week 9 pt.2 and week 10 (confidence intervals)

interpretation and significance

• reject the null or fail to reject (not accepting the null hypothesis)

---

• statistical significanceL there was enough statistical evidence to reject the null

  

interpretation example

summary: one sample z-test

intervel estimation

• not the same as hypothesis testing (inferences of sample to a population

• interval estimation: involves generating point and interval estimates

  • point estimates are the calculation of a SINGLE statistic

confidence intervals

• confidence represents your studies ability to correctly retain (fail to reject) the null when the null is true

  • success rate of the method

• CI of the mean is used to make an interval estimation of the population mean from the sample data

• confidence = 1- alpha =

• we want high confidence

  ---

  MEANINGS:

• 95% CI: We believe would contain the true population mean and 5 would not

---

to calculate CI: margin of error

• margin of error: the distance from the center (point estimate) of the interval

A. for e 2 sided test: z value equal to one (z1-a/2)

• for a 2 sided test at a=0.05 → ±1.96 for 95% CI

CONFIDENCE INTERVAL LENGTH of 2 sided interval: L = 2m

• 2 times the margin of error

• for a 2 sided test: (larger sample size is better)

  • larger SD → larger margin/length (larger CI since larger cl results in larger critical value of the statistic)

  • larger sample size →

confidence interval and interpretation

• interval: (lower limit, upper limit)

• CL: xbar sample mean -

---

• interpretation: