Assessing Convergent and Discriminant Validity of Your Measures in SEM Model

After initially establishing that each indicator is loading on its respective construct and the model has an acceptable fit to the data, we now need to assess the convergent and discri- minant validity of your measures. Just to recap, the basic difference between convergent and discriminant validity is that convergent validity tests whether indicators

If I Have a Weak Factor Loading in SEM Model, Should I Drop It?

If you have a factor loading that is near or below the .70 threshold, it does not mean you need to drop the indicator from the analysis. Complex or newly developed constructs will often have numerous indicators in an attempt to capture a comprehensive aspect of a construct. If you have numerous indicators that

How to Calculate Composite Reliability for SEM Model

I have discussed earlier that reliability of indicators can be explored by calculating Cronbach’s alpha, but this test has had its detractors. Yang and Green (2011) has noted that Cronbach’s alpha is known to overestimate and underestimate values and is “not to be construed as a coef- ficient of internal consistency but rather an

What If I Have a Standardized Factor Loading Greater Than 1 in SEM Model?

A standardized factor loading greater than 1 is stating that you are explaining more than 100% of the variance in an indicator. In this instance, you will also see a negative number in your error term, which is often called a “Heywood case”.The causes of a Heywood case are often the result of an

How Do I Report My CFA Results of SEM Model?

One of the challenges in performing research is analyzing the data, but an equally challenging task is how to report my findings when I am writing it up for a journal or publication. I will give you two templates to follow for writing up CFA results in a table format.There are many accept- able

Measurement Model Invariance Across Groups in SEM Model

Up to this point, we have discussed performing a CFA with a single group. With SEM requir- ing a large sample size, it is pretty common to collect data from two different methods such as online and face-to-face surveys. You will also come across instances where you are surveying two different groups but still

Full vs. Partial Metric Invariance in SEM Model

The test described in the previous section is considered a “full” invariance test. When a model gets more complex, it is increasingly difficult to achieve full invariance. Partial invariance is where multiple estimates/factor loadings per construct need to be equiva- lent (but not all). Based on Hair et al. (2009), if two parameters per

How Do I Know Which Parameter to Free in a Partial Invariance Test in SEM Model?

If you have a significant chi-square invariance test, you need to see what indicators/factor loadings are differing across the groups.You can look at the individual factor loadings for each group and see where you have large differences across the groups. It may just be one construct causing the significant invariance test. Sometimes the differences

Common Method Bias in SEM Model

Over the last 10 years, more and more attention has been paid to the idea of common method bias (CMB) in the measurement analysis phase. Common method bias is the inflation (or in rare cases deflation) of the true correlation among observable variables in a study. Research has shown that because respondents are replying

What If My Test for Common Method Bias Is Significant in SEM Model?

You will need to include the common method variable and its relationships to all the indi- cators when you start testing the structural relationships between constructs. By including the common method variable in the testing of the structural relationships, you are con- trolling for the potential bias of the common method. If you want

What if My Common Method Bias Test Is Non-Significant in SEM Model?

You can state in your research that common method bias is not a substantial concern in your research (do not state that it is not present—it could be present but at very low levels). In your research, you could even present the chi-square difference test showing that common method bias was not prevalent. With

What Is a Second Order Confirmatory Factor Analysis Model in SEM?

A second order CFA is a “higher order” construct that is measured by latent constructs. For instance, let’s say we want to model a construct called “Unique Experience”. This construct is the degree to which customers think their experience was unique or different compared to a normal experience. This second order construct of “Unique

Reflective vs. Formative Indicators of SEM Model

Based on classical test theory (Lord and Novick 1968), when an unobservable variable causes changes in its indicators, the specification of these indicators is reflective (For- nell and Bookstein 1982) or effect indicators (Bollen and Lennox 1991). By far, this is the most common type of measurement model in which the direction of influence

Potential Identification Problems With Formative Indicators in SEM Model

A formative construct will have identification problems in SEM if additional steps are not taken. The first way to rectify the identification problem is to include two reflective indica- tors from the formative unobservable construct. These two reflective indicators should try to capture the overall theme of the formative construct.You need to know if

Dependent Relationships With Formative Constructs in SEM Model

As stated earlier, a formative construct is defined by its indicators. A change in a formative unobserved construct cannot take place without a change also taking place in the indicators. Hence, I think it is a good idea to discuss the appropriate way to model relationships in a SEM model with formative constructs. If

Higher Order Formative Construct Example of SEM Model

Let’s look at an example of a higher order (second order) construct that has formative indi- cators in AMOS. Using our example from earlier in the chapter, the higher order construct of Unique Experience is formed by two first order constructs of Surprise and Empathy.We have also included two reflective indicators off the higher

How Do I Present the Result of My Higher Order Formative SEM Model?

The results of the CFA for the first order dimensions should be presented in a dedicated CFA table. For the first order constructs’ relationship to the higher order construct, I prefer to put this information in a structural relationship table. I also like to include the actual wording of the identification items to give

Error Messages in AMOS

If you encounter an error message in AMOS while testing your measurement model, here are some of the most common causes I have found along with potential remedies. The model is unidentified—after running the analysis, this error message will appear in the “Notes for Model” link of the output. Figure 4.52 Error Message for

Path Analysis in SEM Model

So far, we have talked only about the measurement model; now let’s talk about the struc- tural model.The structural model’s focus is on examining the relationships between constructs. Hence, we will look at how independent constructs influence dependent constructs or, in more complex models, how dependent variables influence other dependent variables. The first type

Can I Use a Correlation Matrix as My Data Input?

Yes, AMOS allows you to use a correlation matrix as data input, but you need to include the standard deviation, mean, and sample size for each construct. Typically, you will see path models used when the data input is a correlation matrix instead of the raw data. For many correlation analyses, the researcher is