Chapter 7 Notes
©2008 by S. Gramlich
updated 5/15/2014
Confidence Intervals
! = Important Note
! These Notes are not meant to replace
7-2
Steps to estimate Population Proportion of Successes (P) Confidence Interval (CI):
1) Check assumptions: Binomial (2 outcomes: "Success" and "Failure"), Normality [x & (n-x) >=5], SRS
2) given or find sample proportion of successes: phat = x/n (that's "p-hat" not "fat")
calculate sample proportion of failures: qhat = 1 - phat
3) given Confidence Level (CL), find critical value zα/2 from Table A-2 NEGATIVE z Scores,
in the body of the table lookup closest value to (1-CL)/2
!
1-CL = α (alpha) which represents the area of the critical region
! critical value is the cutoff value (along horizontal axis) for the critical region
4) calculate margin of error E = zα/2 * √(phat*qhat/n)
5) calculate the CI: phat - E < P < phat + E
Estimating Sample Size (N)
if phat known: n = [zα/2]2 * phat*qhat/ E2
if phat
unknown, let phat =50%: n
= [zα/2]2 * .5*.5/ E2 = [zα/2]2 * .25/
E2
*always round N up (not roundoff)
7-3
Student t (Gosset) Distribution Properties:
1) Bell-Shaped & Symmetric (Frequency Polygon with zero skewness)
2) Total Area Under the curve is 100% or 1 (so each half is 50% or .5)
3) Continuous
4) Asymptotic to the horizontal axis (runs infinitesimally close to but never touches)
5)
Different for different sample sizes (approaches
6) mean (μ) = 0, standard deviation (σ) > 1
Determining whether to use t, z, or neither:
1) sigma (population standard deviation) known? yes-> z, no-> t
2) normal?
3) n > 30?
if answer is no to 2) & 3) then neither.
Steps to estimate Population Mean (μ) Confidence Interval (CI):
1) Check assumptions: SRS, approximately
2) given or find mean and standard deviation
3) given Confidence Level (CL), find critical value tα/2 in body of Table A-3 with df = n-1,
! top row gives alpha α (alpha) = 1 - CL, always use 2 tail for CI"s
4) calculate margin of error E = tα/2 * s/√n
5)
calculate the CI: xbar - E < μ < xbar
+ E
TECHNOLOGY
using StatCrunch:
! Don't need to enter any info in StatCrunch spreadsheet for critical value procedures
to find critical zα/2: Stat - Calculators -
to find critical tα/2: Stat - Calculators - T - (df=n-
! 3rd box will give critical value in both above procedures
! Both CI Procedures below will only give the CI and not the Critical Value or Margin of Error
for P CI:
! Don't need to enter any info in StatCrunch spreadsheet for this procedure
Stat - Proportions - One sample - with summary - (enter # successes, # observations) - Next -
-Confidence Interval - (enter CL for Level) - Calculate
for μ CI:
! Don't need to enter any info in StatCrunch spreadsheet for this procedure
Stat - T statistics - One sample - with summary
-(select column) - Next - Confidence Interval - (enter CL for Level) - Calculate
! this procedure is for using the original data set (not the summary),
entered into a column of the StatCrunch spreadsheet first
Stat - T statistics - One sample - with data
-(select column) - Next - Confidence Interval - (enter CL for Level) - Calculate
EXCEL commands:
Find Excel
Command
critical zα/2 =NormSinv(α/2) {! will only give the negative critical z value}
critical tα/2 =tinv(α,df) {! will only give the positive critical t value, enter α not α/2}
The "Using Technology" section in the text gives Excel procedures using DDXL.
The DDXL add-in can be found in CourseCompass under Chapter Contents (in left margin) - Tools for Success -DDXL.
! DDXL must be downloaded from CourseCompass and then added into the Excel software. If you are using a college computer, the add-in will be removed when the computer is reset.
! In Excel 2007, you have to highlight the data before launching into a DDXL procedure.
to find Margin of Error E for μ CI:
Excel 2003: Tools - Data Analysis - Descriptive Statistics - (check Confidence Level for Mean and enter CL) - OK
Excel 2007: Data - Data Analysis - Descriptive Statistics - (check Confidence Level for Mean and enter CL) - OK