How do we learn to feel safe or threatened?


Anxiety disorders are a kind of mental disorder defined by problematic anxiety, fear and avoidance (Craske & Stein, 2016). They are the most common type of mental disorder worldwide, develop early in life, and tend to be chronic. Those with anxiety disorders have a heightened sense of threat and danger that is excessive and not in proportion to the situation of concern. Exposure therapy is an effective psychological treatment for anxiety disorders. Exposure means approaching a feared (but actually safe) situation or thing (as in “exposing” yourself to what you are scared of). Exposure therapy involves approaching feared (but actually safe) situations with the support of a therapist until it is learned that these situations are safe. Decades or research and clinical experience demonstrates that exposure therapy improves all types of anxiety disorder. Despite this, researchers are not sure how and why exposure works. The leading view currently is that exposure works by providing experiences where clients can learn that their fears don't come true. The technical term for when a fear doesn't come true is "threat prediction-error". Discrepancy between anxious predictions and actual outcome vary from being a lot worse than expected, as-expected, to a lot better than expected. Current theory posits that larger threat prediction error during an exposure will lead to more safety learning. Exposure therapists should thus try to create conditions where people are surprised by the outcome and see that their fears don’t come true to the greatest extent possible. Evidence supporting this theory of exposure therapy comes from basic fear learning research with animals and humans and using controlled experimental procedures in laboratory settings. This fear learning research often involves participants viewing trials of abstract geometric shapes on a computer screen. Some shapes pair with an electric shock or loud noise and some don't. Researchers measure fear to see how it changes through different stimuli combinations. This research shows that fear in this kind of procedure reduces after threat prediction errors. Some researchers assume that the findings from this kind of research will translate to the "realworld" context of therapy in the clinic. However, findings may not transfer to the more complex and "messy" settings that characterise everyday life. We need more research to determine whether the threat prediction error model of exposure translates to real-world contexts outside of the laboratory. It is important to know the therapeutic mechanisms of our treatments. If we don't know the true treatment mechanisms, we may be subject clients to unnecessary procedures. Instead, we could drop unneeded procedures to focus on the real mechanisms of change. This study aims to address these gaps by exploring whether threat prediction error is associated with anxiety reduction through a procedure that resembles exposure therapy in the clinic. This procedure involves give six short speeches to a researcher over two days and rating anxiety, threat perception, and perception of threat outcome for each speech.

Research Questions / Hypotheses


  1. How often do speeches turn out better, worse, and as-expected? 2. How does threat expectancy and perceived outcome relate to speech anxiety change over time? 3. How does social anxiety symptom level relate to speech anxiety, threat prediction, and perceived outcome? Hypotheses: 1. Anxiety, threat expectancy, and perceived outcome scores will vary significantly between participants, and also with within participants over time.
  2. Discrepancy between speech threat expectancy and perceived outcome will relate to speech anxiety change over time.

    i. Speech anxiety will be lower after speeches that went better than expected. The better the outcome was compared to what was expected, the lower anxiety will be at the next speech.

    ii. Speech anxiety will be higher after speeches that went worse than expected. The worse the outcome was compared to what was expected, the higher anxiety will be at the next speech.

    iii. Speech anxiety will remain similar after speeches that went as-expected.

  3. Higher social anxiety symptoms will be associated with higher overall levels of speech anxiety, threat expectancy, and perceived outcome.


69 people have participated in the study so far. 57 completed both part one and two. We need 100 participants in total and will keep recruiting next semester.


It is a longitudinal and observational study conducted remotely via Zoom. Participants give six 3 minute speeches to a researcher with one minute to prepare, while rating their anxiety, sense of threat, and threat outcome. These six speeches are done across two sessions separated by a week (3x speeches per session). Social anxiety and mood questionnaires are completed before each session.


Multi level modelling will be used where threat expectancy, threat outcome, and anxiety at time X will be entered as predictors of anxiety at time X + 1. A discrepancy effect will be supported if threat expectancy and outcome predictors are both significant and opposite in sign (Humberg et al, 2018), and this is what we are expecting. We will also test whether the joint relationship between threat expectancy / outcome and anxiety is more complex than a simple discrepancy effect by adding a set of higher order terms to the model (e.g see Nestler et al, 2019) and evaluating whether this improves fit with the data (Edwards, 2002). If adding higher order terms do not improve fit, then it allows more confidence in establishing that the relationship between the two predictors and anxiety represent a simple discrepancy effect.


The findings from this study will help determine whether threat prediction-errors are related to safety learning in settings that are similar to those in exposure therapy. Basic research in the laboratory suggests that they will be, but this proposition hasn't been tested enough in naturalistic settings. To our knowledge, no other studies to date have explored this explicitly in the area of social anxiety. Giving a speech is more complex than experimental procedures used in basic fear learning research, so it can't be assumed that safety learning works the same way in these two scenarios. Clinicians are conducting exposure based on the prediction error model, so it is important to know if this practice is correct or whether they should focus their practice on other mechanisms. This study will contribute to knowledge about how fear and safety learning works for humans in everyday settings. Achieving these aims will help refine understanding of exposure therapy mechanisms and may lead to better treatment outcomes for people living with anxiety. Anxiety disorders are the most common mental disorder worldwide, so research that may help improve outcomes for these issues is important.