恐懼的性質：從視覺注意力到眺匿理論

本研究之目的，乃是在於恐懼作為一種情緒反應，與其他認知因素之關聯，以及在行為上造成的影響。研究整體承襲自Nasar & Jones(1997)的實驗，該研究在夜間的校園空間中讓受試者散步，並使用錄音筆記錄恐懼因素，得到了間接隱匿與退路受阻在校園空間中對恐懼感具有重要地位的結論。間接隱匿出自Appleton(1975)的眺匿理論，該理論基於生物的「偵測威脅」與「迴避威脅」本能而提出，認為提供這兩種特性的環境應該較受到偏好：而這兩種特性反映在視覺上就是眺望與藏匿。同時Appleton也認為，這兩種特性都可以再進一步分為直接(primary，在目前的觀景位置上的情況)以及間接(secondary，從目前的觀景位置推測整個環境中的情況)兩種特性，以及眺望與藏匿之間的平衡可能隨環境而改變的可能。眺匿理論雖然受到多方研究引用，卻經常有解釋與脈絡不一的問題，尤其在日間與夜間的研究上，夜間研究較少且不齊全，更缺乏統合日間與夜間的研究。因此本研究決定先用新的方法對校園空間的恐懼因素及視覺注意力行為進行再檢視，再從這些恐懼因素的特性發展到眺匿理論的補足研究、以及該理論對景觀評價的影響與其隨日夜產生的變化。研究以兩段實驗進行。第一段實驗採取Nasar & Jones(1997)的散步實驗方法，並採用新的儀器：瞳位追蹤儀，試圖透過受試者實地走訪的動態資料收集了解校園空間中的恐懼因素。實驗首先針對校園恐懼路段的特性進行簡單的訪談，再依訪談結果規劃散步路線。接著研究者進行瞳位追蹤儀的散步實驗。受試者在散步實驗後，也被要求回報各路段的恐懼因素。結果雖然引起恐懼的因素並未看出在視覺注意力上有甚麼特別的行為，但是卻得到了恐懼感較低者比較有四處觀看的餘裕的結果，符合恐懼感造成迴避行為的特性；也從散步後訪談的各路段恐懼因素上再度得到了與Nasar & Jones(1997)一致的實驗結果，顯示出間接隱匿與退路受阻在夜間校園空間中具有恐懼的效果，並且進一步指出社會監控與熟悉度兩個降低恐懼的因子。第二段實驗採取照片問卷法。本研究在台大校園內分層隨機抽選42個地點，使用網路問卷、並透過iPad於台大校園內針對每個地點在日、夜的情境下的直接眺望、間接眺望、直接藏匿、間接藏匿進行問卷調查，並依調查結果將各地點進行叢集分析，從每一叢集中選出一個地點，以確保地點具有隨機性與代表性。接著研究者再度使用網路問卷、透過iPad於台大校園內進行便利抽樣，對每個地點的兩張照片進行眺望、藏匿、偏好及恐懼的調查與分析。實驗結果共得197名受試者，結果顯示直接與間接眺望/藏匿在空間中之評值相差不顯著；眺望特性均對偏好有正面效果、對恐懼有負面效果，而藏匿特性白天與偏好無關、夜晚負相關，與恐懼則皆為正相關；眺望特性與偏好的正相關、恐懼的負相關在夜間都較日間強，而藏匿特性與偏好的負相關、與恐懼的正相關在夜間也都較日間強。最後研究者針對實驗結果進行討論，並將討論結果化為對未來研究的建議。瞳位追蹤儀研究並未得到新的恐懼因素，在視覺注意力上僅得到與恐懼感呈現負相關的注視目標：天空、行人、告示牌與總注視次數。這可能是因為引發恐懼的是整體環境脈絡，而不是可以直接注視的實在目標。而從日夜變化的眺匿研究結果，研究者推論Appleton(1975)所提及的日夜反轉的戲劇效果可能確實存在，但是因為藏匿特性可能同時具有偏好與恐懼的兩面性，而導致偏好的效果被恐懼所掩蓋。影響其效果的關鍵因素仍有待研究。本研究建議未來研究於動態空間中操作瞳位追蹤儀應採取短操作時間、多受試者並單一受試環境的策略；進行眺匿理論相關研究時，應注意光線環境之影響、藏匿特性之兩面性，並且採用多元的環境進行研究；在恐懼感相關研究上，應注意可能引起同時引發恐懼感與偏好的各種環境特性，並試圖找出控制恐懼感、或進一步將這些特性導向偏好的關鍵因素。

