This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Sections 7.47.5 7.4 Small Sample Intervals Based on a Normal Population Distribution • tdistribution • confidence intervals, upper and lower bounds for a small sample distribution • prediction intervals We know from CLT that 1. If n is large , then for any population s n x z ) ( μ = is approximately standard normal, and that 2. If population is normal and standard deviation σ is known then for any n the sampling distribution of x is normally distributed and σ μ n x z ) ( = has standard normal distribution. However, if in the formula above we replace σ by s and n is small, then the distribution of the above statistic can be far from 1 standard normal. We introduce a new probability distribution for a small sample. tdistribution Definition : Let x 1 , ... x n be a random sample from a normal distribution.. Then the standardized variable Has a so called tdistribution with n1 degrees of freedom Properties of tcurves: • Symmetric and centered at zero • Thicker tails if compared to zcurve • Different for different n (n1 called degrees of freedom) • When n → ∞, the tcurves approach zcurve 2 n s x t μ = Example 1: Find the t critical value for a) Central area = .95 for df = 10 b) Upper tail area = .01, df = 25 c) Lower tail area = .025, df = 5 Solution : (a) From Table IV ( row 10, central area = .95), the critical value is 2.228. (b) An upper tail area of .01 corresponds to a cumulative area of .99, so from Table IV ( row 25, cumulative area area = .99), the critical value is 2.485. (c) A lower tail area of .025 corresponds to a central area of .95, so from Table IV ( row 5, central area = .95), the critical value is 2.571. One Sample t Confidence Interval for μ Assumptions: • Random sample from normal population ) , ( σ μ N No assumption about n 3 The twosided CI for μ is of the form n s t x c ± , where c t is the central area critical value for t distribution with df = n1 to be found in Table IV. ) (1 100 α % Upper Confidence Bound: n s t x c 1 + < μ where 1 c t is tcritical value for cumulative (one sided) area ) (1 100 α % Lower Confidence Bound: n s t x c 1 μ , where ) 1 ( 1 1 α = c t c is tcritical value for cumulative (one sided) area. Exercise 39: An article contained the following observations on degree of polymerization for paper specimens for which viscosity times concentration fell in a certain middle range: 418 412 421 422 425 427 431 434 437 439 446 447 448 453 454 463 465 a) Construct a boxplot of the data and comment on any interesting features. 4 112 470 460 450 440 430 420 No rm a l qua ntile b) Is it plausible that the given sample observations were selected from a normal distribution?...
View
Full
Document
 Summer '08
 Palaniappan
 Normal Distribution, µ, Student's tdistribution

Click to edit the document details