O - Week Flashcards
What is research methodology?
Collection of approaches, paradigms, procedures that give us confidence in our data
What is statistical analysis?
Algorithms to process data to arrive at conclusion
What is the line of best fit?
Gives us predicted Y for every possible change in X
What is R2?
Measure of total association between X and Y
What happens when more than one X?
The size of b and sr2 indicates the relative strength of association
What do we want to know in regression?
The explained variance of the outcome (y)
Why is each data point less or more than the mean on Y
Falsification, Null hypothesis significance testing
Is the underlying population in which the null is true have given rise to the sample statistic
If null is true b-weight should be 0 (Close to 0)
Sampling distribution
Statistic provide expected value and statistics in population
Z-test
Apply normal curve theory
Divide statistic by its own SE
Used to find probability of obtaining a b-weight that is extreme from the sampling distribution under the null
What is good regression?
B-weight reflect the relationship between X and Y for all data points
What is bad regression?
B-weight strongly affected by the inclusion/exclusion of a small number of data points (influential data points)
Distance
Points far from the regression line
Above 3.3 require checking
Studentised residuals
Leverage
Identify the multivariate outliers - outliers in the space of the criterion (Y)
Checked using mahalanobis distance (high and significant values problematic)
Influence
Distance and Leverage
Cooks distance
Regression Assumptions
Assumption Normality
Independence of errors
Homoscedasticity
Assumption linearity