資料與概念趨勢訊息於登錄階段的語文遮蔽效果

The Verbal Overshadowing Effect of the Data-Driven and Conceptural-Driven Information During the Encoding Stage

Paivio 的雙碼理論（Dual-coding Theory）認為，語文與視覺刺激的處理歷程會相互輔助，而促進刺激的記憶表現。但Schooler and Engstler-Schooler（1990）發現視覺刺激在提取階段，有可能因語文處理歷程的存在，而產生干擾的現象。Marks（1991）更發現它會隨著視覺刺激的屬性差異─資料趨勢（date-driven）與概念趨勢（conceptual-driven），而有不同的影響結果：視覺刺激的資料趨勢訊息，不論語文處理歷程是否出現，都不會出現語文遮蔽效果；視覺刺激的概念趨勢訊息，只有當個體以視覺處理歷程處理視覺刺激，才不會出現語文遮蔽效果；而當個體採取語文處理模式時，則因語文遮蔽效果而干擾表現。本研究主要探討刺激登錄階段中，不同的視覺刺激屬性與語文處理歷程間的關係，以及其可能導因。實驗一主要是利用Marks（1991）的實驗，重新檢驗不同視覺刺屬性於登錄階段的VOE，是否有所差異。研究者藉由操弄「登錄型式」與「視覺訊息類別」的方式，檢驗此二獨變項之間的交互作用。實驗結果證實Marks的假設在登錄階段，仍然可以成立。實驗二主要探討不同刺激屬性於登錄階段出現VOE的原因。研究者以Dodson， Johnson， and Schooler（1997）的研究作為實驗理論根據，並假設：視覺刺激的概念趨勢訊息於登錄階段時，若尚未以語文處理歷程登錄，則將可以「提示」的方式避免VOE的干擾；但若個體已將視覺刺激以語文處理歷程登錄，則將無法以「提示」的方式避免VOE的干擾。研究者以受試者是否親自描述視覺刺激的方式，作為是否完成語文處理登錄的依據。研究結果支持假設：概念趨勢訊息，若以「親自描述」的方式進行語文登錄，則不論是否給予提示，都將因VOE而抑制辨識表現；若以「閱讀他人描述」的方式進行語文登錄，則可以提示的方式，讓個體區辨視覺與語文訊息的差異，進而避免VOE的干擾作用。資料趨勢訊息，不論以何種方式進行登錄，均不出現VOE。

According to the Dual-coding Theory， Raivio claimed that the processes of verbal and visual stimulus are complementary， which will facilitate memory performance. However， Schooler and Engstler-Schooler（1990）found that， at the retrieval stage， the process of visual stimulus will be interfered with by the verbal process， named the verbal overshadowing effect（VOE）. Also， Brandimonte， Hitch， and Bishop（1992a， b）found this interference occurred both at the memory and encoding stage. Moreover， Marks（1991）found the VOE varies with the category of visual stimulus：the date-driven and conceptual-driven information. According to his research， the data-driven information of visual stimulus will not cause the VOE no matter whether the verbal process is present or not. The conceptual-driven information will not either， if the subjects handle their visual stimulus with the visual process. But， if they handle the visual stimulus with the verbal process， the conceptual-driven information will then cause the VOE and eliminate the memory performance.This research focused on the relationships between the category of visual stimulus and the verbal process at the encoding stage， and intended to sort out the cause of the VOE. Based on the framework of Marks’ experiment， experiment 1 was to verify the hypothesis：At the encoding stage， different categories of visual stimulus will cause different VOEs. The researchers manipulated two independent variables：the type of encoding process and the category of visual stimulus. The result confirmed their hypothesis.Experiment 2 was to find the cause of the VOE at the encoding stage. The researchers used the research of Dodson， Johnson， and Schooler（1997） as the basis of their theory. And they hypothesized that if the conceptual-driven information of visual stimulus， at the encoding stage， have not been encoded by the verbal process， the subjects can avoid the VOE by the hint of ignoring the verbal encoding. If they have already been encoded by the verbal process， the hint will be useless. The researchers take the verbal describing types， describing the visual stimulus by himself or reading others’ descriptions， as the manipulation which the subjects have or have not encoded. The results confirmed this hypothesis. If the subjects encoded the visual stimulus according to self-description， no matter whether there is a hint， the conceptual-driven information will always cause the VOE and eliminate the memory performance. If they encoded according to others’ descriptions， by using the hint we can avoid the presence of VOE. Of course， the data-driven visual information will never cause the VOE under any conditions.