Seeking clarification on reporting EFA, KMO & Bartlett’s in Pilot Study (PLS-SEM with adapted scales)

[SWEETY DAS]

Dear All,

I hope you are doing well.
I’m writing to seek methodological clarification regarding the pilot study section of my doctoral research, which employs PLS-SEM for data analysis.

Context:
My study uses adapted measurement scales (not entirely new, but contextually modified) for constructs like Experiential Marketing, Brand Relationship Quality, Green Brand Innovation, Brand Loyalty, and Brand Advocacy. The final analysis is being conducted through SmartPLS.

In my pilot study, I collected around 102 responses and initially checked only convergent validity and reliability (outer loadings, Composite Reliability, and AVE) using SmartPLS. However, while reviewing other theses and methodological guidelines, I noticed that many researchers also report Exploratory Factor Analysis (EFA), along with KMO and Bartlett’s Tests, even when the main analysis is based on SmartPLS (PLS-SEM).

Hence, I’m confused about what exactly should be reported in the pilot section for a PLS-SEM-based study.

My specific queries are:

1. Since I have used adapted scales, is it mandatory or advisable to run EFA (with KMO and Bartlett’s) for the pilot study, even though my main analysis relies on the SmartPLS measurement model?

2. If I already checked reliability and convergent validity in SmartPLS for the pilot dataset, would that be considered sufficient — or should I still perform and report EFA for scale validation before confirmatory analysis?

3. What is the ideal sequence and structure for reporting the pilot study in the thesis for a PLS-SEM approach (e.g., Content Validity → KMO/Bartlett’s → EFA → Reliability → Convergent Validity)?

4. Finally, under what circumstances is it acceptable to skip EFA entirely when using adapted scales in SmartPLS-based research?

I want to ensure that my pilot study section is methodologically sound and consistent with doctoral research expectations.

Thank you in advance for your time and guidance.
Looking forward to your advice on how best to report these steps in my thesis.

Sincerely,

Sweety Das

Assistant Professor

Research Scholar- Amity University, Noida

Contact No : 9654046037

New Delhi, India

[Shahzada Irfan Bashir]

At the doctoral level, the most robust and defensible path is to perform and report both EFA and the SmartPLS measurement model assessment on your pilot data. This demonstrates thoroughness and a deep understanding of scale development and validation processes, which will satisfy even the most critical thesis reviewers.

1. Is EFA mandatory for adapted scales?

No, but it is highly advisable. For a doctoral thesis using adapted scales, EFA provides essential evidence of a clean factor structure before confirmatory analysis, strengthening methodological rigor.

2. Is SmartPLS assessment sufficient?

No. SmartPLS (loadings, CR, AVE) assesses reliability and convergent validity confirmity. EFA first checks the underlying factor structure exploratory.

They are complementary steps for adapted scales.

3. What is the ideal reporting sequence?

1. Content Validity: Expert review of adapted items.

2. EFA: Report KMO (>0.70), Bartlett's Test (p < .05), and factor loadings. Remove problem items (cross-loadings > 0.40, low communalities).

3. Measurement Model (SmartPLS): Report outer loadings (>0.708), Composite Reliability (>0.70), and AVE (>0.50) on the purified scales.

4. When can you skip EFA?

Only if using verbatim, well-established scales in a identical context. For adapted scales, skipping EFA is not recommended for a doctoral thesis.

.........................

Shahzada Irfan Bashir

Research Interests:
Strategy, Entrepreneurship, Family Business/Firms, Innovation

Email: profsh...@gmail.com
          

--
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/CAJ8y0n_K0TCnFH0Hd3WgFdjeNj6Oiq2j6HebjmvTrd8Qe3P-Bw%40mail.gmail.com.

[RUTH MATHEWS 2090027]

Greetings, 

Whenever a researcher conducts his study using either new scales or modified scales, the EFA and CFA ( structural model and measurement model) are mandatory. It is important to check the reliability and validity of our modified instrument as it is based on new sampled population.

Addressing 

1. EFA should be run on your unique dataset  using SPSS
2. The reliability ( alpha, composite reliability, ) and validity ( convergent, discriminant validity) should be checked for the entire dataset using SMART PLS

3. During pilot testing, the Cronbach alpha test should be run. The EFA helps to know what indicators should be included in the final study ( adding/ deleting). I used SPSS for pilot testing. 

The CFA consists of a structural model (SM) and a measurement model (MM). The SM and MM was run in Smart PLS where the following things were reported - indicator reliability, cronbach alpha, composite reliability, convergent validity, discriminant validity.  

4. We should always run EFA for even an adapted scale as it's on a new sampled population. 

--
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/CAJ8y0n_K0TCnFH0Hd3WgFdjeNj6Oiq2j6HebjmvTrd8Qe3P-Bw%40mail.gmail.com.

--

We grow through what we go through

Yours Sincerely 

Ruth Mathews

PhD Scholar

CHRIST( Deemed to be University), Delhi NCR Campus

[Neeraj Kaushik]

Just a small addition to the replies in this thread:
Sweety's question is about the pilot testing results.

EFA shd not be used on pilot testing as typical pilot testing sample size is around 30-40 respondents and EFA shd be applied on a min sample of No of items in scale x 10

For the pilot testing data, we can apply Cronbach's alpha and check the sequence of questions, item wording, and clarity of instructions.

[SWEETY DAS]

Dear All,

Thank you so much for taking the time to respond to my query. I truly appreciate your inputs and guidance.  It has helped me gain better clarity on the matter.

To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CAAd%3Dc8OfA%3DBx9nFsMVM3qNpTnvNrsf2qSAPQZ5RMMGfQ9KXiCg%40mail.gmail.com.