Unit 6, Inference for Categorical Data Proportions
Confidence Interval and Level
Confidence Interval
Interpretation: We are---% confident that interval from_to __ captures the true
statistics±(critical value)(standard error of statistics)
Confidence Level
Interpretation: If we take many samples of the same size from this population, about ---% of them will result in an interval that captures the true parameter
Constructing Confidence Interval
C.I. for Population Proportion
Formula for one sample-z interval for proportions
Conditions(Plan)
- Random: SRS of ....
- Independence: Assume at least (n*10)
- Normality: np̂≥10 and n(1-p̂)≥10
Conclude
Use the Confidence Interval Interpretation
Formula for finding the sample size from given Marginal Error
Types of Error
Type I Error
Type II Error
P(Type I error)=α(significance Level)
P(Type II error)=1-power
Type I error is made when the null hypothesis is actually true but the alternative hypothesis is chosen
- Reject Ho when it is true
Type II error is made when failing to reject the null hypothesis when it is false
- Fail to reject Ho when it is false
Significance Test
Hypthoesis
Null Hypothesis(Ho)
Alternative Hypothesis(Ha)
"No difference" statement.
Null means NOT
Ho: P=
Claim we are find evidence for
Ha: P >, <, ≠
P<α
- Reject the null, there is convincing evidence for Ha(alternative)
P>α
- Fail to reject the null, there is no convincing evidence for Ha(alternative)
P-Value Interpretation
- If the [null hypothesis is true], the probability of getting [sample size] is [p-value]
Constructing significance test for population proportions
- State:
-Null Hypothesis
-Alternative Hypothesis
-Define Parameter
-Significance Level - Plan:
-Random
-Independence
-Normality:np̂≥10 and n(1-p̂)≥10 - Do:
-1-sample z-test
-Normalcdf() - Conclude:
-Reject/Fail to reject the null(Ho), Convincing evidence/No Convincing evidence for alternative(Ha)
1-sample z-test formula
Constructing significance test for difference in two population proportions
- State:
-Ho: p1-p2=0
-Ha: p1-p2 >, < 0 - Plan
-Random
-Independence
-Normality:n1pc≥10 and n1(1-pc)≥10, n2pc≥10 and n2(1-pc)≥10 - Do
-2 Prop Z-Test - Conclude
-Reject/Fail to reject the null(Ho), Convincing evidence/No Convincing evidence for alternative(Ha)
CI for the Diff. of Two Proportions
Conditions(Plan)
- Random: SRS of ....
- Independence: Assume at least (n*10)
- Normality: n1p̂1≥10 and n1(1-p̂1)≥10, n2p̂2≥10 and n2(1-p̂2)≥10
Do
- Use 2-Prop Z-Interval