PSYC 200W Lecture Notes - Lecture 11: Matching Law, Linear Regression, Regression Analysis
28 Jun 2018
School
Department
Course
Professor
Regression is used to study correlations further and develop hypotheses about these relationships
Regression analysis - ability to predict scores on one variable from 1 or more other variables
Regression equation - helps to predict scores on one variable on the basis of the scores of another
variable
We find the line that best fits the data which will help us create an equation to predict the variables
"Advanced correlation study"
Assess correlations between variables
Run regression predicting one variable from > 1 others
Slope intercept formula -
M = slope
b= y-intercept
Regression Equation -
Y = dependent/outcome/criterion variable
X = predictor variable
B0 = regression constant (y-intercept)
B1 = regression coefficient (slope of line)
oChange in y associated with 1-unit increase in x
Simple Linear Regression
Exam 2 score = 40 + .59x (Exam 1 score)
oX = 62
Y = 76.58
oX = 73
Y = 83.07
oX = 88
Y = 91.92
oX = 96
Y = 96.64
Standardized Regression Coefficients (B)
Convey strength and direction of relation between outcome and a particular predictor
PRO - they are standardized. Can directly compare Bs in multiple regression to see which predictor is
most influential
Multiple Regression
Inclusion of >1 predictor in regression equation
oTypically allows us to improve the accuracy of our prediction
oCan accommodate both categorical and continuous predictors
Ex: college success predicted by high school GPA, financial aid, and familial support
Regression Table answers:
High school GPA, familial support, not financial aid
Positive relationship
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Regression is used to study correlations further and develop hypotheses about these relationships. Regression analysis - ability to predict scores on one variable from 1 or more other variables. Regression equation - helps to predict scores on one variable on the basis of the scores of another variable. We find the line that best fits the data which will help us create an equation to predict the variables. Run regression predicting one variable from > 1 others. B1 = regression coefficient (slope of line) Change in y associated with 1-unit increase in x o. Exam 2 score = 40 + . 59x (exam 1 score) o o o o. Convey strength and direction of relation between outcome and a particular predictor. Can directly compare bs in multiple regression to see which predictor is most influential. Inclusion of >1 predictor in regression equation o o. Typically allows us to improve the accuracy of our prediction.
