An example motivated discourse of the independent samples t-test and the Welch test
Other
Analyses are given for two example scenarios. Both scenarios comprise two independent samples. One scenario is analysed using the independent samples t-test; the other the Welch test. Commentary on the designs and limitations is given through a qu... Read More about An example motivated discourse of the independent samples t-test and the Welch test.
A primer on validity and design terminology in comparative designs
Other
An overview of terminology (independent and dependent designs, dependent and independent variables) and validity (internal, external, and measurement) is given using six different scenarios. For simplicity of exposition, all examples have factors at... Read More about A primer on validity and design terminology in comparative designs.
An example motivated discourse of the chi-squared test of association (2 by 2)
Other
An example application of the chi-square test of association is given through a set of questions and answers. The emphasis is on methodological aspects and what can be legitimately inferred, if anything. It is strongly highlighted that statistical... Read More about An example motivated discourse of the chi-squared test of association (2 by 2).
An example motivated discourse of the paired samples t-test
Other
An application of the paired samples t-test is used to discuss the logic underpinning the test and to consider what may be legitimately inferred.
A primer on statistical hypotheses and statistical errors
Other
An abstract overview of some of the logic and statistical considerations in quantitative inference is given
A reverse look at p-values
Other
An overview of p-values is given. The usual way of being introduced to p-values is by considering the rigorous development of statistical tests, then their use for decision making. This note takes a reverse view by firstly considering their use in... Read More about A reverse look at p-values.
That assumption of normality
Other
Many statistical tests are based around an assumption of “normality”. The reasoning for this choice of distribution, whether to test, what to test, and how to test for normality is covered along with practical recommendations.
Cohen’s d for two independent samples
Other
An exposition of standardized effect for two independent samples (under an assumption of normality) is given along with an insight into interpretation and reporting