start430 exam2

help

regression residuals

the estimates of the true error in a regression [observed error]

significance level a

the probability of a type 1 error

type 1 error

false positive conclusion, incorrectly reject null when its true

type 2 error

false negative conclusion, fail to reject null when its false

sampling distribution

probability distribution of random sample-based statistic

confidence interval

a range of values used to estimate a population parameter

distribution

the variable values and how often they occur

law of large numbers

explains why as the number of observations increases, the sample mean gets closer to the mean u

level of confidence

the degree of certainty that a calculated interval contains the true population parameter

alpha

probability of a type 1 error based on the sample

p-value

probability of another sample producing a mean as far or further in the tail end

statistic

descriptive measure of a sample

test statistic

how many standard deviations our statistic from the sample is in the null hypothesis distribution

standard error

poopulation SD / sqrt of n

PROC ANOVA

to see if 3 or more distributions have the same mean, independent distributions, normal distribution, and have the same variance

PROC CORR

gives statistics and a table with pearson correlation coefficients

PROC REG

gives how strong the linear relationship is and if the conditions are met

if you saw scatterplot was linear so you run PROC REG, what do you look at to see if the linear regression model is ‘good’

r-square, p-value from variance, parameter estimates, residuals

adjacent residuals should be correlated with each other

false

CLT

regardless of the shape of population, the sampling distribution becomes approximately normal for n

linear correlation coefficient

measures the strength of the linear correlation between x and y values, also known as pearson product correlation coefficient