We have tried to make log10 of one variable normal. While it is still non-normal, its skewness is reduced by 50%. Can we take the log again, or is there a different technique we should try?
Hi Nidhi,
You can transform your data. Indeed log-transform the variable is one option. If this did not work (completely) you can also try to LN-transform. As you probably know, you can do this with the compute window with point-click or in syntax by typing:
COMPUTE VariableName_LN=LN(VariableName).
EXECUTE.
--> fill in the name of your variable you want to transform instead of 'VariableName'
Good luck!
Hi Nidhi,
You can transform your data. Indeed log-transform the variable is one option. If this did not work (completely) you can also try to LN-transform. As you probably know, you can do this with the compute window with point-click or in syntax by typing:
COMPUTE VariableName_LN=LN(VariableName).
EXECUTE.
--> fill in the name of your variable you want to transform instead of 'VariableName'
Good luck!
You're welcom! You mean transform the already transformed data? I would not do that.. If the data is still not normal after log10, LN and sqrt, I would use non-parametric tests.
I think Muhammed Mustafa is correct. It happened to my data also. I used large size of 710 samples. My data shows huge multivariate outliers, which is doubted as non normal data distribution. I am using SEM, AMOS for data analysis. It is non parametric test as it considers chi-square value. Can anyone correct me if I am wrong.
I have one more question. In reporting the skewness and kurtosis value using SPSS do one need to consider the raw values of these both or z scores of these both?
In some literature it says skewness and kurtosis values should be between -2 and +2. Other studies says the z score of these two much be greater than +/- 1.96. I want to know which values need to documented in thesis to say data is normally distributed.
also you can go to this link:
https://statistics.laerd.com/spss-tutorials/transforming-data-in-spss-statistics.php
do not convert your data
Article TO DETERMINE SKEWNESS, MEAN AND DEVIATION WITH A NEW APPROAC...
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS