Suppose a survey was conducted with a lower sample size (less than 100) for factor analysis of the items (less than 20). Now would a larger sample size increase the inter-item correlation?
What is the recommended sample size in this case?
Dear Hasan Mahbub Tusher
The sample size is usually recommended to follow response to item ratio between 10 and 20. This implies that the sample size should be a minimum of 200 or 400.
Regarding how the sample size influences the inter-item correlation, I think the collected sample is relatively small, making it more difficult to conclude with high confidence.
However, I can say that if the pilot study results showed a particular tendency either to be positively correlated or negatively, and if the measurement model faced minor issues for passing the validity and reliability tests, combined with respondents who answered with a proper understanding of the items. Therefore, the inter-term correlation should be confirmed and further supported.
Best Regards,
Belal Edries
Dear Belal Edries, many thanks for your detailed reply. Is there any reference for the recommended response to item ratio between 10 and 20?
Dear Hasan Mahbub Tusher
I would like to refer you to the following questions that were asked in this regard.
- https://www.researchgate.net/post/What-is-rule-of-thumb-mean-in-sample-size-estimation-Is-there-a-book-that-clearly-describes-the-rule-of-thumb
- https://www.researchgate.net/post/What_is_the_sample_size_for_an_unknown_population
I hope the discussion they have would fill full your question.
Best Regards,
Belal Edries
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size