![least square means asreml least square means asreml](http://www.biosci.global/wp-content/uploads/2020/06/Data1-300x288.png)
We suggest that the use of the second method.
![least square means asreml least square means asreml](https://img.yumpu.com/6919982/1/500x640/asreml-tutorial.jpg)
Back-transformed means ‘estimate’ the medians of the original populations, which may be regarded as better measures of central tendency for skewed data.
![least square means asreml least square means asreml](https://www.mdpi.com/dairy/dairy-02-00049/article_deploy/html/images/dairy-02-00049-g001.png)
Accordingly, if we have done, e.g., a logarithmic transformation, we can exponentiate the means of transformed data and report them back to the original measurement unit. Furthermore, please remember that the means of original data may not be a good measure of central tendency, if the original population is strongly ‘asymmetric’ (skewed)! This is clearly less than optimal, if we want to suggest more than the bare variability of the observed sample.