(PS3-68) Cognitive Behavior Therapy versus "Therapy" on Tiktok: A Comparison of User Attitudes, Claims, and Video Content Characteristics

Background: Therapy is an incredibly popular topic on TikTok, with videos tagged #therapy currently having over 16.9 billion views. Recent research has shown that mental health misinformation and misleading content are highly prevalent on TikTok. In prior work (Lorenzo-Luaces et al., 2023), we demonstrated that 23% of videos about cognitive behavior therapy (CBT) expressed negative attitudes about CBT and commonly included false claims. In the present study, we aimed to extend this work by including a comparison group for therapy generally, in order to differentiate a potentially harmful trend of negative attitudes about evidence-based therapy specifically from base rates of attitudes about therapy.  

Method: We manually coded (κ = 0.79-0.92 across variables) N = 300 videos from a TikTok search (#therapy, #cognitivebehavioraltherapy). Comparison groups were generated by videos’ mentions of “therapy” (unspecified or other therapies) and/or CBT specifically. Coding of study variables allowed for both therapy categories and both positive and negative attitudes in a single video. 

Results: We found that n = 141 videos (47.0%) were relevant and codable, with n = 68 CBT mentions and n = 71 “therapy” mentions. CBT mentions were more commonly negative (29.4% of CBT mentions) in attitude than therapy mentions (9.9%; χ2 = 7.98, p = .004). However, CBT mentions were also more commonly positive (64.7%) than therapy mentions (28.2%; χ2 = 19.04, p < .001). Only a small minority of CBT mentions were neutral (16.2%) versus a clear majority of neutral therapy mentions (64.8%; χ2 = 34.06, p < .001). CBT videos were far more likely to contain claims, educational content, or advice (82.4%) relative to videos about therapy alone (38.4%; χ2 = 24.7, p < .001), whereas videos only mentioning therapy were far more likely to primarily consist of jokes (42.5%) than those mentioning CBT (5.88%; χ2 = 21.9, p < .001). The most common negative claims about CBT were that it is bad for neurodiverse people (15.9% of negative CBT claims) and invalidating (13.6%); neither claim was made about general therapy. The most common positive claim about CBT was that it is effective (26%), whereas the most common positive claim about therapy was that it helped the user personally (13%). 

Discussion: We found that the content of TikTok videos mentioning CBT versus therapy more generally differs greatly, such that mentions of therapy generally are more likely to be valence-neutral and include jokes, whereas mentions of CBT are more likely to include informational claims and express attitudes about the treatment. Content differences (e.g., educational vs. anecdotal) may modify videos’ impacts on viewers, such that future work should consider the effects of posts’ content type rather than their informational content alone. Future work should also use new sampling techniques to generate a more analogous comparison group against which to evaluate CBT attitudes. Finally, differences in claims about CBT versus therapy generally suggest that negative attitudes about CBT may be due to users’ specific experiences with CBT rather than reflective of general public attitudes about therapy. Future work should aim to understand the origins of CBT-specific attitudes, e.g., in the quality of CBT received.