Can mindlessness in CASA effects be measured independently?

Eun-Ju Lee argues that mindlessness, the dominant explanation for why people treat computers as social actors (the CASA paradigm), has never been independently measured or directly tested in most of the research that invokes it. "Mindlessness as a psychological state was not separately measured, nor was it directly tested. Instead, mindlessness was either presumed or inferred from people's reactions to computers." The inference runs backward from effect to cause: "if people are mindful, they should not apply social rules to a computer—given that they did, they must have been mindless."

Lee's own comparison sharpens the problem: this reasoning is "analogous to saying that if argument quality made a difference in persuasion outcomes, people must have engaged in systematic processing." Without a mediator measured independently of the outcome it is meant to explain, the claim cannot be wrong. She notes a rare exception, Lee (2008), which manipulated multitasking load and output modality (speech versus text) to vary mindlessness directly and then checked whether gender-typed responses to computers followed. The results supported the prediction, but Lee observes that even a null result would not have settled anything: a failure to find the predicted pattern could mean either that the manipulation failed to induce mindlessness, or that mindlessness failed to produce the social response — and without an independent measure of the mindless state itself, the two possibilities are impossible to tell apart. "Unless we measure mindlessness independently and test these two processes separately, the presumed connection between mindlessness and social responses to computers remains virtually unfalsifiable."

This bears directly on claims elsewhere in the human-machine communication literature that rest on the same inferential move. Do humans apply human-human scripts to AI interactions? treats systematic shifts in people's responses across repeated interactions with media agents as evidence that a second, media-specific script system is applied just as mindlessly as the original human-human scripts — but the mindlessness of that application is inferred from the systematicity of the behavior change itself, not measured apart from it. Lee's critique does not refute the extended-CASA account, but it identifies the same evidentiary gap: systematic behavior is consistent with mindless processing, but it does not establish it. The companion claim in this batch, Does comparing machines to humans actually reduce social responses?, shows that the rival explanatory mechanism has its own, different set of problems — which is why Lee treats neither account as adequate on its own.

The excerpt does not claim that mindlessness is false or that CASA effects don't occur; Lee's point is narrower and more corrosive: the standard evidentiary chain linking a mindless psychological state to observed social responses has rarely been tested, only assumed. The implication for design-facing claims about users "mindlessly" applying human social rules to AI interfaces is that such claims, when they rest only on behavioral pattern-matching rather than an independently measured mediator, are stating a plausible interpretation, not a demonstrated mechanism.