Module 5 Confidence Interval Estimation Learning Objectives • Estimation process o Point estimates o Interval estimates • Confidence interval estimation for the o Population Mean, μ • when Population Standard Deviation σ is Known • when Population Standard Deviation σ is Unknown 3/08/2020 RMIT University Vietnam 2 Learning Objectives 2 • Determining Sample Size, n o when estimating the Population Mean, μ 3/08/2020 RMIT University Vietnam 3 Point and Interval Estimates • A point estimate is the value of a single sample statistic. • A confidence interval provides a range of values constructed around the point estimate. Lower Confidence Limit Point Estimate Upper Confidence Limit Width of confidence interval 3/08/2020 RMIT University Vietnam 4 Point Estimates We can estimate a Population Parameter… Mean 3/08/2020 μ RMIT University Vietnam with a Sample Statistic (Point Estimate). X 5 Confidence Interval Estimate • An interval gives a range of values: o Takes into consideration variation in sample statistics from sample to sample. o Based on observations from 1 sample. o Gives information about closeness to unknown population parameters. o Stated in terms of level of confidence. o Can never be 100% confident. 3/08/2020 RMIT University Vietnam 6 Estimation Process Random Sample Population Mean X = 50 (mean, μ, is unknown) I am 95% confident that μ is between 40 & 60. Sample 3/08/2020 RMIT University Vietnam 7 Confidence Interval • The general formula for all confidence intervals is: Point Estimate ± (Critical Value)*(Standard Error) Represents confidence for which the interval will contain the unknown population parameter. 3/08/2020 RMIT University Vietnam 8 Confidence Level (1-) • Common confidence levels = 90%, 95% or 99%: o Also written (1 - ) = 0.90, 0.95 or 0.99 • A relative frequency interpretation: o In the long run, 90%, 95% or 99% of all the confidence intervals that can be constructed (in repeated samples) will contain the unknown true parameter. 3/08/2020 RMIT University Vietnam 9 Confidence Level (1-) 2 • A specific interval will either contain or will not contain the true parameter. o No probability involved in a specific interval. 3/08/2020 RMIT University Vietnam 10 Confidence Intervals Confidence Intervals Population Population Mean Proportion (not part of this course) σ Known 3/08/2020 σ Unknown RMIT University Vietnam 11 Confidence Interval for μ (σ Known) • Assumptions: o Population standard deviation σ is known o Population is normally distributed o If population is not normal, check n and hence CLT. • Confidence interval estimate σ XZ n 3/08/2020 RMIT University Vietnam 12 Confidence Interval for μ (σ Known)2 Where X is the point estimate Z is the normal distribution critical σ/ n value for a probability of /2 in each tail is the standard error σ XZ n 3/08/2020 RMIT University Vietnam 13 Finding the Critical Z Value • Consider a 95% confidence interval Z 1.96 1 0.95 α 0.025 2 α 0.025 2 Z units: Z= -1.96 X units: 3/08/2020 Lower Confidence Limit 0 Point Estimate RMIT University Vietnam Z= 1.96 Upper Confidence Limit 14 Common Levels of Confidence 90%, 95% and 99% Confidence Level 80% 90% 95% 98% 99% 99.8% 99.9% 3/08/2020 Confidence Coefficient 1 0.80 0.90 0.95 0.98 0.99 0.998 0.999 RMIT University Vietnam Z value 1.28 1.645 1.96 2.33 2.575 3.08 3.27 15 Intervals and Level of Confidence Sampling Distribution of the Mean /2 Intervals extend from σ XZ n to 3/08/2020 σ XZ n 1 /2 x μx μ X1 X X 2 3 X 4 Xn Confidence Intervals RMIT University Vietnam (1-)x100% of intervals constructed contain μ; ()x100% do not 16 Confidence Interval Example • A sample of 11 pizza shops from a large normal population has a mean pizza cooking time of 2.20 minutes. We know from past testing that the population standard deviation is 0.35 minutes. • Determine a 95% confidence interval for the true mean pizza cooking time of the population. 3/08/2020 RMIT University Vietnam 17 Confidence Interval Example 2 • We are 95% confident that the true mean cooking time is between 1.9932 and 2.4068 minutes. • Although the true mean may or may not be in this interval, 95% of XZ σ n intervals formed in this manner (in repeated samples) will contain 2.20 1.96 (0.35/ 11 ) the true mean. 2.20 0.2068 1.9932 2.4068 3/08/2020 RMIT University Vietnam 18 Confidence Intervals Confidence Intervals Population Population Mean Proportion (not part of the course) σ Known 3/08/2020 σ Unknown RMIT University Vietnam 19 Confidence Interval for μ (σ Unknown) • If the population standard deviation σ is unknown, we can substitute the sample standard deviation, S. • This introduces extra uncertainty, since S is variable from sample to sample. 