This effect was stronger in rule-in blocks, leading to a rule type (rule-in, rule-out) x decision (commit, defer) interaction on response times (F 1,19) = 27.8, p < 0.0001) with no main effect of rule type (F 1,19) = 1.1, p = 0.3).
Confronted with Dinneen's account of their interactions, Icahn deferred to his lawyer, Jesse Lynn, who disputed several points.
Accordingly, we have toned down our language in asserting this claim; nevertheless, in the Discussion section we continue to call upon the neural data from the dACC which show both a main effect (of commit vs. defer) and an interaction exclude/include x value to support the argument that there are separable additive and multiplicative biases, and that the dACC signals encode these biases.
We therefore do not agree with Ward et al. that proposals for handling difference and equivalence tests, setting equivalence limits based on observed natural variation, and displaying the combined results in a graphical form should be deferred "until the issue of interactions has been resolved".
A statistically indistinguishable pattern of activity was observed more ventrally in the medial orbitofrontal cortex (BA 10), a region that has been labeled arMFC or vmPFC (rule x value interaction on defer trials: peak −2, 52, −10; t(19) = 4.85, p < 0.00006) and where activity typically correlates the expected value during independent choices (O'Doherty et al., 2001; Plassmann et al., 2007).
In a subsequent analysis, we additionally extracted the BOLD timeseries from the peak group response to rule x decision x value in the dACC (−6, 32, 34) and included it (and its interaction with defer vs commit) as an additional regressor.
Necessary maintenance was deferred.
More like deferred frustration.
