Top-Down Control of Selective Attention: Directed Inhibition of Previewed Distractors in the Stroop Color Naming Task

本研究旨在探討選擇注意力作業中進行注意力控制的歷程。本研究以Stroop色字唸名作業（Stroop, 1935）呈現人們的主動抑制的能力。實驗一到實驗三呈現了干擾物預視的效果。倘若預先呈現線索告知參與者（participants）後續的干擾字為何，參與者能夠利用此一由上而下（top-down）的知識預先抑制干擾字。透過此一主動抑制的歷程，Stroop作業中的干擾字有較弱的干擾效果，而使得參與者有較快的唸色表現。實驗四以中性組與不一致組的對比反映Stroop干擾效果的大小，並顯示干擾物預視確實可以減少Stroop干擾效果。實驗五到實驗十針對干擾物預視效果的可能解釋進行檢驗，這些可能解釋包括干擾字處理加速假說、注意部分干擾字假說、時間區隔假說、顏色數減少假說與可能目標準備假說、以及主動抑制假說。實驗五探討一致組加入對於干擾物預視效果的影響。結果顯示，不一致組仍有穩定的干擾物預視效果；而一致組同樣有干擾物預視效果，但此效果小於不一致組內之效果。實驗六到實驗八探討線索與干擾字間知覺改變的影響。結果顯示，知覺改變後仍然有干擾物預視效果，但其效果比知覺不變下之效果要小。實驗九則引入線索成為目標色而非干擾字的衝突線索組，而顯示衝突線索組的表現慢於中性線索組。實驗十移除線索的預測力，而發現線索不再影響唸色的速度。綜合實驗五到實驗十的實驗結果，注意力的主動抑制假說有最強的解釋力。實驗十一到實驗十六就被動抑制與主動抑制之特質進行比較。被動抑制的指標乃Stroop版本的負向促發效果。實驗十一到實驗十四的結果顯示被動抑制受到記憶需求的影響，而主動抑制則不受到記憶需求的影響。實驗十五與實驗十六則顯示被動抑制無法存續達6,500毫秒，而主動抑制則能夠存續達6,500毫秒。這些實驗顯示了被動抑制與主動抑制的差異。

The current study investigated the effect of intentional inhibition on selective attention. Using a Stroop color naming task, a previewed condition was designed that the distractor of the upcoming Stroop stimulus was presented in advance as a cue. With the foreknowledge of the upcoming distractor, participants were instructed to inhibit this previewed distractor. Experiments 1-3 indicated that participants can benefit from distractor previewing. Color naming was faster when the distractor was previewed than when no cue or no valid cue was presented. By estimating the Stroop interference effect as the difference between the incompatible condition and the neutral condition, Experiment 4 further demonstrated that distractor previewing can decrease the Stroop interference effect. Experiments 5-10 were designed to evaluate hypotheses of distractor previewing, including facilitation of distractor processing, attending part of the cue, temporal segregation, expecting possible targets, and intentional inhibition. Experiment 5 showed significant effect of distractor viewing when incompatible and compatible trials were used. However, the benefit of distractor previewing was decreased when compatible trials were presented. Experiments 6-8 demonstrated significant effect of distractor previewing when cues and distractors differed in forms. Moreover, the benefit of distractor previewing was less when cues and distractors were different in forms than when they were identical in forms. In Experiment 9, there were conflict trials that the previewed item cued the target color rather than the distractor word. In comparison to the neutral condition, a cost was obtained for these conflict trials. Finally, the cue validity was removed in Experiment 10, and no effect of distractor previewing was observed. These results from Experiments 5-10 support the hypothesis of intentional inhibition. Using Stroop negative priming as the index of passive inhibition, intentional inhibition was compared to passive inhibition in Experiments 11-16. Experiments 11-14 showed that while passive inhibition was sensitive to short-term memory load, intentional inhibition was less sensitive to the load of short-term memory. Experiments 15-16 indicated that while the effect of intentional inhibition could last at least 6,500 ms, the effect of passive inhibition could not last as long as intentional inhibition. These results demonstrate differences between passive and intentional inhibition.