Super video. Many thanks. Dear professor, can you give me a reference that contains the computation of the correlation implied by a non-recursive model. Many thanks for considering my request.
If you are looking for a cite how the calculation is done in AMOS, then I would use: Arbuckle, James L. (2017). IBM SPSS AMOS 25 User’s Guide, IBM corporation If you are looking for the method on how to analyze non-recursive models, you can cite my book. Link is in the description
Hi Dr. Collier. I have a question. No estimates are provided in the tables at around 7 minutes for the relationships between the instrumental variables and the endogenous variables which they are instrumental in explaining. Is this because those paths have been constrained to zero? I'm not familiar with the theory behind these variables but it seems that one could make an argument for the existence of a relationship between gratitude and trust which would preclude gratitude from being included as an instrumental variable. Do instrumental variables have to be shown to demonstrably not have a relationship with the variable they are instrumentally predicting, or can we constrain them to zero if we feel a sufficiently small influence exists? The reason I ask is that I have read a few papers where it seems the authors have based their selection of instrumental variables on the fact that the instrumental relationship has not been examined in-depth in the prior literature. But just because a relationship hasn't been examined doesn't mean that the relationship doesn't exist, right?
Lot to unpack here so I'll tackle it one piece at a time. Why are there no estimates from the instrumental variable and the DV it is helping to explain? For instance, in the example used in the video Gratitude only has a direct relationship to Confidence in the Retailer. Gratitude is supposed to help explain Trust but it has no direct relationship. This is done purposely so in the AMOS. The instrumental variable can only have a relationship to one side of the feedback loop. In essence, it creates the identification for one side of the feedback loop. As for your second question, your are correct you could make an argument for that Gratitude to Trust is present. The argument of the researcher is that the influence of Gratitude flows through "confidence of the retailer" construct to Trust. It does not have a direct effect (or a weak effect)....thus allowing it to be an instrumental variable of trust. I know that is a little confusing and overly complicated. As for the third part, does a construct like gratitude have to show a near zero relationship with Trust for it to be an instrumental variable. The short answer, no. This is where the theory is going to come into play and make the argument that gratitude has more of an influence on confidence to the retailer construct and indirectly it impacts Trust. Last part, Just because a relationship has not been examined before in no way says the relationship doesn't exist. It simply means no one has tackled that question yet. Hope that helps.
@@joelcollier9387 Thanks a lot, Dr. Collier, that's very helpful. I have a follow-up question which I I was hoping you could give me your opinion on. I have specified a model for an indirect feedback loop (A-->B-->C-->A) as opposed to the A-->B-->A relationship discussed in the video. My model is similar to the one found in Kline (2016, p.151). I know that this structure is difficult to specify, but assuming I specified it a priori based on sound theoretical foundations, can this model be run in AMOS? I only ask because although I have found quite a few research models employing the structure outlined in your video, I haven't found many with the A-->B-->C-->A as per Kline.
I could not properly describe it in the comments, it is just too lengthy. If you want a full step by step process, I would encourage you to check out my book "Applied Structural Equation Modeling using Amos". It has step by step instructions. Here is a link: www.amazon.com/Applied-Structural-Equation-Modeling-using/dp/0367435268/ref=tmm_pap_swatch_0?_encoding=UTF8&qid=1668697577&sr=8-1
Thanks for putting out this video
Super video.
Many thanks. Dear professor, can you give me a reference that contains the computation of the correlation implied by a non-recursive model.
Many thanks for considering my request.
If you are looking for a cite how the calculation is done in AMOS, then I would use:
Arbuckle, James L. (2017). IBM SPSS AMOS 25 User’s Guide, IBM corporation
If you are looking for the method on how to analyze non-recursive models, you can cite my book. Link is in the description
@@joelcollier9387 It's clear. Many Thanks
Hi Dr. Collier. I have a question. No estimates are provided in the tables at around 7 minutes for the relationships between the instrumental variables and the endogenous variables which they are instrumental in explaining. Is this because those paths have been constrained to zero? I'm not familiar with the theory behind these variables but it seems that one could make an argument for the existence of a relationship between gratitude and trust which would preclude gratitude from being included as an instrumental variable. Do instrumental variables have to be shown to demonstrably not have a relationship with the variable they are instrumentally predicting, or can we constrain them to zero if we feel a sufficiently small influence exists? The reason I ask is that I have read a few papers where it seems the authors have based their selection of instrumental variables on the fact that the instrumental relationship has not been examined in-depth in the prior literature. But just because a relationship hasn't been examined doesn't mean that the relationship doesn't exist, right?
Lot to unpack here so I'll tackle it one piece at a time. Why are there no estimates from the instrumental variable and the DV it is helping to explain? For instance, in the example used in the video Gratitude only has a direct relationship to Confidence in the Retailer. Gratitude is supposed to help explain Trust but it has no direct relationship. This is done purposely so in the AMOS. The instrumental variable can only have a relationship to one side of the feedback loop. In essence, it creates the identification for one side of the feedback loop. As for your second question, your are correct you could make an argument for that Gratitude to Trust is present. The argument of the researcher is that the influence of Gratitude flows through "confidence of the retailer" construct to Trust. It does not have a direct effect (or a weak effect)....thus allowing it to be an instrumental variable of trust. I know that is a little confusing and overly complicated. As for the third part, does a construct like gratitude have to show a near zero relationship with Trust for it to be an instrumental variable. The short answer, no. This is where the theory is going to come into play and make the argument that gratitude has more of an influence on confidence to the retailer construct and indirectly it impacts Trust. Last part, Just because a relationship has not been examined before in no way says the relationship doesn't exist. It simply means no one has tackled that question yet. Hope that helps.
@@joelcollier9387 Thanks a lot, Dr. Collier, that's very helpful. I have a follow-up question which I I was hoping you could give me your opinion on. I have specified a model for an indirect feedback loop (A-->B-->C-->A) as opposed to the A-->B-->A relationship discussed in the video. My model is similar to the one found in Kline (2016, p.151). I know that this structure is difficult to specify, but assuming I specified it a priori based on sound theoretical foundations, can this model be run in AMOS? I only ask because although I have found quite a few research models employing the structure outlined in your video, I haven't found many with the A-->B-->C-->A as per Kline.
Hello sir, for the mediation value in the bootstrap P-value = 0.021.
Lower bound=0.351
Upper bound=1.251.
Is this significant or not?
If the lower bound to upper bound confidence interval does not cross "0", then it is significant at .05 level
@@joelcollier9387 What about the above conditions?
Can it be called significant or not?
@@makmursujarwo9458 Yes, since it does not cross zero then it is significant at the .05 level.
hello sir,How to set path in non recursive model ?how to do please give step by step process?Am a beginner of non reursive model as well as AMOS
I could not properly describe it in the comments, it is just too lengthy. If you want a full step by step process, I would encourage you to check out my book "Applied Structural Equation Modeling using Amos". It has step by step instructions. Here is a link: www.amazon.com/Applied-Structural-Equation-Modeling-using/dp/0367435268/ref=tmm_pap_swatch_0?_encoding=UTF8&qid=1668697577&sr=8-1