correlations & experiments

• what do correlations measure?

  A relationship/associations between two naturally changing co-variables.
• when using correlations can the variable be manipulated?

  No. When using correlations you cannot manipulate the variable(s) - correlations are used for relationships not differences.
• can you determine a cause and effect relationship with correlations?

  No. A cause and effect relationship cannot be determined when using correlations.
• the three key factors of a correlation

  -continuous data

  -two sets of data

  -relationships
• what do experiments measure?

  Experiments measure a difference. They look at how the independent variable changes the dependent variable.
• when using experiments can the variable be manipulated?

  Yes. In experiments the independent variable is manipulated. This is done to be able to see a difference in the dependent variable.
• can you determine a cause and effect relationship with experiments?

  Yes. A cause and effect relationship can be determined when using experiments. This is because experiments are highly controlled and scientific.
• correlations - hypotheses

  Correlations have an alternative and null hypothesis.

  Correlation hypotheses use language that relates to a relationship and has both co-variables (operationalised) in it.
• differences between correlations and experiments

  In experiments the researcher manipulates/controls the independent variable, in the correlation the variable is not manipulated.

  Experiments can measure a cause and effect relationship - correlations cannot.

  In experiments the hypothesis states a difference, in correlations the hypothesis states a relationship/association.
• experiments - hypotheses

  Experiments have an alternative and null hypothesis.

  Experiment hypotheses use language that relates to a difference, include all levels of the IV and the DV (operationalised).
• correlations - coefficients

  A statistical test that gives a numerical value between -1 and +1. Coefficients show us the strength and direction of the relationship between two variables.
• positive correlation coefficient

  Above +0 to +1.
• negative correlation coefficient

  Below -0 to -1.
• different types/levels of correlations based on the coefficient

  -strong positive

  -weak positive

  -strong negative

  -weak negative
• what numerical value is needed for something to be strong (positive or negative)

  The numerical value of 0.70

  (+0.70 and -0.70 = strong)
• AO3 - correlations/coefficients

  strength(s):

  -objective. The results can only be interpreted in one way.

  -correlations are easy to read/understand.

  -cheap to do (secondary data can be used + do not have to pay for things like labs).

  
  

  weakness(es):

  -third variable problem. Correlations only show two variables when there might be more variables involved that make the results what they are.

  -the direction of causality cannot be determined (cannot tell what caused what).

  -cannot see/tell the cause and effect.
• causation - correlation coefficient

  "Correlation does not imply causation."