7. P- Values And Confidence Intervals and Ch. 5, 6 of E-book Flashcards
What is the null hypothesis
H0
Chosen to assume no effect
Alternative hypothesis
H1 or Ha
That there is an effect
Two sided - the effect can either be beneficial or harmful
What are the two assumptions you need to make for a t test
- The variable to be tested is, at least approximately, normally distributed
- The samples are independent
Calculation for t statistic
(Mean - mu)/ standard error
- mu values varies depending on the null hypothesis, eg. If the null hypothesis is that a certain drug has no effect on blood pressure, then mu = 0
Is a confidence interval adequate for hypothesis testing?
Yes. But we usually test using a test statistic and a p-value
What does the p-value tell us about the data?
how likely it is that the null hypothesis is correct
The p-value gives us the probability that the value we got is true if the null hypothesis were correct
What does a smaller p-value indicate about the likelihood that the result occurred by chance alone
the smaller the p-value, the less likely the result is to have occurred by chance alone
If the degrees of freedom are greater, does the t-distribution resemble the normal distribution more or less closely
the greater the degrees of freedom, the more closely the t-distribution resembles the normal distribution
what does the p-value measure
a p-value is a measure of statistical significance
If a p-value is less than 0.05, is it statistically significant? does it support/ go against the null hypothesis
A p-value of less than 0.05 is considered to be a statistically significant result
Provides evidence against the null hypothesis
Do we need to consider the clinical significance of a study where the results are not statistically significant?
Yes
It is possible to get a non-significant statistical result when there is a real clinical difference.
