DV | Within dataset (new/original): DV ∼ fROI + (1|ID) Across datasets: DV ∼ fROI + dataset + fROI × dataset + (1 + dataset|ID) | |||||||
---|---|---|---|---|---|---|---|---|
Language LH | Language RH | MDfp | MDco | |||||
New | Original | New | Original | New | Original | New | Original | |
ISCb | 0.22 | 0.28 | 0.18 | 0.21 | 0.11 | 0.15 | 0.09 | 0.11 |
Z = 16.2 | Z = 13.4 | Z = 13.5 | Z = 10.1 | Z = 8.2 | Z = 7.6 | Z = 6.3 | Z = 5.0 | |
p ≈ 0 | p ≈ 0 | p ≈ 0 | p ≈ 0 | p < 10−15 | p < 10−13 | p < 10−9 | <10−5 | |
Z = 4.78, p < 10−5 | Z = 2.38, p = 0.03 | Z = 4.05, p < 10−4 | Z = 1.5, p = 0.21 | |||||
Language localizerc: sentences >; nonwords | 0.17 | 0.63 | −0.01 | 0.21 | −0.24 | −0.38 | −0.03 | −0.13 |
Z = 4.4 | Z = 12.0 | Z = −0.4 | Z = 4.0 | Z = −7.0 | Z = −7.7 | Z = −0.7 | Z = −2.4 | |
p < 10−4 | p ≈ 0 | p = 1 | p = 10−4 | p < 10−11 | p < 10−12 | p = 0.77 | p = 0.03 | |
Z = 11.9, p ≈ 0 | Z = 5.9, p < 10−7 | Z = −3.9, p < 10−3 | Z = −2.3, p = 0.04 | |||||
MD localizerc: hard >; easy | −0.03 | −0.12 | 0.06 | 0.01 | 0.37 | 0.95 | 0.04 | 0.52 |
Z = −0.4 | Z = −1.4 | Z = 0.8 | Z = 0.09 | Z = 5.22 | Z = 10.5 | Z = 0.5 | Z = 5.3 | |
p = 1 | p = 0.48 | p = 0.97 | p = 1 | p < 10−6 | p ≈ 0 | p = 1 | p < 10−6 | |
Z = 1.4, p = 0.42 | Z = 0.7, p = 0.94 | Z = 8.4, p ≈ 0 | Z = 5.9, p < 10−7 |
↵aWe compare the data of the first replication reported in the manuscript (“original” dataset) to data derived from the “alternative” fROIs (“new” dataset). The first replication was chosen because it had a sufficient number of participants (n = 13) who listened to the same two stories, namely, 1 and 3.
↵bFor the new dataset, one story was used for defining “alternative” fROIs, and the held out story was used to estimate their ISCs independently of the criteria used to define them. The process was then repeated with the two stories in reversed roles, and the resulting two estimates for each fROI were averaged.
↵cFor language (MD) fROIs in the original dataset, one run of the language (MD) localizer was used to define fROIs and the second run was then used to estimate their responses independently of the criteria used to define them. The process was then repeated with the two runs in reversed roles, and the resulting two estimates for each fROI were averaged.