ppt课件-how can cost effectiveness analysis be made more relevantto（成本效益分析可以更加relevantto如何）.pptVIP

下载本文档

7
0
约8.13千字
约 37页
2017-01-25 发布于四川
举报
版权申诉

ppt课件-how can cost effectiveness analysis be made more relevantto（成本效益分析可以更加relevantto如何）.ppt

1、原创力文档（book118）网站文档一经付费（服务费），不意味着购买了该文档的版权，仅供个人/单位学习、研究之用，不得用于商业用途，未经授权，严禁复制、发行、汇编、翻译或者网络传播等，侵权必究。。
2、本站所有内容均由合作方或网友上传，本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺！文档内容仅供研究参考，付费前请自行鉴别。如您付费，意味着您自己接受本站规则且自行承担风险，本站不退款、不进行额外附加服务；查看《如何避免下载的几个坑》。如果您已付费下载过本站文档，您可以点击这里二次下载。
3、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等，请点击“版权申诉”（推荐），也可以打举报电话：400-050-0827(电话支持时间：9:00-18:30)。
4、该文档为VIP文档，如果想要下载，成为VIP会员后，下载免费。
5、成为VIP后，下载本文档将扣除1次下载权益。下载后，不支持退款、换文档。如有疑问请联系我们。
6、成为VIP后，您将拥有八大权益，权益包括：VIP文档下载权益、阅读免打扰、文档格式转换、高级专利检索、专属身份标志、高级客服、多端互通、版权登记。
7、VIP文档为合作方或网友上传，每下载1次，网站将根据用户上传文档的质量评分、类型等，对文档贡献者给予高额补贴、流量扶持。如果你也想贡献VIP文档。上传文档

* * * * * * * * * * * SAS Collin option SAS Model Y= var1 … varN/collin; Continue newborn example Birth weight and gestational age very correlated. R=0.56 Simple model, only BW, GA, Black Interpreting Collin output Condition index 10 indicates a collinearity problem Condition index 100 indicates an extreme problem There is strong correlation in the variance proportion if 2 or more variables have values 0.50. SAS Collin output Eigen value Condition index Black BW GA 3.18 1.00 0.02 0.00 0.00 0.77 2.03 0.96 0.00 0.00 0.04 9.27 0.01 0.80 0.02 0.001 18.41 0.00 0.39 0.98 Fixing multicollinearity,NICU example Used dummy variables for BW in 100g intervals to 1000g, then 250g intervals. Separate BW dummies for singleton males, singleton females, and multiple births, Gestation in 2 week intervals. Max condition index 8, i.e. no serious collinearity problem. Model predictions also improved. Dummy Variables To Fix Collinearity References Belsley, D.A., Kuh, E., and Welsch, R.E. (1980) Regression Diagnositics. New York, John Wiley Sons. Next lecture Limited Dependent Variables Ciaran Phibbs November 13, 2013 * * * * * * * * * * * * * * * * * * * * * * * * * Right Hand Side (Independent) Variables Ciaran S. Phibbs November 6, 2013 Independent Variables Regression models make several assumptions about the independent variables The purpose of this talk is to examine some of the more common problems, and some methods of fixing them Focus on things not likely to be covered in standard econometric classes Outline Heteroskedasticity Clustering of observations Functional form Testing for multicollinearity Heteroskedasticity Yi = β0 + βX + εi Assumes that the error terms are independent of xi. Common pattern, as x gets bigger, e gets bigger. Heteroskedasticity Biased standard errors Parameter estimates unbiased, but inefficient Heteroskedasticity Simple solution, robust option in Stata uses Huber-White method to correct standard errors. May also consider transf