Testing new ideas about how self-injurious thoughts and behaviors work
A major takeaway from our meta-analytic work is that science currently knows little about how self-injurious thoughts and behaviors (SITBs) work. We've accordingly done away with old assumptions and conducted several projects aimed at testing new ideas about SITBs. This work involves experimental designs and typically includes multiple levels of measurement (explicit, implicit, behavioral, physiological). There is surprisingly little of this kind of work in the SITB field. Yet this work is crucial to advancing knowledge because, compared to correlational (and/or solely self-report) studies, experimental work allows us to more directly evaluate ideas/questions and permits more epistemologically valid conclusions. In short, it is difficult to advance knowledge about something without stringent experimentation.
Most of our completed experimental work has been aimed at answering two major questions. First, we reason that people only do things if there's some kind of benefit to doing them. The benefits of SITBs aren't obvious, but still they must carry benefits; so what are these benefits? As described in more detail below (click on the picture), we've found that SITBs can make people feel better via something called "pain offset relief" and via the satisfaction of self-punishment motives. Second, we reason that SITBs are scary, painful, and generally counter to evolutionarily conserved self-preservation instincts. What are the barriers to SITBs and how do some people overcome them? Click on the picture link below to see some of our preliminary answers to this question.
Ongoing and Upcoming Experimental Work
In addition to continuing projects aimed at answering the two questions noted above, we're now using experiments to fill in several other important gaps in knowledge about SITBs:
(1) How do suicidal thoughts come about? We know that suicidal thoughts exist (~2% of the population each year), we know that they explode in prevalence in early adolescence, and we know that they are correlated with certain disorders and negative life events -- but we don't know much else. We're interested not only in the more distal factors that might lead to suicidal thoughts (e.g., a romantic breakup), but also in the specific, proximal cognitive, emotional, and social processes that directly produce suicidal thoughts. In other words, with these experiments we're not looking for the predictors of suicide ideation; we're looking for the mechanisms of suicide ideation.
(2) In the moment, what causes someone to initiate suicidal behavior? Differences between people who only think about suicide vs. people who actually engage in suicidal behavior has become a hot topic recently, but most of the work so far has focused on correlates of ideators vs. attempters. This kind of work hints at the processes that give rise to suicidal behavior; we are interested in specifying and directly studying these processes. As with ideation, we're primarily interested in the specific, proximal processes that directly produce suicidal behavior rather than the predictors of suicidal behavior.
(3) How can we interrupt the processes that produce suicidal thoughts and behaviors? So far, we've identified two such processes (negative association with the self; diminished aversion to self-injury/suicide/death). We have a long way to go to fully understand these processes, but we have identified at least one way to counteract part of these processes: conditioning, specifically evaluative conditioning. Our early work on this, described in the treatment section of this website, uses a broad technique to condition new associations. We are currently testing a range of other techniques that may lead to more powerful and enduring conditioning.
Within these projects, we will be adding to our methodological repertoire, which already includes things like pain manipulations, psychophysiological measures, and a range of implicit assessment tools. Most notably, our newer work includes:
(1) Custom virtual reality paradigms. We have an Oculus Rift and an HTC Vive, and specific lab space devoted to their use. We work with programmers to develop tailor-made paradigms to help us experimentally test our major questions of interest. Compared to our traditional experimental capabilities (e.g., imagining a scenario, looking at a picture, answering a question), virtual reality allows us to immerse the participant into a life-like environment with limitless experimental options.
(2) Machine learning techniques. We take advantage of the machine learning expertise of Dr. Jessica Ribeiro's lab at FSU to apply this approach to help answer our major questions of interest. As Dr. Ribeiro recently pointed out (see here), this approach is far superior to traditional statistical techniques and has already generated major advances in the prediction of SITBs. The TAP Lab uses these techniques to extract, combine, and optimize experimental data in novels ways that deepen our understanding of how SITBs work.
(3) Mobile technologies. Just as our interventions are scaled to mobile technologies (see treatment section of this website), some of our experiments are implemented on a world-wide scale. This includes mobile versions of implicit tasks (e.g., the Implicit Association Test, the Affect Misattribution Procedure), custom surveys, and mobile virtual reality paradigms (e.g., via Google Cardboard).