Student Bump up Your Grade

Student bump up your grade

Name ...... Class ...... Date ......

Keywords / Correlation, causation, bias, fallacy, placebo, double-blind, randomised controlled trials, peer review

Correlation and causation

Specification references

B2.2.6 The effect of lifestyle on some non-communicable diseases

Aims

Medical scientists often have to discern the factors that may lead to a person getting a disease. To do this they have to analyse data. As a consequence of thisthey may see a link between certain patterns of behaviour and the likelihood of the occurrence of a disease. However, this does not necessarily mean that one is causing the other. This exercise will help you to understand the difference between correlation and causation.

Learning outcomes

After completing this exercise, you should be able to:

explain the meaning of the words correlation and causation
distinguish between the two

describe the importance of maintaining this distinction.

Task

First ensure that you have read Topic B7.1, Non-communicable diseases in the student book.

Then read the passage below and answer the questions.

It is incorrect to assume that, just because two occurrences happen simultaneously, one occurrence has caused the other. There are several different ways in which assumptions of causality can be incorrect.
Windmills turn faster when it is windy, but that does not mean that they make the wind. This would be an example of reverse causation – getting it the wrong way round.
There may also be a common factor causing both occurrences. For example, a medical research study found that young children who sleep with the light on are more likely to develop myopia (short sightedness) as they grew older. The conclusion was reached that it was this that was causing the eye problem. However, a later study found that this was not the case. They found that myopic parents are more likely to leave the light on so they can see their sleeping infants, and that an increased likelihood of developing myopia can be genetically inherited. So one cause was resulting in both effects.
Occurrences may also be linked so that the phenomena both affect each other. A good example of this from biology would be predator-prey relationships, where the numbers of a prey species affect the numbers of a predator species and vice versa.
Another source of error can be that the apparent correlation is in fact just coincidental. For example, there is a greater than 95% correlation between the number people who drown after falling out of a fishing boat and the marriage rate in Kentucky. The two occurrences are not connected, but a correlation can be used to suggest that they are.
In order to overcome the problems that misinterpretation of links can create, medical scientists have developed systems of ensuring that an apparent correlation is valid. Randomised controlled trials (RCTs) are commonly used. In these, the allocation of participants in a group trial is randomised and often carried out in a concealed manner so that neither the patients nor the scientists or doctors carrying out the trial know who has received the real treatment and who has received the placebo (fake treatment designed to appear exactly the same as the drug itself). These trials are often referred to as being “double-blind”.
Fear can lead the population to accept anecdotal evidence, as was demonstrated in the antivaccination scares regarding MMR in 1998. Parties with ulterior motives or vested interests may try to use correlations for their own ends. They may also place undue emphasis on the fact that correlation does not in itself constitute proof. Tobacco companies have used this to argue that lung cancer can correlate with cigarette smoking but may not be caused by it.
Amassing evidence from a number of different studies (triangulation) and from projection can provide weight to an appeal for acceptance of causation. Publication of research is another part of the procedure, as it allows the scientific and medical community to scrutinise and repeat any trials. Peer review is essential.

Questions

1Give another example of a reverse causation fallacy.

(1 mark)

2Describe the mistake the researchers made when they stated that leaving the light on at night causes myopia.

(1 mark)

3Why we should expect unexpected occurrences of correlation to occur.

(1 mark)

4Give an example of circumstances where a randomised controlled trial would be appropriate.

(1 mark)

5The use of placebos can sometimes have a statistically significant positive effect. Suggest why this might be the case.

(2 marks)

6Explain why some people might deliberately want to overemphasise the relative weakness of correlation to determine causation.

(1 mark)

7Suggest why being “double-blind” is an advantage in RCTs.

(1 mark)

8Describe what may be meant by use of the word ‘projection’ in the final paragraph.

(1 mark)

9Suggest any drawbacks that are inherent in the process of peer review as far as drug development is concerned.

(2 marks)

10Explain why the media might have ulterior motives regarding the relationship between causation and correlation.

(1 mark)

This resource sheet may have been changed from the original.1