These videos explain the concept of ILLMO, but use an older release of the program. They are valuable to help those already experienced in statistics, to understand how well-known statistical methods can be implemented in ILLMO.
Interactive Statistics with ILLMO – basic principles
Shows how to do basic statistics such as ANOVA and linear regression in an interactive way. The example data came from a Gaussian number generator, so that the ground truth is known (averages increase in constant steps of 3 and the standard deviation is 1), and such that the assumptions made by basic statistical tests are valid.
Interactive Statistics with ILLMO – data transformations
Shows how a transformation of data that doesn’t comply with traditional statistical tests (i.e., normally distributed with equal variance across conditions) can greatly improve the statistical analysis. The data were time measurements from an actual Fitts law experiment, performed by 8 subjects, for 9 different conditions (3 target distances x 3 target sizes).
Interactive Statistics with ILLMO – between- and within-subject analysis
Shows how both within-subject and between-subject analyses of data can be performed within a single interface. Methods in between classical within- and across-subject are also available in ILLMO. The data are time measurements from an actual Fitts law experiment, performed by 8 subjects, within 9 conditions (3 target distances x 3 target sizes).
Interactive Statistics with ILLMO – ordinal data analysis
This shows how analyzing discrete (Likert scale) data can be just as easy as analyzing continuous data. The data came from an actual experiment in which subjects needed to classify scatterplots between two variables (with 40 points per plot) into 5 categories: highly positively correlated (+2), positively correlated (+1), uncorrelated (0), negatively correlated (-1) and highly negatively correlated (-2).