No graph needed (probably)

From the educational welfare officer in a school: (fuzziness is from the original)

With the helpful commentary:

"It clearly shows that those students with attendance rates below 85% achieved far fewer GCSE’s at grades A to C than those with regular attendance." -mmMMM

Firstly, pedant mode - This is an awful chart - badly executed and the legend:
What if my attendance was 85% -- which box do I go in? Basic rules of labelling -- make sure your chart actually says what you think it does.

Sticking to my rules about horizontal axes showing numerical, not categoric data: (I made a call on 85% or hinger)
Now, I don't like that, with the repeating colour patterns vertically - so lets do a back to back comparison instead:
More worryingly, is the commentary actually correct? Does the original chart actually show what they think it does?
There are 49 learners with less than 85% attendance. Of the 162 learners who have above 85% attendance, 88 have C grades or higher. So even if ALL the less than 85% learners got C or above (a great result)- the statement as written would still be correct - but I'm sure that's not what they meant when they wrote it.

We need to compare percentages so that the cohort size does not matter.
Or back to back:

I wonder if a chart / distribution is needed at all - after all, if we are obsessing with "C's and above", we can reduce this to:

  • 54% of learners with attendance of 85%+ get C compared to only 20% for those less than 85% attendance.

OR: "Learners with 85%+ attendance are more than 2.5X more likely to get C+ grade in maths"

comments powered by Disqus