The Japanese Journal of Psychonomic Science

Sixty-one subjects were asked to judge subjective truthfulness of single propositions and conjunctions in order to investigate how the human integrates fuzzy logical information. The single propositions and conjunctions used in this study stated class-membership relations for vehicle category and furniture category. Three rules for descriptions of cognitive operations that determine the subjective truthfulness of conjunctions, that is, minimum rule, multiplying rule, and averaging rule, were compared. The root mean-squared deviation of the rule from the data showed that the minimum rule best approximated subjects' cognitive operations. However, regression analyses using Akaike's information criterion (AIC) revealed that subjects used both the product of the truthfulness of the component propositions and the average truthfulness of the component propositions.

View full abstract

The present study examined the extent to which imageability of sentences depends upon imageability of their component parts. In Experiment I, subjects rated the sentences and their component parts (3 propositions and 5 words for each sentence) on imagery scale. Multiple-regression analyses were performed on the imagery ratings, regarding the imagery ratings of sentences as dependent variable and the imagery ratings of words or propositions as independent variables. The results indicated that the imageability of words made significant contributions to the imageability of sentences, but the imageability of propositions did not make significant contributions. In Experiment II, type of postposition ("ga" and "wa") and word order (SOV and OSV) of the sentences were varied. The results indicated that the sentence involving "ga" was rated higher than that involving "wa", and the SOV sentence was rated higher than the OSV sentence. Multiple-regression analyses indicated that contributions of the independent variables to the dependent variable were different between the SOV sentence and the OSV sentence. Theoretical implications of these findings were discussed.

View full abstract

Show abstractHide abstract

The present paper focussed on the effects of the figure orientation on the magnitude of the Ponzo illusion. The Ponzo illusion figures used in this study were composed of the inducing figures and two induced lines. Experiment 1, in which inducing figures were two converging lines, examined whether there were systematic deviations of the illusion magnitudes according to the orientations of the inducing lines or not. The orientations of the converging lines were top-converging, right-converging, bottom-converging and left-converging. The illusion magnitude of the top-converging figure was maximum and the difference of the illusion magnitude between the top-converging and the bottom-converging figures was significant. The illusion magnitudes of the right-converging and the left-converging figures lay between the top- and the bottom-converging figures. In Experiment 2, the effects of the main factors, the orientations of the inducing figures, the distance between the standard and the variable lines, and the kinds of inducing figures were all significant. In addition, the effects of the orientation of the inducing figures were significant in the gradient texture and the subjective contour figures.

View full abstract

Download PDF (988K)

To examine the effects of script type on search performance, an experiment was conducted in which subjects searched for a target in a prose passage/a scrambled-letter passage written in hiragana only or in kanji and kana. The target used was a symbol, an inverted-letter, or a misspelling. In a symbol search task requiring a graphemic matching between target representation and each background item, there was no difference in search time between two script types. In an inverted-letter search task requiring an identification of each item and a decision of its orientation, the target in the passage written in hiragana only was searched faster than in kanji and kana. In a proofreading task requiring semantic processing, on the other hand, the prose passage written in kanji and kana was read faster. These results showed that the effect of script type depended upon the processing demand required for the task.

View full abstract

Register with J-STAGE for free!