tailieunhanh - Ebook Statistics for business - Decision making and analysis (2nd edition): Part 2

(BQ) Part 2 book "Statistics for business - Decision making and analysis" has contents: Inference for counts, linear patterns, curved patterns, the simple regression model, regression diagnostics, multiple regression, building regression models, categorical explanatory variables, alternative approaches to inference,.and other contents. | Inference for Counts From Chapter 18 of Statistics for Business: Decision Making and Analysis, Second Edition. Robert Stine, Dean Foster. Copyright © 2014 by Pearson Education, Inc. All rights reserved. 485 Inference for Counts Online marketing allOws a retailer tO custOmize sales pitches. In place of a generic store- 1 2 3 4 front, Web sites can tailor the merchandise they offer to specific customers. The better the retailer knows its customers, the more it can specialize its offerings. If you regularly buy music from a Web site, the next time you visit, you’ll see songs related to your prior purchases. And if you’ve been spending a lot, the retailer will offer you high-end products, hoping to get your attention before you surf to a different site. Customizing the shopping experience isn’t easy. If you’ve ever bought gifts for a younger relative, you know what can happen. All of a sudden, the chi-squared tests Web site thinks you are that person and offers you test Of independence toys for kids rather than products you want. To avoid general Versus specific such mistakes, Web sites try to learn more about hypOtheses customers. Some guess incomes to tailor the shopping tests Of gOOdness Of fit experience. Does income matter? Do shoppers with summary higher incomes buy, for instance, different music? When a shopper visits the electronics department at Amazon, does income anticipate whether they are looking for, say, a new camera or a new phone? Data can answer these questions. If data show that income and shopping choices are independent, then Amazon can skip using income to guess shoppers’ preferences. All we need is a test for independence. The chi-squared sTaTisTic TesTs hypoTheses abouT a conTingency Table and The disTribuTion of a caTegorical variable. As with other tests, a null hypothesis states a claim about a population. The chi-squared statistic measures how far our data deviate from this claim. This chapter examines chi-squared tests for .