Hi,
I'd like to simulate a variable with a negatively skewed distribution, bound by zero on the lower end. The only data available on the structure is the median and IQR.
Is there any advice on how to do this? I have tried following this thread using rbeta but am unsure how to calculate what inputs I need (eg, if my data is median (IQR), 5 (3-13)) - or what extra information I need to be able to use this.
I have also looked at sknor (ssc install sknor) as a possible resolution. But again, unsure of how to calculate my inputs aside from an estimate of the mean and variance.
Thank you for any help.
Megan
I'd like to simulate a variable with a negatively skewed distribution, bound by zero on the lower end. The only data available on the structure is the median and IQR.
Is there any advice on how to do this? I have tried following this thread using rbeta but am unsure how to calculate what inputs I need (eg, if my data is median (IQR), 5 (3-13)) - or what extra information I need to be able to use this.
I have also looked at sknor (ssc install sknor) as a possible resolution. But again, unsure of how to calculate my inputs aside from an estimate of the mean and variance.
Thank you for any help.
Megan
Comment