Lecture #18 - Chance 2 Flashcards
What are p-values?
- Probability of getting study estimate………
- If probability is really low then……..
- Use logic of what?
The higher the value, the higher the probability that event you’re observing can be explained by ______ alone
-
Steps of hypothesis testing:
- The null hypothesis (Ho)
- What is it? So parameter = ?
- What are the ratio measures and difference measures in this case? - The alternative hypothesis (Ha)
- What is it? So parameter = ?
- What are the ratio and difference measures?
Go read the example
-
What is type 1 error rate and what is it usually set to?
It’s when we find association due to sampling error when there isn’t one
Set it to 0.05 so we’re okay with 5% of studies (or 1 in 20 studies) telling us there is an association when there really isn’t one
So what is the threshold for p-values? When can we accept or reject the two hypothesises and say the association is stat sig or not
-
What are the three statements you need to write in your interpretation of a RR = 2.1 and a p of 0.01 and p of 0.15
-
Fundamental concept #4:
Probability of getting study estimate (or an estimate further from the null when there’s really no association because of sampling error (chance).
-
- What is type-2 error?
- P value should have been what but was what?
- Why would you typically get a type-2 error rate?
- What do staticians calculate to find out how many participants are needed to minimise…….
-
What is the relationship between p-values and CI?
Explain
-
What’re the three main limitations of p-values?
-
Arbitrary threshold
- Threshold is _____ and ______
- first point about p-values
- always useful to report what rather than just “stat sig” or “not stat sig”
- At 5% threshold, still what? (what kind of error?)
-
Only about Ho
- Just give _____ for what?
- Doesn’t say anything about what? You can’t tell how _____ you estimate the parameter
- Best presented with what?
-
