About formative indicators

[Manisha Soreng]

Dear Professors and group members,

I have query regarding formative construct. What if I don't have a global item but I have checked VIF and no multicollinearity issue was found. But many of the indicators are showing low weights and loadings. Please suggest possible solution.

With regards,

Manisha Soreng

[Pravash ghosh]

Dear Manisha,

1. Global item is used to check the redundancy analysis, Which is mandatory if you are having a formative construct.

2.If your outer loadings are less than .7 then you have to check the AVE. AVE should be greater than .5.

--
The members of this group are expected to follow the following Protocols:
1. Please search previous posts in the group before posting the question.
2. Don't write the query in someone's post. Always use the option of New topic for the new question. You can do this by writing to dataanaly...@googlegroups.com
3. It’s better to give a proper subject to your post/query. It'll help others while searching.
4. Never write Open-ended queries. This group intends to help research scholars, NOT TO WORK FOR THEM.
5. Never write words like URGENT in your posts. People will help when they are free.
6. Never upload any information about National Seminars/Conferences. Send such information
in personal emails and feel free to share any RESEARCH-related information.
7. No Happy New Year, Happy Diwali, Happy Holi, Happy Birthday, Happy Anniversary, etc. allowed in this group.
8. Asking or sharing Research Papers is NOT ALLOWED.
9. You can share your questionnaire only once.
---
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CAAKWW2XTnenr1MzTxEANQ8TFHCZ7reY5E-K%2B8J7eNrDP3gYUJQ%40mail.gmail.com.

[Anubhuti Dwivedi]

Dear Manisha,

For formative constructs you have to check the significance of the outer weights, not only the size of the weight. Run bootstrapping and check which weights are significant. If few weights are insignificant check the outer loadings of those items and if the loading is 0.5 or above then you should retain that indicator.

0.7 item loading and AVE above 0.5 are thresholds for reflective constructs assessment, not formative constructs.

As you did not include a global single item in your questionnaire, you will not be able to run redundancy analysis which is done for assessing the convergent validity of the formative construct. You can still go ahead with formative assessment based on collinearity analysis and outer weights significance. However reviewers can raise the point of redundancy analysis (for checking construct validity) as that is an important step.

Best,

Anubhuti Dwivedi

To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CA%2BSJB1g2f6uHybSw4EZTBjiZ0Q%2Baqyb3G7L6afacf8mJTAJNeQ%40mail.gmail.com.

[Rajinder Kumar]

A. For Formative construct, global item is essential as redundancy analysis is a compulsory step towards the SEM. 
If you really have no solution, For low factor loadings 1. start removing the items from lower loading, if still not able to get good results, increase the sample, if still not able to get the loading drop this study or use some alternate methods.

For more details you can visit https://www.youtube.com/@NeerajKaushik

On Fri, Mar 27, 2026 at 8:44 AM Manisha Soreng <manishas...@gmail.com> wrote:

--

[Manisha Soreng]

Thank you for your reply. One more question I wanted to ask, I have one construct and it has 6 dimensions for that have constituted Higher order model, and that as a formative construct for that also global item is needed?

To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CAG9B3sW9LNhAn35O6M%3Dh8Xzr4SXU2cZ9Z4Hmjg%2B%2BpH9qEOhfeA%40mail.gmail.com.

[Muhammad R Siregar]

Dear Manisha,

Your formative indicators (or dimensions) are not expected to correlate with one another, so they need a Global Item as an 'anchor' to establish convergent validity. You do Redundancy Analysis. The path coefficient between your formative HOC and the Global Item needs to be ≥ 0.707, meaning that your 6 dimensions explain at least 50% of the variance (R^2 = 0.50) in that Global Item.

Best,

MRS

To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CAAKWW2WYcfyam_SuDj0a%3D%2BS9xGjYOfZU-vRfp8LaB7OG%2BDJdFg%40mail.gmail.com.

[Rajinder Kumar]

For a Formative construct, a global item is essential. 

To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CAAKWW2WYcfyam_SuDj0a%3D%2BS9xGjYOfZU-vRfp8LaB7OG%2BDJdFg%40mail.gmail.com.

[Anubhuti Dwivedi]

Yes it is needed

Regards,

Anubhuti Dwivedi

To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CAAKWW2WYcfyam_SuDj0a%3D%2BS9xGjYOfZU-vRfp8LaB7OG%2BDJdFg%40mail.gmail.com.

[Ehsan Saeedi]

A relevant question in this context, I have a reflective-formative construct (Type II) in my study. 

The construct is represented by two dimensions: knowledge and confidence forming literacy. I am confused here on how to design a single global item for this. 

For example, is it correct to write it like this, "Overall, on a scale of 1 to 5, where 1 means VERY LOW and 5 VERY HIGH, how would you assess your knowledge and confidence in X?" 

Or should I write; "Overall,  on a scale of 1 to 5, where 1 means VERY LOW and 5 VERY HIGH, how would you assess your literacy in X?"

If I use the term "literacy" alone, I fear that respondents would perceive it as knowledge alone. However, if I use both "confidence" and "knowledge," I fear that it would violate the signal global item's psychometric idea of measuring one underlying idea only.  

Experts, I would sincerely appreciate it if you could share your opinion on this, as it has been a while that I have been contemplating it and cannot decide what to do. Thank you.

To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CAG9B3sVgi1TyaCPBY_73-q%3Dh5eeXKFW_c46gkTqtVPjo6DDuZA%40mail.gmail.com.