When I first saw those horrendous venn diagrams on junkcharts, I considered sending in a bar chart similar to this. I think I would only adjust the category order along the vertical, placing Clinician third after Test A and Test B, so the single methods are listed together above the double methods.
I was thinking of something along the same lines, but I think Igor also wanted to highlight the inefficiency of the tests (the percentages in parens in the original venn diagram).
I also do prefer the bar chart to both former plots though i'd rather use absolute values for subsample sizes than percentages. I.e. in this version it's impossible to see that those examined by only a clinician are in fact only two individuals.
@Nathan:
I'm not quite sure how good a conclusion about the accuracy of the tests can be drawn from the data since to me it's not clear what a combination of tests (and clinician) means. Does it mean that in a combination all parties come to a positive diagnostic or that the results of the involved tests or clinician might contradict each other. In the first case this would mean that the accuracy of the tests actually isn't that horrible somewhere around 70-80%, while as long as we don't know how results for example Test A gave in the combination cases it might well be that simply flipping a coin would have given a more accurate result.
I've now made a stab at it myself, presenting the numbers as a pair of stacked tornado charts, one each for autism proper and autism spectrum disorder diagnoses.
That one assumes an exclusive diagnosis, in other words, if it says the clinician and the interview agree, that means the observation dissents. This one includes the unanimous diagnoses as well, in other words it's as if the third measure were never taken at all. It's a little harder to see that the numbers are the same as the Venn diagram in this version, but on the other hand it lets you answer the question "how would the diagnoses differ if I had only used two of the three measures?"
moveabletype keeps making the graphs extend beyond the white border of the blog, and the unfortunate use of frames means that most links to the outside web open in the Statistical Modeling… frame, which makes them have the wrong title and URL. Can I vote for another template or different blog software?
I revised the horizontal bar chart based on the comments here and Dereks' comments on JunkCharts blog.
Derek, you can see my latest revision and post your version @ perceptualedge.com under discussion / graph design / alternative to venn diagram.
I have had a look at the original venn diagram created, as well as some of the ones created in response to a better way to display the data and I would have to say that this graph is by far the most efficient and easy to read.
Edit to this post:
Test A – Pre Linguistic (PL-ADOS)
Test B – Interview (ADI-R)
When I first saw those horrendous venn diagrams on junkcharts, I considered sending in a bar chart similar to this. I think I would only adjust the category order along the vertical, placing Clinician third after Test A and Test B, so the single methods are listed together above the double methods.
I was thinking of something along the same lines, but I think Igor also wanted to highlight the inefficiency of the tests (the percentages in parens in the original venn diagram).
I also do prefer the bar chart to both former plots though i'd rather use absolute values for subsample sizes than percentages. I.e. in this version it's impossible to see that those examined by only a clinician are in fact only two individuals.
@Nathan:
I'm not quite sure how good a conclusion about the accuracy of the tests can be drawn from the data since to me it's not clear what a combination of tests (and clinician) means. Does it mean that in a combination all parties come to a positive diagnostic or that the results of the involved tests or clinician might contradict each other. In the first case this would mean that the accuracy of the tests actually isn't that horrible somewhere around 70-80%, while as long as we don't know how results for example Test A gave in the combination cases it might well be that simply flipping a coin would have given a more accurate result.
I've now made a stab at it myself, presenting the numbers as a pair of stacked tornado charts, one each for autism proper and autism spectrum disorder diagnoses.
That one assumes an exclusive diagnosis, in other words, if it says the clinician and the interview agree, that means the observation dissents. This one includes the unanimous diagnoses as well, in other words it's as if the third measure were never taken at all. It's a little harder to see that the numbers are the same as the Venn diagram in this version, but on the other hand it lets you answer the question "how would the diagnoses differ if I had only used two of the three measures?"
moveabletype keeps making the graphs extend beyond the white border of the blog, and the unfortunate use of frames means that most links to the outside web open in the Statistical Modeling… frame, which makes them have the wrong title and URL. Can I vote for another template or different blog software?
I revised the horizontal bar chart based on the comments here and Dereks' comments on JunkCharts blog.
Derek, you can see my latest revision and post your version @ perceptualedge.com under discussion / graph design / alternative to venn diagram.
I have had a look at the original venn diagram created, as well as some of the ones created in response to a better way to display the data and I would have to say that this graph is by far the most efficient and easy to read.