search

NDCLab / readAloud-valence-alpha

analysis | real-world reading, lexical valence, and word frequency
GNU Affero General Public License v3.0
1 stars 1 forks source link

analysis: initial readme/contributing release #8

Closed jessb0t closed 2 years ago

jessb0t commented 2 years ago

@georgebuzzell Kindly review the initial readme and contributing files for the readAloud-valence-alpha analysis project. Note that I have transferred the summaries of the planned analyses here on the dev-transfer branch. :pray:

jessb0t commented 2 years ago

@georgebuzzell I have now incorporated the changes that we made to the readAloud-valence-dataset readme. I also tweaked presentation of the first paragraph (to separate it into bullets) as I felt that improved readability.

I am hoping you can assist with a concise addition for post-error behavior? I'm excited to add it to the analyses, but don't yet have a clear vision on your precise prediction. Is it that, after an error in a negative passage portion, people are less likely to make additional errors than they are when they make an error in a positive passage portion (due to congruency between the negative passage valence and the negative error)? Thank you in advance for the guidance here! :pray:

georgebuzzell commented 2 years ago

@jessb0t Sorry for the delay here!

Re the post-error hypothesis. I am not sure if i (yet) have a strong hypothesis of the exact behavioral change we should see, post error, as a function of valence. But, the overarching hypothesis is that valence will impact error monitoring and in turn lead to changes (of some sort) in post error behavior. The basic idea, is essentially similar to what you suggest. I.e. that in the negative valence condition, there will be more of a "match" between the experience of an error and the overal context. However, this could lead individuals to be more likely to detect errors, and thus, exhibit greater behavioral changes in either a beneficial or negative way, or, it could lead individuals to be more likely to quickly/efficiently process errors, leading to less dramatic post-error behavior changes.

In some, the hypothesis, at least in my mind (currently) is that valence will lead to changes in post-error behavior, although we do not have a specific hypothesis of the kind of change (exploratory at that level). could change as we approach data collection though...

georgebuzzell commented 2 years ago

GREAT WORK, @jessb0t ! As always :)