Unit 6: Inferences for Categorical Data: Proportions Flashcards
Interpret a Confidence Level.
After many, many samples of same size and from the same population __% of the confidence intervals will capture the true proportion of ___.
Interpret a One-Sided Confidence Interval for a Population Proportion.
We are __% confident that the interval (__,__) captures the true mean of ____.
What affects the width of a confidence interval?
The size of sample (n) and the confidence level, when increased, the interval is wider, and when decreased, the interval is narrower.
Interpret the p-value.
Assuming the H0 is true, there is a ___ (p-value) probability of obtaining a test statistic greater than (less than) or equal to the observed test statisitic.
Interpret a Hypothesis Test for a Single Population.
Because the p-value is less than (or greater than) alpha = ___ (alpha value), we fail to reject (or reject) the H0. We do not (or do) have convincing evidence that Ha is true.
What is Type I Error?
This is when we reject the null hypothesis and it’s actually true.
What is Type II Error?
This is when we fail to reject the null hypothesis, but the alternative hypothesis is true.
What increases power?
Increasing sample size
Increasing alpha value (but also increases the probability of Type I Error)
Increasing distance to H0 (Difference from H0 and Ha gets greater)
Interpret the power of a test.
If Ha is true (context and specific value), there is a ____ (power) probability that we find convincing evidence to reject the H0 (give context and specific value).
2-Sample Z-Int: Interpret a Difference in Two Proportions
We are __% confident that the interval (__,__) captures the true difference between the proportions of (First proportion and Second proportion).
2-Sample Z-Test: Interpret a Difference in Two Proportions
Because the p-value ___ is less than/greater than alpha = ____, we REJECT (FAIL TO REJECT) H0, and ACCEPT (DO NOT ACCEPT) Ha.
What condition check for Independence?
10% condition
What condition checks for Normality?
Large Counts
What condition checks for Random?
None, look at the question and interpret.
What does the “Choose” consist of in the 4C’s?
Choose the procedure (1-Prop Z-Int, 2-Prop Z-Int, 1-Prop Z-Test, 2-Prop Z-Test)
DEFINE p (true proportion of __)
Write Hypotheses (H0: p = __) (Ha: p </>/NOT EQUAL to __)
State alpha-value
Define p-hat (the value AND representation)
EVIDENCE for Ha: (If in the problem, it gives p as a different value than before, write it down)
What does the “Check” consist of in the 4C’s?
Conditions:
Independence: 10%
Normality: Large Counts
Random: Check Prompt
Name of Procedure: List it
