ABSTRACT
Semantic similarity across distractor types. Consistent with Kay et al.’s claims, there is a significant difference in semantic similarity across the three distractor types (Page’s L test: L=426.00, z=– 6.97, p=.000). Furthermore, close semantically related distractors are rated as significantly more semantically similar to their targets than both the distant semantic distractors (t=8.803, df 39, p=.000), and the unrelated distractors (t=24.961, df 39, p=.000). In addition, the distant semantic distractors were rated as significantly more semantically related than the unrelated distractors (t=12.012, df 39, p=.000). This semantic similarity gradient (CSD>DSD>URD) is true for 39/40 (97.5%) items. This is not the case for item 35 “needle” where the unrelated distractor (tweezers) is rated as more semantically similar to the target than the distant semantic distractor (spinningwheel). Ideally, all items should consistently show semantic similarity gradients that are in the same direction.
