Common method bias

[Tija Ragelienė]

Dear Terrence,

I did like this:

Com=~a*SN1 + a*SN2 + a*SN3 + a*SN4 + a*VPF1 + a*VPF2 + a*VPF3 + a*SSEF1 

+ a*SSEF2 + a*SSEF3 + a*NPAF1 + a*NPAF2 + a*NPAF3 + a*FBTPF1 + a*FBTPF2 

+ a*FBTPF3

to make constrains equal but then in the output I get them equal to a and not to some variance number as it was in the video I sent to you about common method bias estimaiton with AMOS.

So, should I add some extra line to the model? or ask R to give me some kind of parameters?

Best,

Tija

[Tija Ragelienė]

On Thursday, August 29, 2019 at 7:46:36 AM UTC+2, Tija Ragelienė wrote:

Dear Terrence and Yves,

I did like this:

Com=~a*SN1 + a*SN2 + a*SN3 + a*SN4 + a*VPF1 + a*VPF2 + a*VPF3 + a*SSEF1 

+ a*SSEF2 + a*SSEF3 + a*NPAF1 + a*NPAF2 + a*NPAF3 + a*FBTPF1 + a*FBTPF2 

+ a*FBTPF3

to make constrains equal but then in the output I get them equal to a and estimate is 1 and not to some variance number as it was in the video I sent to you about common method bias estimaiton with AMOS.

[Tija Ragelienė]

if I ask R to give me stand parameters

I get this warning message:

Warning messages:
1: In sqrt(ETA2) : NaNs produced
2: In sqrt(ETA2) : NaNs produced
3: In sqrt(ETA2) : NaNs produced

On Thursday, August 29, 2019 at 7:46:36 AM UTC+2, Tija Ragelienė wrote:

[Mauricio Garnier-Villarreal]

When you do this

Com=~a*SN1 + a*SN2 + a*SN3 + a*SN4 + a*VPF1 + a*VPF2 + a*VPF3 + a*SSEF1 

+ a*SSEF2 + a*SSEF3 + a*NPAF1 + a*NPAF2 + a*NPAF3 + a*FBTPF1 + a*FBTPF2 

+ a*FBTPF3

lavaan will estimate 1 parameter that will be the same for all the factor loadings. If you want to equate them to a specific value you could add a == 0.75, for example

Is that what you are looking for?

[Terrence Jorgensen]

You didn't provide any other syntax (namely, your call to lavaan() or any output), so I don't know how you are identifying your relative to how the Amos video assumed you would.  If you are using the default std.lv=FALSE, then the first loading will be fixed to 1 for identification, in which case you can change all your "a" labels to values 1 to equate them.  

If you set std.lv=TRUE, then your syntax should do what you are asking, but that doesn't mean the model is correct for your data.  The warning messages indicate that at least one of your latent variables (ETAs) may have negative variance.  Check your summary() output.

Terrence D. Jorgensen

Assistant Professor, Methods and Statistics

Research Institute for Child Development and Education, the University of Amsterdam

http://www.uva.nl/profile/t.d.jorgensen

 

[Harmanjit Singh]

I would appreciate it if someone can share the sample code to check common method bias using the method (I assume this is a common latent factor approach as showcased in the following video {from 6 - 10 mins} https://www.youtube.com/watch?v=CFBUECZgUuo).

[Rönkkö, Mikko]

Hi

Here is an example using Lavaan:

https://www.youtube.com/watch?v=zqLeboNwo1o&list=PL6tc6IBlZmOX0GpHkTjYzzmDnTALk_sEp&index=6&t=447

Note that ULMC (unmeasured latent method contract) models are rarely identified and are not very useful. You might find the entire playlist useful, though it gets a bit technical at some points.

Mikko

-- 
You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lavaan+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/365c5319-1e1b-448b-8883-74f3a2b42ef0n%40googlegroups.com.