filtering concgrams

The point of it

In order to select which items are "associated", we need some sort of suitable statistical procedures. Before filtering, the members of each concgram are at present merely associated by co-occurring at least a certain minimum number of times, as explained in generating them.

How to do it

The filtering settings in the Controller allow you to specify, for example, that you want to see only those which are associated with a Specific Mutual Information (MI) score of 3.0 and/or a Log Likelihood score of 3.0.

Ensure the statistics you need are checked and set to suitable thresholds, and decide whether any one of the thresholds have to be met (in the case above either MI and Log Likelihood would have to score 3.0 at least) or all of them (in the case above would require both MI at 3.0 or above and Log Likelihood at 3.0 or above). You can also optionally insist on certain words being in your filtered results.

What you see

When you press the filter button (), you will see something like this:

where the items which meet the filter requirements are separated out and selected ready for concordancing; any others are hidden. The thorough checkbox setting gives you details of which of the collocates met the requirements.To the right you see that the head-word HANDS here relates to ON, HIS and EYES but not FIXED because of the thresholds set when I required both MI and LL of 3.0 for the Dickens hands data filtering.