This evidence is conceptually similar to the 5-item per category (i.e., 10 items altogether) group in Table 2 that had a similar effect size but only marginal significance.
← all excerpts
On the emergence of the in-out effect across trials: two items do the trick.
3
—
—
The sentences
We did not establish such evidence, although the stimuli used in the present experiments were carefully crafted according to the respective phonation rules in German, English, and Portuguese and had produced reliable in–out effects in within-subjects designs in previous publications, and proved reliable in additional within-subjects replications reported here (for Experiments 1a–c). Furthermore, according to our knowledge being the first to implement a pronounceability manipulation in a one-shot between-subjects design, we did find a marginally significant effect (Experiment 1e).
When only 1 item per category was presented, the in–out effect emerged with an already non-neglectable but insignificant effect size of d z = 0.22, increased in effect size approaching significance (mind Bonferroni-correction for multiple testings) when 5 items per category were presented and reached a solid plateau of highly significant effects for 10 to 50 items per category.