Correlations

Experimental methods
 Manipulation of IV and DV, control
 Lab, field, natural/quasi
Non-experimental methods
 Correlations
 Interviews and questionnaires
 Observations
 Case Studies
Children who watch television shows that
contain a lot of violence are more likely to
show violent behaviour. Therefore violent
television causes violent behaviour
It is raining outside. You see lots of people
with umbrellas. Therefore, you argue that
whenever people put up umbrellas, it
causes rain.
What is wrong with these arguments?
What is a correlation?
 A relationship between two variables.
 (Eg; TV viewing & intelligence)
Correlational design
 Participant provides data for both variables.
In a correlational design, there are no
independent or dependant variables, but
co-variables.
When do we use a correlational design?
 To test a hypothesis about a relationship
between two variables, for example:
When looking for a relationship that would
be unethical to manipulate for an
experiment, for example:
Decide what co-variables you will
be measuring, and
operationalise your variables.
Measure each participant on both
co-variables.
Plot the values on the scatter
graph to see if there is a
relationship. Carry out a statistical
test to see if the relationship is
significant or due to chance.
The correlation coefficient is measured
-1 to ___
+1
from ___
A correlation can be positive, negative or
no correlation. The larger
_____ the number, the
stronger the correlation
Have a go at matching the coefficients to
the scatter graphs
As one variable
increases, the other
also increases
Between 0 to +1
For example: the
taller a person is the
heavier they are likely
to be
As one variable
increases, the other
decreases
Between 0 and -1
For example: the
more alcohol you
drink, the less you
are able to
remember
There is no
relationship between
the variables
Close to 0
For example, shoe
size and IQ
Work in pairs and have a go at activity 4.
Say what type of correlation you would
expect and why.
You have performed a statistical test, and
found that two variables are correlated
(there is a relationship).
What does this tell us about cause?
When looking at correlation, it is important
to remember that:
CORRELATION DOES NOT MEAN
CAUSATION
 Alternative Hypotheses:
 Directional Hypotheses should state if the
relationship between variables is positive or
negative:
 Eg: stress & caffeine consumption:
 There will be a positive correlation between p’s
stress levels and number of cups of coffee drunk
in a day.
 Non-Directional Hypothesis:
 There will be a correlation between between p’s
stress levels and number of cups of coffee drunk
in a day.
 Null Hypotheses:
 Stress and caffeine consumption:
 There will be no correlation between p’s stress
levels and number of cups of coffee drunk per
day.
 NB: Good hypotheses always indicate how variables are
operationalised
 Alcohol and driving accidents
Violent computer games and aggressive
behaviour
Sleep and problem solving
Homework and GCSE results
Watching TV and intelligence
 In recent years there was much discussion in
the media about a link between the MMR
vaccine and autism.
 A report showed that there was a correlation
between children who received the MMR
vaccine and the development of autism.
 The authors of the report claimed that this
showed that autism was caused by the MMR
vaccine.
 How else could the results be interpreted?