Correlation

correlation

statistical technique that is used to measure and describe the relationship between two variables

variables

in correlation, its simply OBSERVED as they exist naturally in the environment

correlation

no attempt to control or manipulate the variables

cause & effect

correlation does not include _____ & _____, unlike experimental

perfect correlation

identified by a correlation of one

perfect correlation

indicates a perfectly consistent relationship

perfect correlation

in this type of correlation, each change in X is accompanied by a perfectly predictable change in Y

correlation 1.00

degree of consistency in change:

perfect correlation (#)

correlation 0.00

degree of consistency in change:

correlation that indicates no consistency at all

correlation 0.00

degree of consistency in change:

data points are scattered randomly with no clear trend

degree of consistency in change

0.00 = less perfect correlation

1.00 = more perfect correlation

correlation coefficient

expresses quantitatively the magnitude and direction of the relationship

describes

correlation simply ________ a relationship between 2 variables

correlation

does not answer the question "why?"

correlation

its value can be affected by the range of scores represented in the data

outliers

one or 2 extreme data points

outliers

can have a dramatic effect on the value of a correlation

restriction range

correlation within _______ could be completely different from the correlation that would be obtained from a full range of scores

restriction range

if its used, it needs to be mentioned

outliers

an individual with X and/or Y value that is substantially different from the values

square

to describe how accurately one variable predicts the other, you must _______ the correlation

proportion

correlation should not be interpreted as a ________

coefficient of determination

squared correlation

coefficient of determination

measures the proportion of variability in the data explained by relationship between X and Y

coefficient of determination

measures the gain in accuracy that is obtained from using the correlation for prediction

3-5

suggested maximum number of outliers (range)

larger

sample size smaller = need magnitude of correlation _________

small sample

requires a VERY LARGE CORRELATION before one can be confident there is a real nonzero rs in the popu