Published on *OpenMx* (http://openmx.psyc.virginia.edu)

By *Hansjansen*

Created *08/22/2013 - 17:10*

Thu, 08/22/2013 - 17:10 — Hansjansen [1]

Hello everybody!

I am new here and I would like to start of by asking you a question regarding a design I made in AMOS. It is meant as part of my graduation thesis in Supply Chain Management.

I am following the structural model of one paper, which uses first- and second-order factors, however, it reports only a simplistic version of this model and it is hard for me to identify if I replicated the model well.

Information about my model:

- My independent latent variables are IC, CO and IG;

- My dependent latent variables are C, Q, FLEX, INN and FIN;

- My mediating variable and at the same time a second-order factor is SCC measured by latent variables: IS, T, IA and JD,

Information about article's model:

- Article's independent latent variables are: Interaction, Coordination, Integration

- Articles dependent latent variables are: Cost, Quality, Flexibility, Innovativeness

- The article's mediating and 2d-order factor is Buyer-supplier collaboration and according to the article it is measured by latent variables information sharing, joint decision-making and incentive alignment.

I would like to know if the input of my model into Amos was correctly done as the article does not show how measures of the second order factor were built into the model. In the attachment you may see the article's simplistic model and the version of my model built in AMOS.

Furthermore, the article reports the following test, which I did not come across in your videos. Quote:

"Statistical evaluation of the second-order factor model can be evaluated by comparing the fit of the first-order model with that of the more restrictive second-order model. The target coefficient (T), which is the ratio of the x2 for the first-order model to the x2 of the second-order model, provides an indication of the degree to which the second-order model accounts for the relations among the first-order factors (Marsh and Hocevar, 1985; Tanriverdi and Venkatraman, 2005). Conducting confirmatory factor analysis and evaluating both a first-order and second-order factor model yields a (T) coefficient of 0.984."

I would like to know whether this paper ran the first-order model without including Buyer-supplier collaboration as a second-order factor, meaning that the latent variables information sharing, joint decision-making and incentive alignment were direct mediators between the independent and dependent variables? And afterwards with Buyer-supplier collaboration and used the ratio of two chi-squares?

Any ideas are highly welcome. Many thanks in advance.

My sincere thanks,

Hans

Attachment | Size |
---|---|

amos model.png [2] | 175.56 KB |

Screen Shot 2013-08-22 at 9.42.51 PM.png [3] | 112.16 KB |

**Links:**

[1] http://openmx.psyc.virginia.edu/users/hansjansen

[2] http://openmx.psyc.virginia.edu/sites/default/files/amos model.png

[3] http://openmx.psyc.virginia.edu/sites/default/files/Screen Shot 2013-08-22 at 9.42.51 PM.png

[4] http://openmx.psyc.virginia.edu/thread/2332

[5] http://openmx.psyc.virginia.edu/thread/2269

[6] http://openmx.psyc.virginia.edu/forums/opensem-forums/general-sem-discussions