1 STAT 500 – Statistics for Managers STAT 500 Statistics for Managers
2 STAT 500 – Statistics for Managers Estimation Confidence Intervals
3 STAT 500 – Statistics for Managers Learning Objectives Define a probability interval. Calculate and interpret a probability interval for a specific random variable. Define a confidence interval. Calculate and interpret a confidence interval for the population mean.
4 STAT 500 – Statistics for Managers Learning Objectives (cont.) Calculate and interpret one-sided confidence intervals, and Apply confidence intervals to understand the mean of 0,1 population
5 STAT 500 – Statistics for Managers Factors Affecting Interval Width Data dispersion –Measured by Sample size X = / n Level of confidence (1 - ) –Affects Z Intervals extend from X - Z X to X + Z X © T/Maker Co.
Confidence Interval Mean ( Unknown) Assumptions –Population standard deviation is unknown –Population must be normally distributed Use Student’s t distribution Confidence interval estimate
Student’s t Distribution 0 t (df = 5) Standard normal t (df = 13) Bell- shaped Symmetric ‘Fatter’ tails Note: As d.f. approach 120, Z and t become very similar
Student’s t Table Assume: n = 3 df= n - 1 = 2 =.10 /2 = t values / 2.05
Degrees of Freedom Number of observations that are free to vary after sample statistic has been calculated Example –Sum of 3 numbers is 6 X 1 = 1 (or any number) X 2 = 2 (or any number) X 3 = 3 (cannot vary) Sum = 6 degrees of freedom = n -1 = 3 -1 = 2
Estimation Example Mean ( Unknown) A random sample of n = 25 has X = 50 & S = 8. Set up a 95% confidence interval estimate for .
11 STAT 500 – Statistics for Managers You’re a time study analyst in manufacturing. You’ve recorded the following task times (min.): 3.6, 4.2, 4.0, 3.5, 3.8, 3.1. What is the 90% confidence interval estimate of the population mean task time?
12 STAT 500 – Statistics for Managers X = 3.7 S = n = 6, df = n - 1 = = 5 S / n = / 6 = t.05,5 = (2.015)(1.592) (2.015)(1.592) 6.908
Confidence Interval for the Mean The Middle of the C.I. is the Sample Mean The Width of the C.I. is Determined by: –The Confidence Desired Higher Confidence Wider Interval -z.025 z.025 -z.005 z.005
14 STAT 500 – Statistics for Managers Confidence Interval for the Mean The Middle of the C.I. is the Sample Mean The Width of the C.I. is Determined by: –The Confidence Desired Higher Confidence Wider Interval –The Variability of the Data: Standard Deviation Greater Variability Wider Interval
15 STAT 500 – Statistics for Managers Confidence Interval for the Mean The Middle of the C.I. is the Sample Mean The Width of the C.I. is Determined by: –The Confidence Desired Higher Confidence Wider Interval –The Variability of the Data: Standard Deviation Greater Variability Wider Interval –The Sample Size, n Larger Sample Size Narrower Interval
16 STAT 500 – Statistics for Managers Two Common Interpretations If many samples were taken and a 95% confidence interval computed from each, the population mean would be contained in about 95% of them. With 95% confidence, the population mean lies within the 95% confidence interval endpoints.
17 STAT 500 – Statistics for Managers Confidence Interval for the Proportion
18 STAT 500 – Statistics for Managers How to compute a confidence interval for a population proportion
19 STAT 500 – Statistics for Managers Pre-Election Poll in Anywhere, USA For prop 1565% Against prop 1535% What is the percentage of all voters who favor Prop 15 ? How much uncertainty is there in the estimated percentage ?
20 STAT 500 – Statistics for Managers Population Parameter p = ??? 1 Inductive Inference
21 STAT 500 – Statistics for Managers Population Parameter Sample Statistic p = ??? p s = Inductive Inference
Population Parameter Sample Statistical Analysis Statistic Inference p = ??? p s = Inductive Inference
Binomial Probabilities : Application to Opinion Polling Assumptions n independent repeatable trials one of two mutually exclusive outcomes p = Pr(success) remains constant (the same) on each trial (Population Size is VERY Large) k = # of successes in the n trials Chance Situation or “Trial” Repeated n Times
24 STAT 500 – Statistics for Managers Central Limit Theorem for Binomial Proportions If Independent Observations Sample Size is Sufficiently Large Then
Confidence Interval Derivation z-z 1- 0
1. Estimate p : Confidence Interval for a Proportion
1. Estimate p : 2. Estimate SE : Confidence Interval for a Proportion
1. Estimate p : 2. Estimate SE : 3. Obtain the z Value (Normal Table) e.g. 95% Confidence Interval, z = 1.96 Confidence Interval for a Proportion
1. Estimate p : 2. Estimate SE : 3. Obtain the z Value (Normal Table) e.g. 95% Confidence Interval, z = Calculate Confidence Interval for a Proportion
30 STAT 500 – Statistics for Managers Latest Poll Suppose n = 1,500 and p s =.55 95% Confidence Interval
31 STAT 500 – Statistics for Managers THANK YOU