Statistical Inference Flashcards
What is inference?
conclusion drawn from data
What is standard error?
how closely does sample estimate the population mean; =standard deviation/ population size; smaller number, more precision
What is a confidence interval?
- estimation of range of values that are likely to include the true population value
- estimate +/- critical value x standard error; critical value depends on sample size and confidence (95% CI is +/- 2 SE)
What is a hypothesis test?
estimation of probability of obtaining observed result if there is no true difference between the means of the control vs. test group
When do you use T distribution test statistic over a Z normal distribution test statistic?
when sample size is small (allows for wider confidence interval)
= difference of population means/ standard error of difference
What is a type I error?
reject the null hypothesis when the null hypothesis is true
What is a type II error?
do not reject the null hypothesis when the null hypothesis is false
What is alpha?
Probability of Type I error, usually set =0.05
What is the p value?
probability of obtaining the observed test statistic IF the null hypothesis is true
How do you decide if evidence is sufficient to reject the null hypothesis?
If p < alpha
- observed value of test statistic is unlikely if null hypothesis is true
- null hypothesis is likely false, can be rejected
- difference is not due to chance, therefore difference is statistically significant
What is beta?
Probability of Type II error
What is Power? What affects it?
= 1-beta depends on: -type I error rate (alpha) -effect size (difference in means or proportions) -variation of outcome measure -sample size
What is the null value?
represents no difference between groups:
ratio null value =1
difference null value =0
What is the relationship between hypothesis tests and confidence intervals?
if 95% CI contains “null value,” p>0.05
if 95% CI does not contain “null value,” p<0.05
