Respected scholars,
I am measuring religious commitment of respondents on a scale having 10 statements. six of them represents intrinsic religious commitment and 4 represent extrinsic. When i am analysing the data on AMOS by applying SEM, i am not getting the required results. on closer investigation, i found that the inter correlation between the two sub scales of religious commitment is very high. it is almost .65 in my case. when i am forming religious commitment a second order construct and run SEM, it is giving better results.Do i need to apply second order SEM here or there are some other options?
And i Request, if anybody is having any paper in which second order SEM has been applied, please forward it to me. Will be most kindness.
Thanks and regards
Typically either in ML data structure or hierarchical/bifactor measurment model specification.
Hope this helps,
Matthew
Respected Matthew Sir,
Can you please elaborate a bit more. Like my scale also has bifactor intrinsic and extrinsic.
Hi Parvinder,
I ran a third-order model some years ago:
Article Developing a multidimensional and hierarchical service quali...
Best regards
Jose
You might see this article at
https://www.ijern.com/journal/2015/March-2015/42.pdf
Hi Parvinder,
Sometimes a second-order structure of a measurement can be dictated by a specific theory. See, for example, a measure of the Team Climate Inventory (that represents team climate for creativity and innovation):
Article Creative and innovative behaviours of corporate directors: a...
Hope this helps. Good Luck and all the best!
Thank you Kyoung Sir and lukasz sir. Does that means that there is no fixed level of intercorrelation between the variables to make it a single variable by using second order sem??
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling