The DHS Program User Forum: Wealth Index » Wealth Index Pakistan 2017-2018

Home » Topics » Wealth Index » Wealth Index Pakistan 2017-2018

Show: Today's Messages :: Show Polls :: Message Navigator

Wealth Index Pakistan 2017-2018 [message #19394]

Wed, 10 June 2020 07:34

NVUKIT
Messages: 1
Registered: June 2020

Member

Dear reader,

For my thesis I'm using the DHS data set. I'm only using the district of the city of Karachi. The distribution of wealth index is in most of the districts heavily skewed to the left. I want to try and get a normal distribution, but I can not figure out how to do this. A log-transformation wont work because there are negatives and '0's' in the data set. So i wanted to ask if there is another way to transform the variable so I can possibly get a normal distribution?

Report message to a moderator

Re: Wealth Index Pakistan 2017-2018 [message #27430 is a reply to message #19394]

Mon, 14 August 2023 04:28

ernestsmedley
Messages: 2
Registered: August 2023

Member

I suggest Box-Cox Transformation. The Box-Cox transformation is a power transformation that can handle data with both positive and zero values. It allows you to find the optimal power parameter lambda (λ) that transforms the data to achieve a more symmetric distribution. The Box-Cox transformation is commonly used for normalizing skewed data.

Report message to a moderator

Previous Topic:	Can't find variables
Next Topic:	Lesotho Urban/Rural Ns

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sat Sep 23 17:09:06 Coordinated Universal Time 2023