3/08/2020 RMIT University Vietnam 20 Confidence Interval for μ (σ Unknown) 2 • So we use the Student t distribution instead of the normal distribution: o The t value depends on degrees of freedom denoted by sample size minus 1 i.e. (d.f = n - 1). o d.f are number of observations that are free to vary after sample mean has been calculated. 3/08/2020 RMIT University Vietnam 21 Confidence Interval for μ (σ Unknown) 3 Confidence Interval Estimate X t n-1 S n Where t is the critical value of the t distribution with n -1 degrees of freedom and an area of α/2 in each tail. 3/08/2020 RMIT University Vietnam 22 Student’s t Distribution • Note: t Z as n increases Standard Normal (t with d.f = ∞) t distributions are bellshaped and symmetric, but have ‘fatter’ tails than the normal t (d.f = 25) t (d.f = 5) 0 3/08/2020 RMIT University Vietnam t 23 Student’s Table Upper Tail Area d.F .25 .10 .05 1 1.000 3.078 6.314 2 3 0.817 1.886 2.920 0.765 1.638 2.353 The body of the table contains t values, not probabilities. 3/08/2020 RMIT University Vietnam Let: n = 3 d.f = n - 1 = 2 = 0.10 /2 = 0.05 /2 = 0.05 0 2.920 t 24 Confidence Interval Example • A random sample of 25 businesswomen has a mean age of 50 and standard deviation of 8. Form a 95% confidence interval for μ. 3/08/2020 RMIT University Vietnam 25 Confidence Interval Example 2 • d.F = 25-1 = 24 t/2 , n1 t 0.025,24 2.0639 X t /2, n-1 S 8 50 (2.0639) n 25 46.698 ≤ μ ≤ 53.302 We are 95% confident that the true mean age of businesswoman is between 46.698 and 53.305 years. 3/08/2020 RMIT University Vietnam 26 Determining Sample Size • If the Sample Size is too big it requires too much resources ($$$) • If the Sample Size is too small it won’t do the job What would be the minimum required sample size? 3/08/2020 RMIT University Vietnam 27 Sampling Error • The required sample size can be found to reach a desired margin of error (e) with a specified level of confidence (1 ). • The margin of error is also called a sampling error. 3/08/2020 RMIT University Vietnam 28 Sampling Error 2 o The amount of imprecision in the estimate of the population parameter. o The amount added and subtracted to the point estimate to form the confidence interval. 3/08/2020 RMIT University Vietnam 29 Determining Sample Size for the Mean • Confidence interval formula for population mean σ XZ n Sampling error (margin of error) σ eZ n denoted e 3/08/2020 RMIT University Vietnam 30 Determining Sample Size for the Mean 2 σ eZ n Z σ n 2 e 2 Now solve for n to get 3/08/2020 RMIT University Vietnam 2 31 Determining Sample Size for the Mean 3 • To determine the required sample size for the mean, you must know: o The desired level of confidence (1 - ), which determines the critical Z value. o The acceptable sampling error, e. o The standard deviation, σ. 3/08/2020 RMIT University Vietnam 32 Determining Sample Size for the Mean 4 If = 45, what sample size is needed to estimate the mean within ± 5 with 90% confidence? Z σ (1.645) (45) n 219.19 2 2 e 5 2 2 2 2 So the required sample size is n = 220 (Always round up) 3/08/2020 RMIT University Vietnam 33 If σ is Unknown • If unknown, σ can be estimated: o From past data using that data’s standard deviation. o If population is normal, range is approximately 6σ so we can estimate σ by dividing the range by 6. o Conduct a pilot study and estimate σ with the sample standard deviation, s. 3/08/2020 RMIT University Vietnam 34 Ethical Issues • A confidence interval estimate (reflecting sampling error) should always be included when reporting a point estimate. • The level of confidence should always be reported. • The sample size should be reported. • An interpretation of the confidence interval estimate should also be provided. 3/08/2020 RMIT University Vietnam 35 Topic Summary • Discussed point estimates. • Developed confidence interval estimates: o for the mean (σ known) o for the mean (σ unknown) • Determined required sample size for mean 3/08/2020 RMIT University Vietnam 36