关键词: 多身份追踪, 范畴, 工作记忆, 分组理论

Abstract:

Research on Multiple Identity Tracking (MIT) focused on the influence of differences in properties of targets’ and non-targets’ identities on tracking performance and identity recognition. It has been found that the uniqueness of single property, but not compounded property, facilitated tracking performance. However, the uniqueness of compounded property could facilitate tracking performance in the condition where the property guided or oriented attention towards targets. The facilitating effect was also affected by both the complexity of identity information and working memory load. One problem with the previous research is that the same category of stimuli (e.g., either numbers or letters) served as targets and non-targets. The current study involved two categories of stimuli (i.e., 0-9 and A、C、E、K、N、P、R、T、U、Y) and manipulated the differences between targets and non-targets to investigate the effects of differences in targets-nontargets identity on tracking performance and identity recognition. Two experiments had been conducted with undergraduate students using the MIT paradigm. Experiment 1A included three conditions: no identity difference (e.g., both targets and non-targets were the same numbers), within-category difference (e.g., targets and non-targets were different numbers), and between-category difference (e.g., targets were numbers and non-targets were letters). Results showed a significantly greater tracking performance in the between-category difference condition relative to the no identity difference condition, with within-category difference in between. Experiment 1B simultaneously manipulated the identity differences (no difference, within-category difference, between-category difference) and the number of targets (4 vs. 5) and found a significant interaction between the two factors. That is, when there were 4 targets (6 non-targets), tracking performance for between-category difference and within-category difference was better than for no identity difference; when there were 5 targets (5 non-targets), tracking performance for between-category difference was better than for within-category and no identity difference. And tracking performance for within-category difference was significantly better than for no identity difference. Experiment 2 altered the degrees of between-category difference. Non-targets in the third experiment were 5 letters. Targets were 4 letters and 1 number (low-difference condition), 2 letters and 3 numbers (medium-difference condition), and 5 numbers (high-difference condition). We observed that both tracking performance and identity recognition increased with the increasing degree of difference. The results of the two experiments indicated that between-category identity difference between targets and non-targets facilitated tracking performance and identity recognition in the multiple identity tracking task. These results could be explained by the working memory model, grouping theory and Model of Multiple Identity Tracking (MOMIT).

Key words: Multiple Identity Tracking, category, working memory, grouping theory