Correlation (L7) Flashcards
what is a correlation?
a measure of how well 2 variables work together
in what situations would you use correlation?
-when we have 2 continuous variables and want to see if they are related to each other
-ex. scatterplot
what are the 3 outcomes of correlation?
positive
no correlation
negative correlation
what 2 things do we need to know to quantify a correlation?
- covariance
- correlation coefficient
We are about how to calculate Pearson’s
correlation coefficient as a statistical test; this
test has some assumptions that we need to
check first, to make sure we can use the test
5 points:
*Continuous Variables
*Homoscedasticity (equal variances in both
variables)
*No Outliers
*Related Pairs (the values come from the same
subject)
*Linearity (the relationship between the values is
best described by a line)
what does A positive covariance mean
as one
variable deviates from the mean, the other variable deviates in the same
direction.
what does a negative covariance mean
as one variable
deviates from the mean (e.g.,
increases), the other deviates from
the mean in the opposite direction
(e.g., decreases).
what does a covariance number mean??
We cannot compare covariance in an objective way: It isn’t standardized
We cannot compare the covariance we found in
our dataset to another dataset unless they are
measured in units with a similar range of values.
You can imagine variables where the same covariance value can have very different
meanings; think things like annual salaries
(possibly 100’s of thousands) vs. temperature
(single degrees, even fractions of a degree).
what do correlation coefficients mean(standardizing covariance)
+1 = perfect correlation (one up, the other up)
-1 = perfect negative correlation (one up, the other down)
0 = no linear relationship
So in the end, we get a standardized measure of an observed
effect.
± .𝟏 = small effect
± .𝟑 = medium effect
± .𝟓 = large effect
