Question | Answer |
True/False: The predicted score is almost always different from the actually scores. | True |
True/False: The dependent variable is always continuous. | True |
Categorical vs. Continuous variable | Categorical: Difference in classification. Continuous: Differences in amount. |
How is the residual or error calculated (in predicted scores)? | - Subtract the predicted score from the observed score. |
Give an example of both direct and indirect effect on a regression line. | Direct: calories vs. fat. Indirect: Olympic 100m times vs. successive Olympic games. |
When is the relationship between 2 variables considered to be an indirect effect? | When there is reason to believe that another variable mediates the relationship between the independent and dependent variable. |
The sum of squares of the predicted scores is called... | Regression sum of squares |
If all of the observed scores equal the predicted scores, what is the value of the sum of squares residual? | SSres = 0 |
How can you calculate the SSreg using SSy and SSres? | SSreg = SSy - SSres |
What is a useful index for the goodness of fit of the regression line? What does a value of one indicate? What about zero? | Ratio of the sum of squares of the predicted scores divided by the sum of squares of the observed scores. Has a maximum value of one (perfect fit), zero means there is no relationship. |
Why is the goodness of fit ratio given the symbol r^2? | Because it is equal to squared simple correlation. |
How is the goodness of fit ratio calculated? | r^2 = SSreg/SSy |
True/False: The goodness of fit ratio always varies between -1 and 1. | False (It is always less than 1, but greater than 0) |
When you represent 2 or more categorical variables as values of 1 and 0. | Dummy coding |
When you use dummy coding, what is always the value of the predicted scores? | Mean for the category |
When using dummy coding, how do you calculate the constant of the regression line? | a = M0 |
When using dummy coding, how do you calculate the slope? | b = M1 - M0 |
Want to create your own Flashcards for free with GoConqr? Learn more.