Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
23 Cards in this Set
- Front
- Back
what is the purpose of regression? |
compute future outcomes from present ones |
|
what must be done first before a regression analysis? |
1. compute correlation 2. create regression equation 3. plot regression line |
|
what is the focus on when using regressions? |
DV being predicted from the IDV |
|
regression line |
best guess as to what score on the DV or Y variable would be predicted by a score on the IDV or X variable. line of best fit minimizes distance between each individual point and that regression line. |
|
Y variable |
DV |
|
X varibale |
IDV |
|
X variable is also known as what? |
abscissa the predictor variable (IDV) |
|
error in prediction is know as what? |
residuals |
|
what do the means of residuals equal to? |
zero |
|
what is a direct reflection of the strength of the correlation between the variables? |
the difference between each data point and the regression line. -error of estimate or error in prediction |
|
standard error of the estimate |
SD of all the residuals how much imprecision there is in the estimate the better the correlation the less error |
|
Bivariate Regression Formula |
Y'=a+bX Y' = predicted score b = slope, or direction of the line a = the point at which the line crosses the y-axis intercept value of Y' where X=0 X = the score being used as the predictor (IDV) |
|
b (the slope) signifies what? |
how many predicted units change (+ or -) in the DV there are for any one unit increase in the IDV. - example: if b = 3 then DV will increase by 3 units for every 1 unit of the IDV. |
|
r is high if.... |
...scattergram points are close to regression line - quantifies degree to which the actual points match up to predicted |
|
null hypothesis for simple regression |
Ho:b=0 |
|
when is multiple regression used? |
when more than one predictor variable (IDVs) or control IDV is being used as variables. |
|
Multiple Regression Formula |
Y' = a+b1X1+b2X2 |
|
what type of coefficients do you use if you want to compare the contributions (to explain or predict DV) of different IDV to the regression equation. |
standard beta coefficients tells how much variation is explained by each independent variable. |
|
how are standard beta coefficients created |
z-scores |
|
general rule of thumb for additional IDVs being added to regression equation? |
more than 3 or 4 IDV and you run the risk of them being correlated. Lose power of prediction. -IDVs should be uncorrelated with each other |
|
Types of Multiple Regressions |
1. Simultaneous - all IDV entered at once 2. stepwise - entered by size of bi-variate correlation between the IDV and the DV 3. Hierarchical - entered in stages. control variables normally entered first |
|
R squared |
used to report the percentage or proportion of the variability in the DV that is accounted for by the combination of the IDVs in the equation. |
|
null hypothesis for multiple regression |
Ho = R2 = 0 - none of the variance is DV is explained by combination of |