Evoked, NEW anatomy, MNE
===
<font size=3> [Sentence Composition study] </font>
<u><b>Evoked analysis (with new anatomy) </u></b>
To get a full sense of data, let's first focus on the uncorrected left-lateral view for all primary contrasts (=Unexp-Exp, Anom-Exp, Unexp-Anom, Pseudo-Exp).
<span style="color:orange">Note: The time labels in the figures are incorrect. Correct labels are (in ms):
1st row: 0-100 | 100-200 | 200-300 | 300-400 | 400-500
2nd row: 500-600| 600-700 | 700-800 | 800-900 | 900-1000 </span>
[unexpec-expec]
[anom-expec]
[unexpec-anom]
[pseudo-expec]
Some patterns to note:
1) 300-400ms (i.e., 4th figure on the first row in each figure). For all contrasts expect for Unexpec-Anom, we see increased activity for Expected against other conditions. This indicates the engagement of the left temporal lobe in the semantic processing of "expected" stimuli. Given its timing, it could be a part of the N400 response.
2) The LIFG increase for non-expected stimuli at 400-500ms. The LIFG increase has been repeatedly reported in the N400 literature. Our data seem to replicate those observations.
3) The engagement of the Left Angular gyrus in Unexpected-Anomalous contrast. At 300-400, and at 800-900ms again, "Unexpected" evokes more activity in the Left AG than "Anomalous" did. The difference between these two conditions is that the composition of a propositional meaning is <i>possible</i> with "unexpected" condition (though with more effort), but not with "anomalous." That being said, the Unexpec>Anom in left AG might be indexing "semantic composition when the expectancy is controlled for (and is low)." It is noticeable this instance of semantic composition (say, "effortful composition") does not engage the left temporal lobe as much. The fact that this Left AG effect occus at two separate timings suggests multiple stages in the semantic composition of unexpected proposition.
4) Lastly, the Pseudo>Expected pattern at 800ms and later might be related to the button press response. That is, the meaningfulness judgment task could be easier with "Pseudo" condition than with "Expected," and thus the response was made earlier for Pseudo than for Expected.
---
See below for full data. I show three types of figures for each contrast: uncorrected, threshodled at p=0.01, multiple comparison applied with Cluster method (p=0.05).
1. Unexpected - Expected
>a. Left
>(uncorrected)
>(p=0.01)
> cluster-corrected (p=0.05)
> 
> cluster-corrected (p=0.1, masked)
> Max corrected (p=0.05, minnb=2)
>b. Right
>(uncorrected)
> (p=0.01)
> 
>cluster-corrected
>
>c. Bottom
>(uncorrected)
>(p=0.01)
>
>cluster-corrected
>
>d. Top
>(uncorrected)
>(p=0.01)
>
>cluster-corrected
>
>e. Front
>f. Left medial
>uncorreced
>p=0.01
>cluster-corrected
>g. Right medial
>uncorrected
>p=0.01
>
>cluster-corrected
>
2. Anomalous - Expected
>a. Left
>uncorrected
>p=0.01
>cluster-corrected
>
> cluster-corrected (p=0.1, masked)
>b. Right
>uncorrected
>p=0.01
>cluster-corrected
>c. Bottom
>uncorrected
>
>p=0.01
>cluster-corrected
>
>d. Top
>uncorrected
>
>p=0.01
>cluster-corrected
>e. Front
>f. Left medial
>uncorrected
>p=0.01
>cluster-corrected
>
>g. Right medial
>uncorrected
>p=0.01
>cluster-corrected
>
3. Unexpected - Anomalous
>a. Left
>uncorrected
>
>p=0.01
>cluster-corrected
>
> cluster-corrected (p=0.1, masked)
>
>b. Right
>uncorrected
>p=0.01
>cluster-corrected
>
>c. Bottom
>uncorrected
>p=0.01
>cluster-corrected
>
>d. Top
>uncorrected
>
>p=0.01
>cluster-corrected
>
>e. Front
>f. Left medial
>uncorrected
>
>p=0.01
>
>cluster-corrected
>
>g. Right medial
>uncorrected
>
>p=0.01
>
>cluster-corrected
>
4. Pseudo - Expected
>uncorrected
>p=0.01
>cluster-corrected
>
> cluster-corrected (p=0.1, masked)
>
>b. Right
>uncorrected
>p=0.01
>cluster-corrected
>
>c. Bottom
>uncorrected
>p=0.01
>cluster-corrected
>
>d. Top
>uncorrected
>p=0.01
>cluster-corrected
>
>f. Left medial
>uncorrected
>
> p=0.01
>cluster-corrected
>
>g. Right medial
>uncorrected
>p=0.01
>
>cluster-corrected
>
Sign in with Wallet
Connect another wallet