Evaluating the evidence base for relational frame theory: a citation analysis.

Simon and his team read every paper that cited Relational Frame Theory between 1991 and 2008. They found 174 articles and sorted each one into 'empirical' or 'just talking about ideas.'

They counted who was studied, what frames were tested, and whether the work happened in a clinic or a college lab.

Only one in three papers had real data. Almost all of those data papers used neurotypical adults sitting at computers. The 'same-as' frame got nearly all the attention; other frames like 'bigger-than' or 'opposite' were rare.

Clinical studies with kids or people with disabilities were still hard to find.

Weber et al. (2024) and Kaur et al. (2025) show the field has moved on. Their big case-series now track real clients with autism or ID who receive functional analyses and treatment. These newer papers fill the gap Simon spotted—more clinical data, fewer college demos.

Feinstein et al. (1988) is a grandparent here. That single-case paper proved you must match treatment to the function of self-injury. Simon’s map shows the idea caught on in print, but mostly as reviews, not new tests.

Dowdy et al. (2022) gives the next step: once you collect those cases, use funnel plots to be sure you are not hiding null results. Simon’s count of mostly non-empirical papers is exactly the pile Dowdy would flag for possible bias.

If you run verbal-language programs, notice the imbalance: we know a lot about 'same' with adults, little about 'before-after' or 'deictic' with kids. Pick a missing frame, add a clinical participant, and you just moved the field forward. When you write it up, run Dowdy’s bias tools so your study strengthens, not just adds to, the pile.