管理学中的问卷研究法(3).ppt

OCB OCB Similarity (X) .17 .11 OBSE (M) .23** Adjusted R2 .10 .36** DR2 must significant Mediated regression Y M X Partial Mediation Y M X Full Mediation Full or partial mediation X Y M Mediated regression Y Y M Individualism (X) .17 .19 .23** Team Resistance (M) .21* Adjusted R2 .10 .36** .28** Y M X Not mediator Y M X Mediator Mediated regression A significant total effect of X on Y (quantified as c) is not necessary for mediation to occur. Bootstrap method to test the indirect effect (Preacher Hayes, 2004; 2008) /ahayes/macros.htm  An example An example Moderated mediation relationships Which stage are your moderating effects?---axz5, bmz20, or bxz20? Does the mediating relationship (X-M-Y) be different across different level of the moderator(Z)? Moderated mediation relationships Edwards and Lambert (2007: PM) “Third, and most importantly, Edwards and Lambert (2007) have shown that the Muller, Judd, and Yzerbyt procedures for assessing moderated mediation are subject to several conceptual and statistical flaws”. /faculty/edwardsj/downloads.htm An example Equation 5: M = a05 + aX5X + aZ5Z +aXZ5XZ + eM5, Equation 20: Y = b020 + bX20X + bM20M + bZ20Z +bXZ20XZ + bMZ20MZ + eY20. An example An example * Take Away Lessons Good survey measures must be grounded on sound theory and conceptual definitions Avoid convenience measurement at all time!!! Ensure your analyzing methods match with your questions. 谢谢！ 联系信息：10号楼317室 电话Email：mnjian@ * * * * * 管理学中的问卷研究法 梁建 安泰经济与管理学院 2011年3月 * Part 3. Data Analysis in Survey Research 3.1 Confirmatory factor analysis 3.2 Cross level analysis 3.3 Mediators and moderators * 3.1 Confirmatory Factor Analysis Confirmatory factor analysis of all items from the same source Separate measurement model from structural model Use mean score or factor score? Using parcels when number of items are is large Saving your SPSS/SAS/STATA program file Never trim items based on EFA and then retest with a CFA using the