W11

Question 1

correlation coefficients

Answer 1

1) test-retest: correlation of scores from the same test at two points in time
2) parallel forms: correlation between scores on different forms of the same test
3) internal consistency: correlation between different parts of a test
4) reliability is basically the correlation of a test with itself

perfect correlation = -1 or 1, 0 = no correlation

Question 2

prediction

Answer 2

to predict one variable from another, compute the correlation between the two variable first

Question 3

multiple regression

Answer 3

when you predict more than one outcome from independent variables

rules for when using multiple predictor variables:
1) select a predictor variable (x) that is related to the criterion (y)
2) the two share something in common
3) when selecting more than one predictor variable, select variables that are independent or uncorrelated with one another but are both related to the outcome or predicted Y variable

Question 4

Pearsons r

Answer 4

r - expresses the direction and strength of linear correlations with one single number

scatter plot - help us assess whether a correlation is strong or weak but it does not tell us exactly how strong it is

+ R = positive correlation
- R = negative correlation
always lies between -1 and 1

Question 5

compute Pearsons r

Answer 5

1) standardize values, turn them into z scores
2) compute mean for both variables x,y
3) compute steeds for both variables
4) multiply each z score of x by each z score of y and sum the values
5) divide this value by n-1

Question 6

finding the line

Answer 6

residuals - distance from the point to the line
you can end up with positive residuals, from cases above the line, to the line and negative residuals, from cases below the line to the line

Question 7

correlation is not causation assumptions

Answer 7

IV x can explain the DV y
- predict the values of the dv based on the values of the iv
- r^2 allows us to assess how well the line fits our data
- never be certain that one variable is the cause of the other, be aware of outliers
- if there is a third variable (z) included in the study that may be affecting the results it is called a confounding variable