is quite a controversial statement, and the evidence is definitely not conclusive. However some of the data points strongly towards this, and it is definitely something to keep in mind when teaching, as it suggests that mindset and psychological conditioning are major factors in successfully resiting assaults of this nature. Higher levels of assertiveness, (shockingly…) correspond to more assertive responses. Feelings of anger, lower concern for injury, less fear and less confusion, less self blame and lower concerns about the relationship all correspond to more forceful and effective responses. The inverse appears to apply as well. Also feelings of sadness correspond with less forceful responses.

This entire topic is very complicated, as well as highly sensitive, so I don’t think any of the above should be taken at face value. It’s interesting data, but more thorough research is absolutely necessary. Things like the interaction between different factors, or the impact of multiple response strategies are not examined at all.

Existing Interventions

The authors also look at existing interventions to reduce sexual assault, including self-defence training. This part of the paper is a little iffy however. They do include some recent studies (Rowe, Jouriles and McDonald 2015) and a review paper (Anderson and Whiston 2005), pointing out an apparent paradox: Based on their previous results, interventions that improve risk assessment and teach more assertive threat responses should reduce assault risks, but while some studies show such outcomes (Senn et al 2015), the review paper ( Anderson and Whiston 2005) shows an insignificant impact. While there really isn’t a huge amount of conclusive empirical evidence that self-defence training reduces assaults, in this case the reason for the contradiction is more simple: The authors made a mistake! The review paper they reference does not actually review self-defence training, but rather sexual assault education programs, which are quite different. In fairness to them, this is not immediately apparent on looking at the review paper, but not being able to tell the difference does suggest that either a) they were superficial in checking their references or b) their understanding of the subject matter is rather limited from a practical standpoint. They do reference other papers which show positive outcomes, but correctly point out that in most cases the outcome measures fall short in some ways, for instance in that there is usually no assessment of the skill improvement of individuals who have received training.

So is the paper useful to SD/MA instructors?

Overall I think the paper is quite valuable, and if you are a self-defence instructor – or just interested in the subject – it’s worth a read if you can get access. The number of studies it reviews alone are quite useful, and depending on how familliar you are with violence and the responses to it, you can get a lot of good information. There are a few things that are a bit iffy, and reading about the varied methodologies used in other studies you begin to understand that a lot of this data is not very reliable – so you still have to think for yourself! There various suggestions for future research, which I sincerely hope somebody will follow up on to get us more accurate data.
I have a few main takeaways from the paper. First that it’s really important to address the psychological and social aspects of acquaintance assaults and how that affects responses in training. Also it supports the view that we need to address mindset, it’s not enough to teach physical techniques. Training needs to include a spectrum of skills, and we need to also help build people’s inner strength in order to be able to utilise their training appropriately. Also, from an aftermath point of view, several of the referenced studies show that there is quite a wide reange of responses to these situations. If we have survivors in our classes, we can use this data to help them come to terms with their own responses, because they may feel that they “should have done something else”. Finally, I really like the “threat perception and threat response” idea as a model, and I’ve incorporated it into some of my seminar programs.

2 thoughts on “Research Breakdown: Behavioural Response to Threat”

  1. Nice breakdown!

    I’ve not read the original paper, but if it’s like every other piece of academic writing I’ve ever read, the point of the article gets buried under tons of stylistic and scientific conventions.

    You seem to have done a good job making the important stuff it talks about accessible.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: