Regarding the correct use of terminology, the way I was taught was:
"Multivariate analysis: when multiple outcome variables;
Multivariable analysis: when multiple explanatory variables".
However, the terms 'Multivariable' and 'Multivariate' are often interchangeably used in the literature. After reading some articles, I got more confused about the right choice of terms.
Can you share your thoughts on this? Thank you.
The term “multivariable” analysis is commonly used in logistic regression models with a single outcome variable and multiple independent or predictor variables.
“Multivariate” analysis on the other hand, involves multiple dependent/outcome variables with a single set of predictor variables.
Like you said, the two terms are commonly used interchangeably in literature, but using the correct terminology is helpful nonetheless.
I found this reference to be brief and helpful:
Hidalgo B, Goodman M. Multivariate or Multivariable Regression? American journal of public health. 2013;103(1):10.2105/AJPH.2012.300897. doi:10.2105/AJPH.2012.300897.
Cheers!
You are right. Multivariate is used for the analysis with multiple outcomes/dependent variables, whereas multivariable is used for the analysis with multiple explanatory/independent variables. However, they are sometimes used interchangeably in the literature as not many researchers are aware of the difference. To have a more clear picture on the articles, the best way is to refer to the tables they present and their objectives of the studies.
Hope my answer helps you. Thanks.
The term “multivariable” analysis is commonly used in logistic regression models with a single outcome variable and multiple independent or predictor variables.
“Multivariate” analysis on the other hand, involves multiple dependent/outcome variables with a single set of predictor variables.
Like you said, the two terms are commonly used interchangeably in literature, but using the correct terminology is helpful nonetheless.
I found this reference to be brief and helpful:
Hidalgo B, Goodman M. Multivariate or Multivariable Regression? American journal of public health. 2013;103(1):10.2105/AJPH.2012.300897. doi:10.2105/AJPH.2012.300897.
Cheers!
@Aruni Seneviratna: Thank you for taking time to answer; i enjoyed reading the article. Cheers!
Hi Himal!
As used exchangeably in the mentioned article, I'm used to the term 'multiple' (= multivariable) regression analysis. It's good that you make us think about the language we use.
Best
John-Kåre
Would multinomial logistic regression be considered a multivariate analysis therefore, given that there are multiple possible outcomes?
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics