Mechanisms Of Unconscious Processing In Subliminal Semantic Activation

Subliminal semantic activation (SSA) had been researched for more than 20 years and four research patterns were used, i.e. congruent and incongruent pattern, indirect without direct effect pattern, indirect effect greater than direct pattern and regression analysis strategy pattern. All of these patterns were based on respective assumptions, such as comparativeness assumption, exhaustiveness assumption, inclusiveness assumption, relative sensitivity assumption, no measurement error assumption and so on. And some of them brought about a lot of criticisms to SSA researches.Three cognitive models were used to interpret the results of SSA researches. They were spreading priming model, response competing model and compound cue model. Recently, Klinger's research (Klinger et al, 2000) supported the assumptions of response competing model. However, there were two problems needed to further study. The first one was if there were sensitivity differences between Response Time (RT) and Correct Ratio (CR) on detecting SSA; the other was if different task demands influenced SSA.To study the two problems, two experiments were designed. The sensitivity differences of RT and CR on detecting SSA were examined in experiment one, and the influence of different task demands on SSA was examined in experiment two.The results showed: (1) The SSA was observed in the first experiment. It meant that unconsciousness was dissociated from the research. As the mechanisms was concerned, the result supported the assumptions of response competing model. (2) There was no sensitivity difference between RT and CR on detecting SSA, which meant either RT or CR could be used to interpret unconscious mechanisms. (3) Task demands deeply influenced on SSA. (4) There were sensitivity differences between affective semantic relationship and animacy semantic relationship, the former was easier to process than the latter.