Experimental Examination of Message Framing to Increase Dissemination of Evidence-Based Treatments

Anxiety disorders are highly prevalent and cause significant health burden. Fortunately, effective forms of treatment are available, such as cognitive behavioral therapies (CBT), which have demonstrated efficacy for treating anxiety disorders (Hofmann, Asnaani, Vonk, Sawyer, & Fang, 2012; Hofmann & Smits, 2008). However, relatively few individuals seek and receive treatment from healthcare professionals. While there are a number of reasons individuals may not seek treatment, one barrier may be that individuals do not realize that evidence-based treatments are available and effective (Gallo, Comer, & Barlow, 2013). This multi-method dissertation focuses on evaluating ways of framing PTSD and anxiety treatment information, in an effort to find effective ways of promoting evidence-based treatment information and increasing potential consumers’ interest in these treatments. Five frameworks and factors associated with health behavior change were used to create messages about evidence-based treatments in the form of subheadings for information on the treatments. The messages were evaluated across four studies on their ability to increase interest and engagement in seeking anxiety treatment.

Study 1 was a set of exploratory field experiments with the American Psychological Association, in which we experimentally manipulated subheadings on three webpages of their post-traumatic stress disorder (PTSD) clinical practice guidelines website. We were particularly interested in website user engagement on the site (e.g., clicking on links to learn more about evidence-based treatments) to assess behaviors tied to seeking treatment. Results indicated that visitors to the For Patients & Families webpage did vary their behavior based on subheadings (in particular, “Treatment works: Say goodbye to symptoms” was associated with the most clicks on the “Find a Psychologist” button). However, subheadings did not affect visitor behavior on the other two pages tested.

Study 2 (N = 578) was a preregistered conceptual replication to study whether similar results would emerge in a web-based research study with a sample of US adults who were not necessarily seeking treatment information. Study 1 results did not replicate in Study 2. Discussion of these results focuses on how subheadings may only influence individuals’ behavior if they are motivated to learn about available treatments. Studies 1 and 2 have been accepted for publication in Clinical Psychological Science (Werntz, Bufka, Adams, & Teachman, accepted).

Study 3 consisted of three focus groups of local community members (two groups of individuals with anxiety symptoms [n = 8], one group of individuals who have a family member or close friend with anxiety [n = 3]). In these focus groups, barriers to seeking (or suggesting) anxiety treatment were discussed, with a particular interest in asking how individuals feel about varying messages about evidence-based anxiety treatments. Themes that emerged from focus group conversations included: avoid negative messaging, avoid ambiguity, and know the audience being targeted.

Study 4 was a preregistered online experiment (N = 582) evaluating the credibility of information about treatment and tested demographic variables as predictors of credibility by message type. Latent Profile Analysis (LPA) was used to identify mutually exclusive groups of participants based on responses to headline preferences; three groups emerged. Of the 13 tested demographic predictors of group membership, only percentage of non-White individuals was significantly different across the three groups.

Taken together, this dissertation provides insight into how psychologists and other stakeholders can more effectively disseminate information about treatments that work to potential consumers. The results highlight the importance of examining this topic using different methodologies to understand what individuals want to know and what is more or less effective for whom. The Discussion outlines convergent and divergent results across studies, the importance of replication efforts, ideas for future research, and concrete suggestions for framing treatment information based on available data.