How not to understand mathematical language
Returning home late this evening I was listening to "Today in Parliament" on BBC Radio Four and heard a classic example of how mathematical and scientific language can come over to the lay person as more alarming than is really appropriate.
The BBC was reporting on evidence given to parliament by a distinguished professor who was talking about the impact of pollution on various health problems.
Referring to the statistical correlation between pollution levels and increased incidence of these diseases he drew what the BBC announcer desicribed as an alarming conclusion.
As edited the broadcast did not explain with complete clarity exactly what the professor was talking about (I suspect his full speech probably did.)
However, to me as a statistician it was evident that he has to have been talking about the best fit line of correlation between the increased incidence of disease and the number of particles of pollution per unit of atmosphere when he said
"The line ... goes through the origin. There is no safe level."
(I was driving and not in a position to record the exact words used: they included all the words quoted in inverted commas above and the first sentence may have been something like this - the line of correlation between pollution and disease goes through the origin.)
I've no problem with what the professor said. I do have a problem with the way the BBC presented it.
The way the presenter emphasised how alarming the professor's evidence was suggests that she may have fallen into the trap of misinterpreting the comment above and particularly the words "There is no safe level" as .meaning "Even an infinitesimal amount of pulltion is highly dangerous."
It is certainly likely that a listeners with no knowledge of correlation or the graphical depiction of relationships would interpret the broadcast that way and therefore be under a misapprehension about what the professor's evidence means.
In the context that the best fit line of correlation goes through the origin, the words "there is no safe level" do NOT mean that even an iufinitesimal level of pollution is highly dangerous, it means that there is no positive level of pollution at which the effects are zero.
The words "The line goes through the origin" imply that a very tiny amount of pollution does a very tiny amount of harm.
One item of very good news this week was that judges are to be provided with guides to scientific issues written in accessible language with input from people like the Royal Society to help them understand scientific concepts relevant to decisions which courts have to make.
One of the first such guides will cover DNA fingerprinting, and the third, I am very pleased to learn, will cover statistics.
I really wish someone would make journalists read something similar.
The BBC was reporting on evidence given to parliament by a distinguished professor who was talking about the impact of pollution on various health problems.
Referring to the statistical correlation between pollution levels and increased incidence of these diseases he drew what the BBC announcer desicribed as an alarming conclusion.
As edited the broadcast did not explain with complete clarity exactly what the professor was talking about (I suspect his full speech probably did.)
However, to me as a statistician it was evident that he has to have been talking about the best fit line of correlation between the increased incidence of disease and the number of particles of pollution per unit of atmosphere when he said
"The line ... goes through the origin. There is no safe level."
(I was driving and not in a position to record the exact words used: they included all the words quoted in inverted commas above and the first sentence may have been something like this - the line of correlation between pollution and disease goes through the origin.)
I've no problem with what the professor said. I do have a problem with the way the BBC presented it.
The way the presenter emphasised how alarming the professor's evidence was suggests that she may have fallen into the trap of misinterpreting the comment above and particularly the words "There is no safe level" as .meaning "Even an infinitesimal amount of pulltion is highly dangerous."
It is certainly likely that a listeners with no knowledge of correlation or the graphical depiction of relationships would interpret the broadcast that way and therefore be under a misapprehension about what the professor's evidence means.
In the context that the best fit line of correlation goes through the origin, the words "there is no safe level" do NOT mean that even an iufinitesimal level of pollution is highly dangerous, it means that there is no positive level of pollution at which the effects are zero.
The words "The line goes through the origin" imply that a very tiny amount of pollution does a very tiny amount of harm.
One item of very good news this week was that judges are to be provided with guides to scientific issues written in accessible language with input from people like the Royal Society to help them understand scientific concepts relevant to decisions which courts have to make.
One of the first such guides will cover DNA fingerprinting, and the third, I am very pleased to learn, will cover statistics.
I really wish someone would make journalists read something similar.
Comments