Two experiments on goodness and complexity judgments of dot patterns in a square matrix framework were conducted based on studies by Hamada et al. (2016, 2017). Four groups of 52 undergraduates (N=208) judged the goodness and complexity of original and expanded patterns consisting of six 8-dot and six 13-dot prototype patterns and thirty-six 21-dot compound patterns. Rotational/reflectional symmetries were invariant, generating cyclic (Cn) and dihedral (Dn) groups (n=1, 2, 4). Results showed that only complexity increased consistently as the number of dots increased. Adding twelve 8-dot and 13-dot patterns increased the complexity of 21-dot patterns. Apart from the complexity of 8-dot D2 patterns, goodness and simplicity increased one-dimensionally with respect to the order of the matrix. The complexity of 8-dot D2 patterns decreased because of spatial filters on linearity, which did not affect goodness. Concerning the original and expanded patterns, configurations of 21-dot patterns did not influence goodness or complexity. In conclusion, results of goodness and complexity judgments supported our group theoretical model and showed that the complexities of dot count and 8-dot D2 patterns were influenced by physical factors.
In this paper, I clarify the problem of null hypothesis significance testing along with Kruschke (2014) and point out five advantages of Bayesian statistics. First, it is not necessary to convert data for NHST. Second, additional assumptions or corrections are not required. Third, there is no need for any preliminary design of the verification plan. Fourth, the Bayesian approach allows an intuitive interpretation of results. Fifth, the sample size does not cause critical problems. Beyond these advantages, Bayesian statistics can be used alongside frequentism and likelihoodism methods. Finally, I argue it is necessary in science communication to clearly express the researcher’s premise as a prior distribution or likelihood function.
The roles of psychological journals on promoting the reproducibility of psychological research are discussed. These roles contain requiring better statistical expressions, detailed description of research method, pre-registration, replication reports, improvement of peer-review, and accelerating open science. A case example of Japanese Journal of Personality is described.
Although sample-size planning is important for research integrity and useful for designing experiments, it appears to be not yet fully spread in cognitive psychology. The present study examined descriptions for sample-size planning in papers that published in Psychonomic Bulletin & Review and Journal of Experimental Psychology: General. Although some researches determined their sample size without referring to statistical power analysis, they derived the size based on prior research. Not a few researches reported results of power analysis without describing exact values of power and effect sizes and types of targeted effects. To communicate findings more correctly, it should be included descriptions specifying targeted sample sizes and their basis that enables readers to calculate the values irrespective of statistical methods or not. Finally, I discussed the risk of over-emphasizing prior power analysis in peer review.