Inferential Confidence Intervals and Equivalence Testing推理的置信区间和等价测试.pptVIP

下载本文档

4
0
约5.59千字
约 22页
2018-06-24 发布于四川
举报
版权申诉

Inferential Confidence Intervals and Equivalence Testing推理的置信区间和等价测试.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

Inferential Confidence Intervals and Equivalence Testing推理的置信区间和等价测试

Differences, Equivalence, Indeterminacy Inferential Confidence Intervals and their Application to Equivalence Testing Confidence interval approach As suggested by many leading quantitative psych guys, using confidence intervals in lieu of emphasizing p-values may make for a better approach for evaluating group differences etc. Confidence intervals can provide all the relevant NHST info as well as a means of graphically displaying statically significant differences One approach is to simply see if our interval for the statistic of interest contains the H0 value. If not, reject H0 Group CIs This is how CIs are often displayed Note how using individual CIs shows overlap, leading one to think that there is no statistical difference The error term for each group is not the same as that used in the actual t-test Confidence Intervals With mean difference scores we would lose individual group data (not that we couldn’t add it in table form) However, might there be a way to retain the individual group mean intervals and show statistical difference? Inferential CIs Tryon, 2001; Goldstein Healy, 1995 Provide a technique for giving CIs in which non-overlap suggests statistical significance at the level specified by the critical value used. Note that for the one sample case we’d just go about business as usual, and see if the H0 value falls within the interval. i.e. nothing would change. Inferential CIs For group differences all that is required is a correction (reduction) to our CI equation Specifically the critical value Reduction term The ratio of the standard error of the difference between means to the sum of the respective group standard errors Holds for unequal sample sizes as is computed individually for each sample Inferential CIs What it boils down to Inferential CIs So now we can compute the group mean CIs as: And now non-overlap suggests statistical significance at the specified ? level Graphically Now we have no overlap and can clearly discern statistica