79 Custom reports (admin only)

At the moment, this functionality is only available for admins, who will see a “Report” tab. When the “link” is visible, a report has already been created and is accessible via that link. At the moment, these reports are not permanent and may have to be recreated by clicking the Create report button. If the link does not appear, press the Refresh button.

image-20211117170828063

The link leads to a separate file which at the moment is in html format although Word and PDF files might be possible. This report is constructed from a simple markdown template which could in future be customisable, i.e. you could write your own report within the app, including or excluding modules to suit. At the moment the output is just a rough hack.

If the sources table (or statements table) includes any columns which end with the word “rank” or “rank#”, then a QuIP-like table of ranked organisations is displayed near the start of the report.

One interesting section at the end shows which links were preferentially mentioned according to different groups e.g. women more than men. We ask:

does the proportion of women vs men who mention this link differ from what you would expect (given the total number of mentions of links by both women vs men)?

Women Men
… other links …
Number of mentions of the link from X to Y 10 9
… other links …
Total number of mentions of any link 60 10

In this case we can see that although women mentioned the link slightly more often than men, women altogether mentioned links twice as often as men. So we can compare the number of mentions of the link with the number of “non-mentions” of the link. So we can work out this table (not shown).

Women Men
Number of mentions of the link from X to Y 10 9
Number of mentions of any other link 50 1

We can do a simple chi-squared test on this table to see if the ratio 10:9 is significantly different from 50:1 (which of course it is) – this is the same question as to whether 10:50 is significantly different from 9:1 (which of course it is). If this test is significant, the row “Number of mentions of the link from X to Y” is shown in the table, and the intensity of the colouring of each cell reflects its chi-squared residual, i.e. how different is the number it contains from the number you would expect, given the other numbers?

This comparison is agnostic as to whether there are, say, many men or a few men who talk a lot.

At the moment, the tests for this are chi-squared tests which would not give special treatment groups which are actually ordinal e.g. low income, medium income, high income: the chi-squared test is weaker than it should be.

Another useful section is a summary map for each question, as in the picture.

image-20211117170742108