Correlation Flashcards
What is a two tailed hypothesis?
It is a hypothesis in which you don’t know the direction that the significance will be go in, ex. There will be a significant correlation from ethnic minority groups appear on television in one day across BBC, etc etc Must be operationalised.
What is Quantitative data?
It’s data that is countable
Types of Quantitative data
Amount
Time
Score
Can correlation analysis be done of qualitative data?
No, only on quantitative data as it is a statistical process
What is the correlation co-efficient?
It’s the number you work out and it’s between -1 and +1
How is correlation shown graphically? How are they categorised?
Using a scattergram. They are also categorised into positive, negative and zero correlations
What is a one-tailed hypothesis?
It is a hypothesis in which you know which direction the correlation is going to go ex. There will be a significant negative correlation between the number of times ethnic minority groups are shown on TV and level of racism as measured by a score on a questionnaire. The variables must be operationalised
What is a perfect positive correlation? What does this mean?
It has a coefficient of +1. This means one variable increases at exactly the same rate as the other increases
What is a perfect negative correlation?
It has a co-efficient of -1. That means one variable decreases at exactly the same time as the other
What is a zero correlation?
It means there is no clear relationship. A true zero correlation is 0.0
Strengths of a scattergraphs?
Correlations may show a relationship between two variables that was not expected, this can lead to further research and possible experimentation
They can be used to determine a relationship when it is not possible to carry out an experiment
They can be used to determine a relationship when it is not possible to carry out an experiment, such as the research on twins to see it schizophrenia has a genetic element
Fairly easy to analyse using Spearman’s compared to case studies where data is a lot more complex
It’s easy to read because it is visual and the trend/ relationship is clear
Weaknesses of correlations
A correlation does not tell us about cause and effect, so we don’t know if violent television causes aggressive behaviour or vice versa
The results may not be valid as the measurements are often self report and subjective, e.g. How can the benefits of psychoanalysis be measured in an objective way
A correlation does not tell us cause and effect, there may be a third factor affecting the results other than the variables of interests
Is a correlation study an experiment
No
What do correlations allow researchers to do?
Allow the, to statistically analyse naturally occurring events that could not be set up experimentally
