20. SPSS AMOS | Assessing Normal Distribution of Data

[@talhamansoor7108]

that's great

[@researchwithfawad]

Thanks. I am glad you liked it.

[@ShinewithDr.Pooja26]

thnku so much sir.. interpretation is well understood.. can you pls tell how to report these values in paper & thesis..

[@researchwithfawad]

Thanks for your interest. You can put in the constructs with the items in a table and report with a single statement if the conditions for normality are met.

[@Alhamzah_F_Abbas]

Helpful video as usual. thank you, prof. is it better to rely on SPSS to test normality?

[@researchwithfawad]

Thanks. I am glad you liked it. Both are fine.

[@aliviyaghosh4878]

Thank you for the explanation, sir can you please tell me...if multivariate mormality should fall in the range of +10 to -10 and +1.96 to -1.96 ?

[@researchwithfawad]

Yes please

[@artimalik1126]

Sir, please upload video of bootstrapping and interpretation

[@researchwithfawad]

In editing process. Hopefully those video will be uploaded soon.

[@rasdianaa2073]

Thank you for your sharing prof. Can I get information book or reference which states that 200 and more sample size is considered large enough? Thank you

[@researchwithfawad]

Thanks for watching. Please watch kzbin.info/www/bejne/iafHqpqeft6blZI

[@abadirroba6747]

As wr wb thank you for your videos which are really helpful. in the data you have shown in the video only bivariate normality exist, but the data is multivariate non-normal as the ketosis and corresponding c.r are large. My question is, does multivariate normality is not a concern so long as our data is bivariate normal?

[@researchwithfawad]

Thank for watching. It can be a concern if it too much deviant.

[@abadirroba6747]

@@researchwithfawad Jaza kalla for fast response. some scholars suggested that if most variables have univariate skewnesses and kurtoses in the range -1.0 to + 1.0, distortion of ML and generalized least squares (GLS) chi-squares and standard errors are very unlikely and one can ignore multivariate non normality unless most skewnesses and/or kurtoses are larger in absolute value than 2.0 (Gao, Mokhtarian and Johnston, 2008).

[@syifamohdazlan7821]

Prof what if my normality (skewness and kurtosis) contains both normall distribution and non normally distribution? Can i proceed with bootstrap for mediation? Tq Prof

[@researchwithfawad]

Thanks for watching. Yes you can use bootstrapping.

[@nissu...5673]

Sir do we have to check normality like skewness and kurtosis for constructs or for items? As i have seen researchers conducted it on constructs/ variables instead of each and every items?

[@researchwithfawad]

You can do it on both. First items then constructs.

[@Hellz706]

Asalam alaikum Sir yiu are doing great job May allah bless you I wanted to ask my values of skewness and kurtosis are not less than -0.7 but all values are negative . Is it normally distributed?

[@researchwithfawad]

Thanks. I am glad you liked it. Yes, it is normal.

[@nikimorrison6858]

I have a question: Do I have to check all the data for normal distribution or just the metric variables? My items were measured categorically, therefore normal distribution will not appear... and what about control variables like age or gender? Do they have to be normally distributed? Thank you very much for your videos, they helped me A LOT!

[@researchwithfawad]

Pleasure. I am glad you liked it. As for Normality, please check for skewness and kurtosis of the metric variables.

[@nikimorrison6858]

@@researchwithfawad Thank you for the answer. But AMOS measures skewness and kurtosis for the items/indicators (mixed model method) and so the multivariate normality index (Mardia) will be biased by these or not?

[@nikimorrison6858]

And when I include control variables - do they have to be normally distributed to use the ML-Method?

[@researchwithfawad]

You can use SPSS for Skewness or Kurtosis. Further, you can check them without the controls in the model.

[@nikimorrison6858]

@@researchwithfawad Thank you, that helps.

[@Lisansustu101]

Prof., I wonder why you did you assess the normal distribution of data after measurement analysis? If we assess normality after measurement analysis and delete some of the bad observations, Is this affect the measurement analysis? and Do we have to redo the measurement analysis?

[@researchwithfawad]

Thanks for the interest and query. You can do it before. Right at the start.

[@Lisansustu101]

@@researchwithfawad thank you for fast reply.🙏🙏