# Correlation and Causation

Correlation: two variables are positively correlated when an increase in one is accompanied by an increase in the other
• Large amounts of data are needed to ensure that the correlation is statistically significant
o This ensures that it didn’t happen by chance
• Causation: two variables are causally linked when a change in one is responsible for a change in the other
• Scientists need to set up experiments with control variables
o This means they can see if altering one variable has the predicted effect To do this a null hypothesis is set up
o They assume that there will be no difference between an experimental group and the control group
o This is then testes using statistical analysis

COHORT STUDIES
• Follow a group of people over time to see who develops the disease and who doesn’t
• People’s exposure to risk factors are recorded
o This means that any correlations between risk factors and the disease can be established
• It may take a long time to develop the disease, so these studies can take years=£££ expensive
CASE CONTROL STUDIES
• A group of people with the disease (cases) are compared with a control group of individuals who do not have the disease
• Information is collected about the risk factors they have been exposed to
o This allows factors that may have contributed to development of the disease to be identified
o The control group should be representative of the population from which the case group was drawn
o Sometimes controls are individually matched to cases
▪ Known disease- risk factors: such as age and sex are then similar in each case and control pair
• This allows scientists to investigate the potential role of unknown risk factors
o Factors used to match the cases and controls can’t be investigated in the study
o So, it’s important not to match on variables which could potentially turn out to be risk factors
FEATURES OF A GOOD STUDY
CLEAR AIM:
o A well-designed study needs a clearly stated hypothesis or aim
o The design of the study must be appropriate to the stated hypothesis to produce results that are valid and reliable
REPRESENTATIVE SAMPLE:
o Selection bias occurs when those who participate in a study are not representative of the target population

o Differences between people asked to take part in a study and those who actuall respond should be considered before generalizing findings to the target population
o Non-participants can differ in important respects from participants
o The proportion of people who drop out of a study after it has begun should be kept to a minimum
o This is particularly important in cohort studies that follow people over a long period of time
o It is important to monitor the characteristics of the remaining participants to ensure that they are still representative of the target population

VALID AND RELIABLE RESULTS
• Methods used must produce valid data from measurements that provide information on what the study is set out to measure
• For example, a study on the effect of blood pressure on development of CVD
o Valid blood pressure measurements would be made using an appropriate blood pressure monitor
• The method used to collect results must be reliable
o A reliable method used at different times or by different people will produce similar results
o A reliable test will also give similar results for repeated measurements
o E.g. if measuring blood pressure, the same type of equipment and same type of procedure should be used each time the measurement is made
o any variables that could affect the measurement should be controlled or taken into account
• a sample must be large enough to produce results that have not occurred by chance
• in cohort studies of a rare disease only a small proportion of the population will develop the disease
• in case control studies only, a few people may have been exposed to the factors under investigation
• the potential effect of all variables that could be correlated with the disease should be considered when designing the study
• Risk: the probability of occurrence of some unwanted event or outcome
PERCEPTION OF RISK
• People will overestimate the risk of something happening if the risk is:
o Involuntary (not under their control)
o Not natural
o Unfamiliar

o Unfair

o Very small
• There is a tendency to overestimate the risk of sudden imposed dangers where the consequences are severe
o And to underestimate the risk if it has a long-term effect on the future, even if the effect is severe
o E.g. health risk associated with smoking or poor diet
DIFFERENT TYPES OF RISK FACTOR
• Averages can make the assumption that everyone may have the same chance of having CVDwhich isn’t the case
• Averages do not take into account any risk factors
o Things that may increase the chance of a harmful outcome
• There are 4 main factors that contribute to health risks
o Heredity
o Physical environment
o Social environment
o Lifestyle and behavior choices