Limitations in correlational studies

Question 1

No causation

Answer 1

Correlations cannot be interpreted in terms of causation

Question 2

The third variable problem

Answer 2

There is always a possibility that a third variable exists that correlates both with X and Y and explains the correlation between them

X (number of salons) may have correlation with Y (number of criminals), but once you take the third variable Z (high population) into account, the correlation mecobes meaningless.

Question 3

Curvilinear relationships

Answer 3

Sometimes variables correlate non-linearly.

X (arousal) may correlate with Y (performance) to a certain point, until correlation begins to decrease; from correlation to negative correlation.

Performance is best when arousal is average, but this is observable only through a graph. Correlation coefficients a linear, thus correlation may end up being medium to low.

Question 4

Spurious correlation

Answer 4

When a research study involves calculating multiple variables, there is a possibility that some of the statistically significant correlations would be the result of random chance. r= .05 (5%) is statistically significant so there is a 95% chance that the correlation is an artifact.

Question 5

Cure for curvilinear correlation

Answer 5

If a curvilinear relationship is suspected, researchers should generate and study scatter plots

Question 6

Cure for the third variable problem

Answer 6

Researchers should consider potential variables in advance and include them to the research in order to explicitly study the links between X and Y and the third variable.