DABEST is a data analysis tool that is intended to make estimation statistics more accessible to scientific communities. Estimation statistics is a superior alternative to null hypothesis significance testing (NHST), in which effect size and confidence intervals are used to interpret results as opposed to dichotomous significance testing.
This code allows the user to visualize the data as scatterplots; calculates the effect size and confidence intervals of the difference between multiple groups; and plots the results on the same figure. This figure design allows for a visual inspection of the observed values distribution, and displays the differences between multiple groups of data.
DABEST-Matlab can be installed via MATLAB Central (https://www.mathworks.com/matlabcentral/fileexchange/65260-dabest-matlab) or GitHub (how to clone a repo: https://help.github.com/articles/cloning-a-repository/).
Moving beyond P values: Everyday data analysis with estimation plots
Joses Ho, Tayfun Tumkaya, Sameer Aryal, Hyungwon Choi, Adam Claridge-Chang
Nature Methods 2019, 1548-7105. 10.1038/s41592-019-0470-3
Paywalled publisher site; Free-to-view PDF
Please report any bugs on the Github issue tracker.
All contributions are welcome; please read the Guidelines for contributing first.
We also have a Code of Conduct to foster an inclusive and productive space.
DABEST is also available in R (dabestr) and Python (DABEST-Python).
Data should be in the csv file format and contain two columns with the headers: Identifiers and Values.
Identifiers are the labels of each data point, and Values are the data points (see the example below).
Note: All the sample data used in this tutorial are taken from S. Champely's anscombe2 dataset, and can be found in DABEST-Matlab/SampleData/.
Depending on the number of groups the data contain, the main function dabest produces various plots, and returns the key information as a table object.
If the data have two different groups, dabest('TwoGroups_sample.csv') generates a two groups plot.
Running dabest('TwoGroups_sample.csv','Paired') generates a paired plot with the two groups data.
If the number of groups is an even number, a multiple groups plot will be automatically generated by dabest('MultipleGroups_sample.csv') command.
If there are more than two groups in the data, dabest('MultipleGroups_sample.csv') generates a shared control plot.
To combine two groups of data and compare to a third group, run dabest('MergedGroups_sample.csv','mergeGroups').
For the data that contain more than three groups -and a number that is divisible by 3, dabest('MultipleMergedGroups_sample.csv','mergeGroups') generates a multiple merged groups plot.
If the data contain more than three groups, dabest('MultipleMergedGroups_sample.csv','mergeGroups') automatically generates a second plot in which all the groups are compared to the merged shared control.
