One of Glossa Psycholinguistics' primary goals is communicating trustworthy, solid science to a broad psycholinguistic community. To this end, Glossa Psycholinguistics prefers statistical analyses that address the key scientific questions in a way that is maximally transparent and accessible to its broad readership. The journal does not require its authors to adopt any particular set of statistical tools (e.g. Null Hypothesis Significance Testing or Bayesian approaches): We encourage authors to adopt the statistical tools and analyses they believe are best suited for the question at hand, and ask that reviewers and editors critically and constructively evaluate the authors’ choices, suggesting more appropriate tools and analyses where appropriate.

Instead, statistical analysis for submissions to Glossa Psycholinguistics should follow three broad principles. Authors should seek to follow these principles in communicating the results of their research, and reviewers in evaluating submissions they receive. The three principles are:

1) TRANSPARENCY: Authors should aim to be maximally transparent about the analytical choices they made in analyzing, presenting, and interpreting their data. There are many ways to create this transparency, including pre-registration, clearly delineating a priori and post-hoc hypotheses, and distinguishing exploratory analysis from confirmatory hypothesis testing. For this reason, releasing the data and code along with a publication in Glossa Psycholinguistics is required (Open Data and Ethics policy), with exceptions considered on a case by case basis.

2) ACCESSIBILITY: Authors are free to adopt the statistical tools best suited for the question at hand, and need not follow any particular tradition (e.g. Null Hypothesis Significance Testing). However, data analysis in Glossa Psycholinguistics should be as accessible as possible to our broad readership. This means that authors should strive to minimize the complexity of the statistical analyses required to answer the scientific question at hand, and minimize assumptions of audience familiarity with specific statistical packages or techniques.

3) CONTEXT: To afford the readers the maximum possible context for interpreting the results of statistical analysis, submissions to Glossa Psycholinguistics should present rich descriptions of data. This includes reporting explicitly defined measures of variability (e.g. confidence intervals) in addition to measures of central tendency, and prioritizing figures over tables for communicating key aspects of the data and analysis.

For further details and guidance, authors may wish to familiarize themselves with best practices for reporting statistical analysis from the Psychonomic Society's statement on Statistical Guidelines.