Likelihood score Collocation

[Joshua Ball]

Hi Anthony,

I am using your software for my dissertation about Donald Trump's Facebook discourse and its certainly very helpful. I was wondering if you could shed light on what the likelihood and effect scores produced by the collocates actually mean. Particularly for likelihood, I can't find how it might relate to p-scores.

Many thanks

[Laurence Anthony]

Hi Joshua,

I'm now sure why I haven't answered this question up till now.

The (log) likelihood score looks at the probability of the target word appearing next to the collocate candidate. If the collocate appears in the list, this means that the pair appear together statistically significantly more likely than chance. The p value is set in the tool settings. 

As for the effect (size) score, this is simply a measure of the degree (size) of closeness. Words that always appear together will have a higher score than those that rarely appear together. The score is normalized for frequency of occurrence.

I hope that helps.

Laurence.