Hi - Thanks for the example. Lets say for your example, the specification for the amount of calcium in each tablet is 35. With the transformed data set, how would evaluate the average and standard deviation against the lower specification limit of 35? Would you transform the 35 limit using the Johnson equation? the issue you can't take a natural log of a negative number.
If you are using data transformations, like Johnson Transformation in this example, the original data is transferred to new values. This is also true for its specification limits. It's critical to understand - why do we need a transformation?
@@learnandapply I understand that the specification limit should be transformed too. But if the specification limit is, for example 35, then when you try to transform it, you get an invalid number. The issue is taking Ln of a negative number. Any ideas? The reason to transform is because the data was not normally distributed.
In this case, identify which distribution your data follows and use an appropriate Nonnormal capability analysis for it, instead of a data transformation.
hi, i have a question. if i have 2 datasets one of them follows normal distribution while the another is not. can i use Johnson transformation for both and then use welch t-test to find the difference in the mean is significant?
A histogram will help you to get an idea about the distribution and central tendency. But see the topic; we are learning here about how to transform data to get meaningful information or to transform it to normal distribution. You won't get the transformation function just by using a Histogram. I hope this answers your question.
What if Johnson transformation does not work? The message says "Fail to select a transformation with P-value>0.05. No transformation is made.". In this case, is there any way to figure out process capability?
Can you please elaborate on it to find the reason? Is your data nonnormal and do you want to calculate the capability analysis for it? You can use the Nonnormal capability analysis in case your data is not following the normal distribution.
@@learnandapply Hello, I have a similar issue to the above person. My data fits a Bimodal distribution. Which fit distribution should I select in Minitab as the Bimodal option is not there? Thank you
@@user-si3wu5yl5x Are you looking for data transformation or to identify the distribution of your data? Can you please help me with a detailed explanation of the question?
Very helpful video.. thank you
You're welcome. Thank you so much for your valuable comments and appreciation 🙏☺
Hi - Thanks for the example. Lets say for your example, the specification for the amount of calcium in each tablet is 35. With the transformed data set, how would evaluate the average and standard deviation against the lower specification limit of 35? Would you transform the 35 limit using the Johnson equation? the issue you can't take a natural log of a negative number.
If you are using data transformations, like Johnson Transformation in this example, the original data is transferred to new values. This is also true for its specification limits. It's critical to understand - why do we need a transformation?
@@learnandapply I understand that the specification limit should be transformed too. But if the specification limit is, for example 35, then when you try to transform it, you get an invalid number. The issue is taking Ln of a negative number. Any ideas? The reason to transform is because the data was not normally distributed.
@@Chris-rx5ce Use the appropriate Nonnormal distribution to conduct an analysis. May I know which analysis you want to do?
@@learnandapply For your example, lets say we want to know the process capability (number of standard deviations away from the specification of 35.).
In this case, identify which distribution your data follows and use an appropriate Nonnormal capability analysis for it, instead of a data transformation.
hi, i have a question. if i have 2 datasets one of them follows normal distribution while the another is not. can i use Johnson transformation for both and then use welch t-test to find the difference in the mean is significant?
It's not required. You can directly perform a t-test on the data.
Thank you for your valuable comments.
why cant we simply plot the histogram of the original calcium data and see if it looks like a normal distribution?
A histogram will help you to get an idea about the distribution and central tendency. But see the topic; we are learning here about how to transform data to get meaningful information or to transform it to normal distribution.
You won't get the transformation function just by using a Histogram.
I hope this answers your question.
What if Johnson transformation does not work? The message says "Fail to select a transformation with P-value>0.05. No transformation is made.".
In this case, is there any way to figure out process capability?
Can you please elaborate on it to find the reason?
Is your data nonnormal and do you want to calculate the capability analysis for it?
You can use the Nonnormal capability analysis in case your data is not following the normal distribution.
@@learnandapply Hello, I have a similar issue to the above person. My data fits a Bimodal distribution. Which fit distribution should I select in Minitab as the Bimodal option is not there? Thank you
@@user-si3wu5yl5x Are you looking for data transformation or to identify the distribution of your data? Can you please help me with a detailed explanation of the question?
sunny same problem Fail to select a transformation with P-Value > 0.05. No transformation is made. how we can do in this case
Use the LRT p-value for the selection of the correct distribution.