Box-Cox transformation coefficient (lambda) in SPSS or R
1
0
Entering edit mode
5.6 years ago
tesfadej2003 ▴ 10

Hello all, I want to transform one variable, which is skewed, to a normal distribution. An appropriate transformation method should be decided based on Box-Cox transformation coefficient (lambda) value. However, I couldn't calculate in SPSS or R.

Anyone who have experience?

Thanks in advance.

Box-Cox Lambda SPSS R • 9.4k views
ADD COMMENT
0
Entering edit mode

Can you add more detail on your specific problem? - there's a box-cox method in the MASS package, and there's even an R-Bloggers tutorial

ADD REPLY
0
Entering edit mode

First of all, thank you very much. I have seen both links you mentioned.

I have non-normally distributed continuous outcome variable. Therefore, I need to transform to a normal distribution using an appropriate method that is selected based on Box-Cox transformation coefficient (λ) value. The decision should be made as follows: λ = 1.00: no transformation needed; produces results identical to original data λ = 0.50: square root transformation λ = 0.33: cube root transformation λ = 0.25: fourth root transformation λ = 0.00: natural log transformation λ = -0.50: reciprocal square root transformation λ = -1.00: reciprocal (inverse) transformation and so forth.

Now, I have got difficulty to calculate λ using SPSS or R.

ADD REPLY
1
Entering edit mode
5.5 years ago
tesfadej2003 ▴ 10

One of my friends advised me to use forecast package in R.

install.packages("forecast")

library(forecast)

lambda2<- BoxCox.lambda(dataname$variablename, method = "loglik", lower = -20, upper = 20)

Then, if lambda (λ);

λ = 1.00: no transformation needed; produces results identical to original data
λ = 0.50: square root transformation
λ = 0.33: cube root transformation
λ = 0.25: fourth root transformation
λ = 0.00: natural log transformation
λ = -0.50: reciprocal square root transformation
λ = -1.00: reciprocal (inverse) transformation and so forth.

If the lambda (λ) estimate is none of the above value, use another transformation method. At least in my experience, I can recommend the two-step transformation method in SPSS. 

Reference: 

1. Osborne JW. Improving your data transformations: Applying the Box-Cox transformation. Practical Assessment, Research & Evaluation. 2010 Oct;15(12):1-9.

2. Templeton GF. A two-step approach for transforming continuous variables to normal: implications and recommendations for IS research. Communications of the Association for Information Systems. 2011 Jan 1;28(4):41-58.

Box-Cox transformation coefficient (lambda) in SPSS or R - ResearchGate. Available from: https://www.researchgate.net/post/Box-Cox_transformation_coefficient_lambda_in_SPSS_or_R [accessed Mar 27, 2017].

ADD COMMENT

Login before adding your answer.

Traffic: 1490 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6