Dear Dr. Grande, somehow I need to normalize skewed distribution and find your video. I thought I mistook you as a different person, the one who did psychological analysis. Finally, I find out that you are the same person. Wow, you are really talented. Also, thanks for this explanation, which is very digestible and absorbable.
Thanks Dr. Todd Grande!! why do you add 1 to the difference in the log10? and how do you interprete logaritmic values when analysing relationship between data? thanks
He added one so that the minimum value in the data set after reflection would be one. Without adding one, the log10 of zero would blow things up. Note that you can add any constant, it doesn’t have to be one. In this case adding one is mathematically convenient.
Sir suppose we ve three groups to compare. And scores of one group need to be transformed. If in that case we can use transformed score along with non transformed score for comparison??
I'm wondering how the transformations under the Levene's test are different or is it just a preview to what will work. It was suggested in the Field book.
In that case, instead of adding one to the max value you would add a constant that would make you smallest value after reflection equal to one. Then when you take the log, you min value will be zero.
![OsamaSon - ik what you did last summer [Official Music Video]](http://i.ytimg.com/vi/4DIQDnu3mr4/mqdefault.jpg)
Dear Dr. Grande, somehow I need to normalize skewed distribution and find your video. I thought I mistook you as a different person, the one who did psychological analysis. Finally, I find out that you are the same person. Wow, you are really talented. Also, thanks for this explanation, which is very digestible and absorbable.
nice to know you this is a very brief description of data transformation(log transformation) method i really like it.
After doing this and finding mean and standard deviation... should we find the exponential for both to go back to their original values?
Thanks Dr. Grande!
Thanks Dr. Todd Grande!! why do you add 1 to the difference in the log10? and how do you interprete logaritmic values when analysing relationship between data? thanks
He added one so that the minimum value in the data set after reflection would be one. Without adding one, the log10 of zero would blow things up. Note that you can add any constant, it doesn’t have to be one. In this case adding one is mathematically convenient.
How to analyze log 10 scores?
what if u have a fair few 0s in your dataset?
Sir suppose we ve three groups to compare. And scores of one group need to be transformed. If in that case we can use transformed score along with non transformed score for comparison??
I'm wondering how the transformations under the Levene's test are different or is it just a preview to what will work. It was suggested in the Field book.
wow! didn't know Excel had all these functions available
Thanks helped a lot
You're welcome!
Just was I was looking for and thanks for covering the difference between positively and negatively skewed distributions.
You are welcome - thank you for watching -
what if half my dataset is in the negatives?
In that case, instead of adding one to the max value you would add a constant that would make you smallest value after reflection equal to one. Then when you take the log, you min value will be zero.
Hi,
Do we need to transform the whole data set or just pick few that is not normally distributed
You must transform the whole data set otherwise you couldn’t really interpret the results in any meaningful way.
Makes a lot of sense, thanks.
Thank you
please next time speak loudly, so we can hear you … thanks