The studies i've seen so far on this had one thing in common: They do not get the actual use of trigger warnings. Either they just put the warning there, but participants have to view the stuff anyway, which ofc means the warning does absolutely nothing, or participants voluntarily view it, which also says little about the use of the warning. Trigger warnings are about giving people the option to avoid certain topics. Many people who use them may or may not do so on a case by case basis, like assumed severity of the material, general mood in that moment, a calculation of their need to stay informed vs. how recent and severe their latest trauma has been and so on. In a study situation (where they will usually not know that the study researches use of trigger warnings, were they will usually not expect particularly gruesome material, and were they may feel that they should be thorough to be a useful participant), it's reasonable to assume that people will be much more likely to view CWed material.