Non-Response Bias

[spiking]

I like to double check if we can run the non-response bias (early versus late respondents) by using the PLS-MGA?

That is, I divide into two groups of data: early versus late.

Then after running PLS-MGA, I look under the p value column to see if less than 0.05 or more than 0.95 to determine significance?

Please advise. Thank you.

[Hengkov]

Hi,

Many researchers prefer to use Levene test or Hartley test to test it.

Best regards,

[spiking]

In the latest SmartPLS 3, when you run the PLS-MGA, it automatically removes the need to check for Levene's test.

We just need to see the p value to determine the significance of the difference in the path coefficients of the two groups of data.

** p<0.05 or p>0.95