Advice Needed: Handling Awareness-Based Missing Data in Influencer Study

[Nikita Singh Dahiya]

Respected Neeraj sir and all group members,

I am conducting a study titled “Examining the Impact of Human Influencers and Virtual Influencers on Consumer Behaviour: The Mediating Roles of Emotional Engagement and Credibility.”

In my questionnaire:

Respondents were asked if they are aware of social media influencers. If they answered “No,” the survey ended there.

Later, they were also asked about awareness of virtual influencers. If “No,” they skipped all VI-related questions.

This has created structural missing data (missing because the section was not applicable). For example, some respondents have data only for HI but not VI, and vice versa.

My main question is: Should I restrict my analysis only to respondents who are aware of both types of influencers, or is there a better way to analyze this dataset?

More specifically:

Can respondents who are aware of only one type (HI or VI) still be included in partial analysis (e.g., multi-group comparisons in SmartPLS)?

Should respondents who are completely unaware of influencers be excluded entirely?

Would forcing all respondents to answer both sections (even if unaware) be methodologically sound, or would that compromise validity?

I would be grateful for your suggestions on the best way forward with this type of dataset.

Best regards,

Nikita

Ph.D. Scholar

SRM University 

[Neeraj Kaushik]

Dear Nikita

This is a very genuine question and the answer lies in the objs.

Actually you've qualifier questions and that's why you'll use samples depending on the obj.

Plz share your obj.

Best wishes

--
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/3082b745-c80f-4a9f-b8e9-5912a73837d9n%40googlegroups.com.

[Nikita Singh Dahiya]

Respected Sir,
The objectives of my study are as follows:

1. To evaluate the impact of human influencers (HI) and virtual influencers (VI) on consumer attitudes, trust, and purchase intentions.

2. To examine the mediating role of emotional engagement in the relationship between human vs. virtual influencers and consumer attitudes, trust, and purchase intentions.

3. To investigate the mediating effect of credibility in the relationship between human vs. virtual influencers and consumer attitudes, trust, and purchase intentions.

4. To compare the effectiveness of human influencers (HI) and virtual influencers (VI) in driving emotional engagement, perceived credibility, and consumer behaviour (attitude, trust, and purchase intention).

I sincerely thank you for your guidance and kind support.
Best regards,

Nikita

To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CAAd%3Dc8MD-zfeGKHez6gxdH3QG6zjf%3D6JpgodGQdH%2B5VfuHAkPQ%40mail.gmail.com.

[Neeraj Kaushik]

Dear NSD

Your IDV influencer is binary (do record their codes as 0 and 1) and everywhere you shd use SEM (first check the reliability and validity of all constructs measured in Likert scale) with Mediation analysis. Use the case (respondents) that are complete in all respects.

Best wishes

To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CAL%3DHsbGrg4ydgkZPDn%3D_w_gQqKJeXUtkGb0jDswT4d_2Wfe%2B-w%40mail.gmail.com.

[Nikita Singh Dahiya]

Respected Sir,

Thank you for your valuable guidance

Best Regards

Nikita

To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CAAd%3Dc8Nwc0OTHmyzEOU2Cq2yYq3H81RNdn6aZr7vQitmMQkWJg%40mail.gmail.com.