There’s further discussion in the second link where the original authors stand by their claim.
The two use different statistical methods to try to demonstrate the conclusion, and that’s where the difference lies.
I’m not a big stats person, but I’m coming away feeling like the original claim is valid since a) it was shown in two different models the original author used and b) it makes intuitive sense to me.
Direct link to paper
And…
As a result of our analysis we find that the authors’ conclusions are not supported by the data. The data do support the conclusion that high-skill players make more positive and fewer negative comments, perhaps consistent with good team-building behaviour.
There’s further discussion in the second link where the original authors stand by their claim.
The two use different statistical methods to try to demonstrate the conclusion, and that’s where the difference lies.
I’m not a big stats person, but I’m coming away feeling like the original claim is valid since a) it was shown in two different models the original author used and b) it makes intuitive sense to me.