| 摘要: | 本研究利用自我參照作業 （self-referencing task） 探討自我概念 （self-concept）是否以及如何在隱性比較的情況下受到手足的影響。四個實驗皆包含了兩個學習—測驗循環，在學習階段中，受試者需要判斷一系列的形容詞是否符合自己或者他人。受試者和他/她的手足會是其中一循環的指涉對象，而擁有與手足相似熟悉程度的朋友則是另一循環的指涉對象。在測驗階段中，我們將使用不同的作業來測試受試者對於形容詞的記憶（實驗一的再認測驗、實驗二的排除作業、實驗三和四自由回憶）。指涉對象（自己或是他人）以及是否有血緣（手足或是朋友）兩個變項，將用來分析肯定率（自我參照作業中回答「是」的比例）和記憶表現。
實驗一的結果觀察到受試者在與手足配對時，對自我評估有較高的肯定率，與朋友配對時則並無顯著差異，說明和比起朋友，受試者和手足配對時對於自己有更高的肯定。再認記憶測驗中，對自己的記憶表現也比他人更好，是為自我參照效應，且在朋友情境中的整體記憶表現比手足情境更好。然而，兩個情境中的自我參照效應大小並無差別。
在手足情境中，對自我評估有較高的肯定率一結果，在實驗二中仍然被觀察到。此外，為了探討朋友情境中的記憶優勢，是否來自於不同的檢索方向或是回憶策略，實驗二採用排除作業並記錄腦波。透過手足情境和朋友情境中，正確拒絕拒絕反應（new item correct-rejection）的腦波差異，發現了顯著的檢索方向效應，說明檢索記憶在兩個情境中的處理方式是不同的。
實驗三則使用特異刺激典範（oddball paradigm）探討朋友情境的記憶優勢是否來自學習階段中，挹注了不同程度的注意力而導致記憶表現更好。在自我參照作業中，四分之一的形容詞用來判斷是否符合自己，而四分之三的形容詞用來判斷是否符合他人。自己和他人所誘發的P300和N400波形差異，將作為手足情境和朋友情境中注意力分配和語意處理的指標。此外，聚類比率分數 （adjusted ratio of clustering）也將用來計算受試者回憶字詞的組織程度。腦波結果顯示在手足情境的P300效應比朋友情境更大，表示和朋友情境相比，受試者在手足情境中分配了更多的注意力。但兩種情境下的聚類比率分數皆無差別，說明回憶字詞的組織程度並沒有因為不同的指涉對象而有所不同。
為了確認實驗三聚類比率分數的結果不是因指涉對象比例不同所致，實驗四將自己和他人的比例調整為一致。實驗四還使用四組短的刺激材料取代兩組長的刺激材料，來降低實驗難度。然而，兩個情境的聚類比率分數依然沒有差異，驗證實驗三的結果並非來自比例的不同。
以上實驗顯示，自我評估會受到手足的影響，進而產生較高的自我肯定率，和朋友配對時則無這現象。同時，實驗中也觀察到自我參照效應，即與自己相關的記憶表現較他人更好。但朋友情境的記憶優勢僅止於一般記憶，並未包含更多細節。雖然對手足和朋友的記憶表現沒有顯著差異，但神經生理結果顯示，兩者在判斷形容詞和回憶檢索時可能有不同的處理機制。
;This study employed the self-referencing task to investigate whether and in what way self-concept could be modified when an individual is being implicitly compared with his or her siblings. Each of the four experiments reported in this thesis comprised two study-test cycles. During the study phases, participants were presented with a series of adjectives, to each of which they had to judge whether it described themselves or another referent well. The participant’s sibling was designated as the companion referent in one of the two cycles and a friend that was of similar familiarity in the other cycle. In the following test phase, the participants’ memory of these adjectives was tested with an old/new recognition test (Experiments I), exclusion task (Experiment II), or free recall task (Experiments III and IV). The affirmation rates (i.e., the proportion of yes responses in the referencing task) and the memory performance were both analyzed with the factors of referent type (self vs. other) and consanguinity (sibling vs. friend).
Experiment 1 showed that the affirmation rate to self was higher in the sibling block than in the friend block, suggesting that the participants were more affirmative about their self-concept when the companion referent was their siblings in comparison to friends. A significant self-reference effect was also observed during the test phase as the adjectives that were referred to the participants themselves were better remembered than those that were referred to the participant’s siblings or friends. Moreover, the recognition memory performance was in general better in the friend block than in the sibling block. The mnemonic self-reference effects were however of equivalent magnitudes in the sibling and friend blocks.
The finding of a higher affirmation rate to the self in the sibling block than in the friend block was replicated in Experiment II. In addition, Experiment II employed the exclusion task and recorded event-related brain potentials (ERPs) to investigate whether the better memory performance in the friend block than in the sibling block resulted from retrieval processes such as retrieval orientation and strategic retrieval. A reliable retrieval orientation effect, indexed by the distinct ERP patterns between the correct rejection trials in the sibling and friend blocks, suggested that the retrieval cues were processed differently in these two blocks.
Experiment III employed the oddball paradigm to investigate mnemonic processes during the encoding phase contributed to better memory performance in the friend block than in the sibling block. In the study phase of both cycles, one-fourth of the adjectives were related to self and the other three-fourths were related to the non-self referent. The P300 and N400 waves elicited by the self and non-self referents were compared to examine the attentional resource allocation and semantic processing in the friend and sibling blocks. In addition, the adjusted ratio of clustering (ARC) score of the recall performance was employed to examine the memory organization in the sibling and friend blocks. The ERP analysis found a greater P300 effect in the sibling block compared to the friend block, suggesting greater attentional allocation when the companion referent was the sibling in comparison to friends. The ARC scores were statistically equivalent in the sibling and friend blocks, revealing no evidence that friends and siblings as different companion referents exerted effects on memory organization.
To ensure that the null result of the ARC analysis in Experiment 3 was not due to the different proportions of self and companion referents in the oddball paradigm, Experiment IV replicated the procedure of Experiment 3 with the modification that the ratio of self to companion referent was 1. In addition, the task difficulty was decreased by presenting four short study lists instead of two long lists. The results from both blocks still remained no difference which verified that the proportion did not influence the results from the previous experiment.
Across the four experiments, the self-evaluation was influenced by the siblings with consanguinity rather than friends. The self-reference effect was observed, while the memory advantage for friend block only exist in general memory but contained less details. Despite that there was no significant difference on memory performance between sibling and friend, neural physiological results suggested that there could be differences in the processes of judging adjectives and retrieval orientation. |
