Invariance Testing with Mplus - School of Social Work：不变性测试与Mplus -社会工作学院.docxVIP

下载本文档

29
0
约1.96万字
约 9页
2019-11-30 发布于江西
举报
版权申诉

Invariance Testing with Mplus - School of Social Work：不变性测试与Mplus -社会工作学院.docx

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

v.2 N.K.Bowen, 2014 PAGE 1 TESTING FOR DIFFERENCES IN MEASUREMENT (CFA) MODELS USING MPLUS’S INVARIANCE SHORTCUT CODE (WLSMV) Measurement invariance occurs when the parameters of a measurement model are statistically equivalent across two or more groups. The parameters of interest in measurement invariance testing are the factor loadings and indicator intercepts (when ML estimation is used) or indicator thresholds (when WLSMV is used with categorical variables). Invariance in residual variances or scale factors can also be tested, but there is consensus that it is not necessary to demonstrate invariance across groups on these parameters. Researchers typically hope to demonstrate that their measures are invariant across groups. When loadings and intercepts or thresholds are invariant across groups, scores on latent variables can be validly compared across the groups and the latent variables can be used in structural models hypothesizing relationships among latent variables. A recent update to Mplus (7.11) provides a convenient shortcut for conducting a sequence of increasingly restrictive invariance tests. This handout gives some non-Mplus-specific information on invariance testing, information on Mplus non-shortcut invariance testing, and then invariance testing using the Mplus shortcut. We focus on tests of invariance with measurement models based on ordinal variables—the most common type of scale variables in social science research. When scales are based on ordinal measures, it is appropriate to use WLSMV estimation and a polychoric correlation matrix for the analysis. Specifying that observed indicator variables are ordinal automatically leads to the use of WLSMV and the polychoric correlation matrix. Note that Mplus allows two different approaches to the specification of multiple group CFAs: delta and theta parameterization. Delta is the default, but we recommend using Theta. Theta parameterization lets you obtain information on residual variances (unexpla