A minimum of 100 respondents is required to conduct SEM using AMOS.

It depends on the number of constructs and items per construct. You could try SmartPLS if you have a very low sample size. It sometimes works better with low sample sizes.

According to Kline , minimum of 200 of samples is needed. For better analysis the sample should be of 400.

Dear Almaamari

The minimum sample size for SEM observation, depends on your analysis:

a) for high magnification observation, you need at least some particles (of powder)

b) for low magnification observation, line scan and map you need at least a sample with size of 1*1 mm

regards

It's a complicated issue. One good discussion paper here, which includes many useful references.

http://repository.um.edu.my/99406/1/Critical%20Story%20in%20Sample%20Size.pdf

In my practic it is good to have as in statistics, 30-100 of one simple size, ann the same number for two other sample size

thank you for all above answers

– Generally SEM requires LARGE sample size

– Kline (2005) suggests: < 100 Small sample size

100 – 200 Medium sample size

> 200 Large sample size

– Critical sample size = 200 (Hoe, 2008)

– Minimum ratio of 10:1 (Ho, 2006)

– Bentler & Chou (1987) suggest a ration of 5 to 1

From Prof. Dr Bahaman Abu Samah ' notes

I suggest: 10 participants for each Item, for example, if a questionnaire includes 20 item, you need 200 participants as minimum sample size. Based on (Ho, 2006)

Many thanks for you Dr Ahmed

As Ronán Michael Conroy mentioned, it is relatively difficult to answer. However, researchers facing sample size problems may use Bootstrapping to overcome the issue.

thanks a lot Dr Masood