Flashcards
RDA2- Correlation to Regression
Flashcards do conjunto (12)
- 1What is data?Model + error
- 2Is the mean a model?Yes, the mean is an estimate that provides an average score
- 3What is the Standard error?The spread, so this indicates the error
- 4What would be a good model (mean)?If the model is small, so it is a pretty good fit. If the mean is large, it is a bad fit.
- 5What is a regression line?Another model, the regression is the estimate of the relationship between two things
You can analyse the variation of scores from the regression line- if variation is smaller with regression line, it is more accurate - 6What is the gradient?The angle of the slope, when you know where the line starts (deviation) and its gradient, you can plot the gradient
- 7What is the difference between mean and regression?The mean is the most simple model, whereas regression tries to be more accurate with prediction of scores
- 8What is the simple regression?Outcome= predictor
Y = a + (b*X)
DV = a + (b*IV) - 9What does the a and b mean in the equation DV= a + (b*IV)a- where the line crosses the y axis (The intercept)
b- how much you have to multiply the x value (The gradient) - 10How do we test simple regressions?We use ANOVAs
- 11How do you work out how much better the regression line is than the mean for predicting scores?Work out the difference between the predicted value from the regression and the predicted value from the mean (the simplest model), for each pps (and square and add them)
- 12How do you work out how much error there is in the regression model?Work out the difference between the predicted value from the regression line and actual value (and square and add them)