MKT 343 EXAM 3 STUDY GUIDE
Types of Regression - Answers - 1. Linear/Curvilinear Regression
2. Bivariate/Multiple
Relationships between variables - Answers - - Presence
- Direction
- Strength of association (no, weak, moderate, strong relationship)
- Type (linear/curvilinear relationship)
Linear Relationship - Answers - Strength and nature of the relationship remains the
same over the range of both variables
Curvilinear Relationship - Answers - Strength and/or direction of their relationship
changes over the range of both variables
Covariation - Answers - The amount of change in one variable that is constantly related
to the change in another variable of interest
Scatter Diagram - Answers - A graphic plot of the relative position of 2 variables using a
horizontal and a vertical axis to represent the values of the respective variables (a way
of describing the covariation btwn 2 variables)
Pearson Correlation Coefficient - Answers - Statistical measure of the strength of a
linear relationship btwn 2 metric variables; varies btwn -1.00 and 1.00 (0 represents
absolutely no association)
Assumptions for Calculating Pearson's Correlation Coefficient - Answers - - the 2
variables have been measured using interval- or ratio-scaled measures
- nature of relationship is linear (straight line describes the relationship)
- variables to be analyzed need to be from a normally distributed population
Weak Correlation - Answers - There is no significant relationship OR the relationship is
not linear (further tests required; curvilinear regression)
coefficient of determination (r squared) - Answers - - A number measuring the
proportion of variation in one variable accounted for by another
- percentage varies from 0.0 to 1.00
- the larger the size of the coefficient of determination, the stronger the linear
relationship btwn the 2 variables being examined
Substantive Significance - Answers - whether an observed effect is large enough to be
meaningful. It was developed because statistical SIGNIFICANCE TESTS can find that
very small effects are significant, even they are too small to matter.
, 4 Steps to evaluate Multiple Regression - Answers - 1. Assess overall model
significance
2. Evaluate model R-square
3. Examine individual regression coefficients (betas) and their t-statistics for significance
4. Compare relative influence of IVs on the DV based on relative size of betas
Multiple Regression Assumptions - Answers - - linear relationship(s)
- normal distribution (shape of distribution of variable equal both above and below
mean)
Regression analysis - Answers - a set of statistical processes for estimating the
relationships between a dependent variable and one or more independent variables
Multiple Regression Analysis - Answers - Analyzes the linear relationship between DV
and multiples IVs by estimating coefficients for the equation for a straight line
Statistical Significance of each coefficient - Answers - Each regression coefficient is
divided by its standard error to produce a t statistic; compared against critical value to
determine whether the null hypothesis can be rejected
Model F statistic - Answers - Compares the amount of variation in the dependent
measure "explained" or associated with the IVs to the "unexplained" or error variance;
larger F statistic indicates that the regression model has more explained variance than
error variance
Means Comparison - Answers - t-tests/ANOVA
T-tests - Answers - - hypothesis test that utilizes the t distribution
- useful when sample size smaller than 30 and std dev is unknown
ANOVA (analysis of variance) - Answers - - statistical technique that determines
whether 3 or more means are statistically different from one another
- Null hypothesis always states that there is no difference between the dependent
variable group
- F-test
- Follow-up tests performed after to determine diff btwn means
Follow-up tests - Answers - test that flags the means that are statistically different from
each other
F-test - Answers - The test used to statistically evaluate the differences between the
group means in ANOVA; (variance btwn groups/variance within groups)
Multicollinearity - Answers - - A situation in which several IVs are highly correlated with
each other
Types of Regression - Answers - 1. Linear/Curvilinear Regression
2. Bivariate/Multiple
Relationships between variables - Answers - - Presence
- Direction
- Strength of association (no, weak, moderate, strong relationship)
- Type (linear/curvilinear relationship)
Linear Relationship - Answers - Strength and nature of the relationship remains the
same over the range of both variables
Curvilinear Relationship - Answers - Strength and/or direction of their relationship
changes over the range of both variables
Covariation - Answers - The amount of change in one variable that is constantly related
to the change in another variable of interest
Scatter Diagram - Answers - A graphic plot of the relative position of 2 variables using a
horizontal and a vertical axis to represent the values of the respective variables (a way
of describing the covariation btwn 2 variables)
Pearson Correlation Coefficient - Answers - Statistical measure of the strength of a
linear relationship btwn 2 metric variables; varies btwn -1.00 and 1.00 (0 represents
absolutely no association)
Assumptions for Calculating Pearson's Correlation Coefficient - Answers - - the 2
variables have been measured using interval- or ratio-scaled measures
- nature of relationship is linear (straight line describes the relationship)
- variables to be analyzed need to be from a normally distributed population
Weak Correlation - Answers - There is no significant relationship OR the relationship is
not linear (further tests required; curvilinear regression)
coefficient of determination (r squared) - Answers - - A number measuring the
proportion of variation in one variable accounted for by another
- percentage varies from 0.0 to 1.00
- the larger the size of the coefficient of determination, the stronger the linear
relationship btwn the 2 variables being examined
Substantive Significance - Answers - whether an observed effect is large enough to be
meaningful. It was developed because statistical SIGNIFICANCE TESTS can find that
very small effects are significant, even they are too small to matter.
, 4 Steps to evaluate Multiple Regression - Answers - 1. Assess overall model
significance
2. Evaluate model R-square
3. Examine individual regression coefficients (betas) and their t-statistics for significance
4. Compare relative influence of IVs on the DV based on relative size of betas
Multiple Regression Assumptions - Answers - - linear relationship(s)
- normal distribution (shape of distribution of variable equal both above and below
mean)
Regression analysis - Answers - a set of statistical processes for estimating the
relationships between a dependent variable and one or more independent variables
Multiple Regression Analysis - Answers - Analyzes the linear relationship between DV
and multiples IVs by estimating coefficients for the equation for a straight line
Statistical Significance of each coefficient - Answers - Each regression coefficient is
divided by its standard error to produce a t statistic; compared against critical value to
determine whether the null hypothesis can be rejected
Model F statistic - Answers - Compares the amount of variation in the dependent
measure "explained" or associated with the IVs to the "unexplained" or error variance;
larger F statistic indicates that the regression model has more explained variance than
error variance
Means Comparison - Answers - t-tests/ANOVA
T-tests - Answers - - hypothesis test that utilizes the t distribution
- useful when sample size smaller than 30 and std dev is unknown
ANOVA (analysis of variance) - Answers - - statistical technique that determines
whether 3 or more means are statistically different from one another
- Null hypothesis always states that there is no difference between the dependent
variable group
- F-test
- Follow-up tests performed after to determine diff btwn means
Follow-up tests - Answers - test that flags the means that are statistically different from
each other
F-test - Answers - The test used to statistically evaluate the differences between the
group means in ANOVA; (variance btwn groups/variance within groups)
Multicollinearity - Answers - - A situation in which several IVs are highly correlated with
each other
