This preview shows pages 1–3. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Chapter 8 CONFIDENCE INTERVALS 8.1 Introduction: The onesample problem for means. The Z interval. 8.2 Confidence interval for a proportion. 8.3 Comparing means of two independent samples. 8.4 Comparing two proportions. For lecture 13: 8.5 The onesample problem for means with unknown variance and small samples. Students tdistribution, the Tinterval. 8.6 Two independent samples, small sample sizes, variances unknown. The twosample t statistic. 8.7 Paired samples. 8.8 Confidence intervals for variances. The chisquared distribution. 8.9 Comparing two variances. The Fdistribution. 1 8.5 THE ONESAMPLE PROBLEM FOR MEANS WITH UNKNOWN VARIANCE AND SMALL SAMPLES Ref: Devore 6e, Pages 299303. Consider the Zinterval of Sec. 8.1. We can substitute S for , when is unknown, but if n is small (say less than 30) we should compensate for this extra uncertainty by using slightly wider confidence intervals. This is done by using the t percentage points of Students t dis tribution instead of standard normal zpercentage points. The pdf of Students t is bellshaped like the normal, but somewhat flatter: f ( x ) = r +1 2 r 2 r 1 1 + x 2 r ( r +1) / 2 Having written down this pdf , we will never have to use it, since the cdf is tabulated (e.g. Devore 6e Table A8, pages 7467). Note that the distribution depends on a parameter r , called the degrees of freedom. Some facts about the t distribution: Mean 0 (its density is symmetric), Variance is r/ ( r 2) > 1, ( r must be greater than 2). Also f ( x ) 1 2 e 1 2 x 2 as r Percentage points of t : t ( ; r ) : (for more complete listings, see Devore 6e Table A8, pages 7467.) degrees of = . 1 .05 .025 .01 freedom, r 5 1.476 2.015 2.571 3.365 15 1.341 1.753 2.131 2.602 30 1.310 1.697 2.042 2.457 (normal) 1.282 1.645 1.960 2.326 2 The remarkable result of William Gossett (pseudonym Student) is that the pivotal quan...
View Full
Document
 Spring '05
 STAFF

Click to edit the document details