All Academic, Inc. Research Logo

Info/CitationFAQResearchAll Academic Inc.
Document

Does c’ Test Help, Anytime? -- On Communication Fallacy of “Effect to Mediate”
Unformatted Document Text:  Does c’ Test Help, Anytime? -- On Communication Fallacy of “Effect to Mediate” Page 13 of 34 large relative to c effect. When it is large enough, it can make c’ error too large, rendering c’ not significant. 3) c reduces the effect of c’ and increases its error c’. This could occur when a*b is significant, c is not, and a*b and c bear opposite signs. Although c cancels out some effect of a*b, it is not enough to make c’ non-significant. It needs the additional error from c, which is by definition large. Possibility 2) is probably the least intuitive because it does not involve any suppression. To show it is not just a hypothetical possibility, we re-analyzed data from a published study (Zhao et al. 1994). The researchers investigated the effect of reading newspapers on Beijing residents’ opinions regarding the economic reform in the late 1980s. Frequent readers were more sympathetic toward the reasons for the reform, and sympathizers were more supportive of the reform. Newspaper reading, however, did not have a direct effect on the support for reform. These expectations constitute an indirect-only model in which newspaper reading affects attitude-reform through attitude-reason. This is clearly a non Shrout-Bolger condition. There was no a priori belief of suppression. The process was not temporally distal, as the effects should take place within days or hours if not instantaneously. The effects were expected to be sizable, as there were no opposing opinions in the media. A stratified random sample of 870 residents in Beijing were visited in their homes and interviewed through a questionnaire. In addition to the dependent, independent and mediating variables, a list of control variables were also measured. More information about the data and the analysis can be found in the original publication. Table 2 shows the results of bootstrap tests using Preacher and Hayes (2008) SPSS macro ( www.comm.ohio-state.edu/ahayes/SPSS%20programs/indirect.htm ):

Authors: Zhao, XinShu., Chen, Qimei. and Tong, Bing.
first   previous   Page 13 of 34   next   last



background image
Does c’ Test Help, Anytime? -- On Communication Fallacy of “Effect to Mediate” 
Page 13 of 34 
 
large relative to c effect.  When it is large enough, it can make c’ error too large, rendering c’ not 
significant.  
3) c reduces the effect of c’ and increases its error c’. This could occur when a*b is 
significant, c is not, and a*b and bear opposite signs.  Although c cancels out some effect of a*b
it is not enough to make c’ non-significant. It needs the additional error from c, which is by 
definition large. 
 
Possibility 2) is probably the least intuitive because it does not involve any suppression. To 
show it is not just a hypothetical possibility, we re-analyzed data from a published study (Zhao et al. 
1994).  The researchers investigated the effect of reading newspapers on Beijing residents’ opinions 
regarding the economic reform in the late 1980s.  Frequent readers were more sympathetic toward 
the reasons for the reform, and sympathizers were more supportive of the reform.  Newspaper 
reading, however, did not have a direct effect on the support for reform. These expectations 
constitute an indirect-only model in which newspaper reading affects attitude-reform through 
attitude-reason
 
This is clearly a non Shrout-Bolger condition. There was no a priori belief of suppression. 
The process was not temporally distal, as the effects should take place within days or hours if not 
instantaneously. The effects were expected to be sizable, as there were no opposing opinions in the 
media.   
 
A stratified random sample of 870 residents in Beijing were visited in their homes and 
interviewed through a questionnaire. In addition to the dependent, independent and mediating 
variables, a list of control variables were also measured.  More information about the data and the 
analysis can be found in the original publication.  
Table 2 shows the results of bootstrap tests using Preacher and Hayes (2008) SPSS macro 
(
www.comm.ohio-state.edu/ahayes/SPSS%20programs/indirect.htm
):  


Convention
All Academic Convention can solve the abstract management needs for any association's annual meeting.
Submission - Custom fields, multiple submission types, tracks, audio visual, multiple upload formats, automatic conversion to pdf.
Review - Peer Review, Bulk reviewer assignment, bulk emails, ranking, z-score statistics, and multiple worksheets!
Reports - Many standard and custom reports generated while you wait. Print programs with participant indexes, event grids, and more!
Scheduling - Flexible and convenient grid scheduling within rooms and buildings. Conflict checking and advanced filtering.
Communication - Bulk email tools to help your administrators send reminders and responses. Use form letters, a message center, and much more!
Management - Search tools, duplicate people management, editing tools, submission transfers, many tools to manage a variety of conference management headaches!
Click here for more information.

first   previous   Page 13 of 34   next   last

©2012 All Academic, Inc.