Correlation Flashcards
Correlation
A form of research in psychology in which we are trying to calculate if there is a significant relationship between 2 covariables
Co-varibales
The 2 variables that are being measured to see if there is a significant relationship affecting each other
Research question
A broad question shout the concept being investigated eg:
Are people more _ when it is _?
Research aims
Having identified the broad question, the research aim is then identifying a more specific concept within the researcher wants to investigate
Are there independent/dependent variables in correlational studies?
No because we are not measuring the cause (iv) and effect (dv) but if there is a relationship between 2 co variables
How must co variables be operationalised in correlation?
Must be quantitative data
For each participant: their 2 data values must be ordinal or above so can be put on a scale. This way a correlation can actually be measured
What’s important in writing hypotheses for correlation?
There will be a significant RELATIONSHIP (not effect)
Between 2 co variables
A one tailed alternative hypothesis (correlation)
There will be a significant positive/negative correlation between co variables X and Y
A two tailed alternative hypothesis (correlation)
There will be a significant correlation between co variables X and Y
A null hypothesis (correlation)
There will be no significant correlation between co variables X and Y. Any relationship is due to chance factors
How are the results of a correlation study shown?
On a scatter graph
What does each cross on the scatter graph of a correlation study results show?
A different pair of values
Eg on the x axis, the participants score in one co variable (how far long it is) and on the y axis, the participant score in other vi variable (how high it is)
What does the total number of crosses on a scatter graph tell us?
The size of the data the researcher has collected eg sample size (each represents a participant)
A positive correlation on a scatter graph
X axis Increases as y axis increases
A negative correlation on a scatter graph
X axis increases as y axis decreases
What do scatter graphs not tell us?
That co variable x has an EFFECT on y (we can only establish a relationship)
What hypothesis is retained or rejected, only do this through inferential statistics eg finding the correlation co efficient
Findings vs conclusions in correlation studies
Findings = refers to raw data of either co variable (mean, median, mode, range, outliers)
Conclusions = broad inferences made from raw data eg direction of correlation if it’s positive or negative + its strength
Correlation co efficient
A value obtained from inferential statistical tests which allows us to determine how strong/weak the correlation is, and what direction (conclusion) based on its value
A positive correlation co efficient
Shows a positive correlation (as one co variable increases, the other co variable increases)
A negative correlation co efficient
Shows a negative correlation (as one co variable increases, the other decreases)
A correlation co efficient close to 0
Shows that there is no correlation
As the correlation co efficient moves away from 0…
The stronger the correlation is
(Further into negative = stronger negative correlation)
Further to 1 = stronger positive correlation)
What is the maximum for correlation co efficient?
Ranges from -1 to +1
What does the correlation co efficient allow us to do?
Retain a hypothesis or reject it (if there is a significant relationship or no significant relationship)
Advantages of correlational studies
Allows us to measure the relationship between variables that are too unethical to manipulate
Test for inter rater reliability in experiments
Starting point for research (establish relationship)
Find something new (new relationship) of pre existing data
Disadvantages of correlational studies
Correlation =/= causation
Inferential statistics don’t always pick up on a relationship eg a might have a correlation coefficient of 0 when there is still related
No qualitative data is collected
