Test Flashcards
A 99% confidence interval is better than a 95% C.I.
False. The interval for a 99% confidence level is often so long as to be useless.
As sample size increases, sample standard deviation decreases.
True. Sample std dev, or standard error, contains n in the denominator.
Small p-value means the magnitude of the difference from the null hypothesis is large.
False. P-value does not speak to magnitude of statistical significance.
Higher R-squared in a regression line implies the line fits well.
True. A higher R-squared value means that more of the variance is explained by the regression line.
Central Limit Theorem
The sampling distribution of any statistic will be approximately normal if the sample size is large enough.
Multicollinearity
Explanatory variables are highly correlated, leading to a computational problem and interpretation difficulties.
Type I Error
Rejecting null hypothesis when it is true. More of these errors occur with a higher test level.
Bernoulli Trial
A trial with only two possible outcomes - success and failure. X~B(n,p) where X is number of successes in n trials, p = success probability.
P-Value
Probability of the observed event, or more extreme events, under the null hypothesis.
What are the four assumptions of a multiple linear regression?
Linearity, independence, normality, and common variance.
If these are not true, results will not be reliable (Multicollinearity, homoskedastic, etc).
k
dF Model
Number of explanatory variables
dF total
n - 1
n is the number of observations.
Be sure to add 1
MSE
SSE/(n - k - 1)
MSE is the estimator of variance (sigma squared)
R^2
R-squared
SSM/SST
% of variance explained by the regression line
F Model
MSM/MSE
c/d