關鍵字

眺匿理論；恐懼感；眺望；藏匿；瞳位追蹤

並列摘要

The purpose of this study is to see into the relationships and effects of fear, as an emotional response to the environment, on cognitive factors and behavior. The study extends the idea of Nasar & Jones (1997),which asked the respondents to walk through campus space at night, using a recorder to record fear factors. They concluded that secondary refuge and a blocked escape contributes a large percentage of fear in campus space. Secondary refuge is a factor comes from the prospect-refuge theory suggested by Appleton (1975). The theory is based on creatures’ natural of detecting and avoiding threat, believing that environment providing these two characteristics visually – namely “prospect” and “refuge”—should be preferred. Also did Appleton mentioned that these two factor could both be distinguished into two level: primary (directly assessed on present vantage point) and secondary (the situation of the whole environment predicted based on present vantage point), and the balance between them might change along with the environment. The theory was widely cited, but there were often problems caused by differences on interpretations and context, especially on the issue of day and night: there were little studies on night situations, and even less on the integrating of day and night. Therefore, the purpose of this study is to use a new way to examine fear factors in campus and its relationships with visual attention and then step into the complement of prospect-refuge factors’ effect on emotional assessment, and the change caused by shifting of day and night. The study was twofold. The first study replicated Nasar & Jones’s “walk” experiment with a new instrument: eye-tracker to collect fear factor of the campus in a dynamic situation. We first used a simple interview experiment on fearful routes of the campus, then planed the route of the “walk” based on the results. The respondents were asked to walk on the route with eye-tracker, and then had a short interview about the fearful factors during the walk. It turned out that no obvious relationship were found on visual attention and fear factors, but only visual targets having negative relationship with fear: sky, pedestrian, placard and the total times of fixation. The reason of these results might be that less fearful routes allowed respondents to look around more, conforming to the basic theory of fear causing avoidance. Also did we get the same result as Nasar & Jones’s experiment on the short interview, showing secondary refuge and blocked escape would result in fear in campus at night, also showing two more factors: social surveillance and familiarity might reduce fear. The second experiment was a photo questionnaire. We picked 42 positions in NTU campus by stratified random sampling. These positions were photographed in day and night, all photos surveyed on primary prospect, secondary prospect, primary refuge, and secondary refuge. These positions were cluster analyzed on all eight items. We pick a position form every cluster to form the formal questionnaire, ensuring that all positions used in the questionnaire were randomized, and general. Then we survey on day and night pictures of these positions on prospect, refuge, mystery, preference, and fear by internet questionnaire. The respondents were all NTU students, sample collected in NTU campus, using an i-Pad as a sampling instrument. There were 197 respondent obtained, results showed that primary and secondary prospect/refuge are not distinguishable; prospect factors have positive relationships with preference and negative relationship with fear, refuge factors have no significant relationships with preference during daytime and negative relationships during night, and positive relationships with fear. The positive relationships between prospect factors and preference and negative relationships with fear were both stronger at night than day, the relationships between refuge factors and preference or fear are also stronger at night. We discuss on the results, and form them into further research suggestion. We found no particularly new fear factors on eye-tracking research, and on visual attention we only found some targets negatively related with fear. It might be that the key to provoke fear lies in environment context thus forming no actual visual target. Respecting the results of second experiment, we infer that the effect of balance between prospect and refuge mentioned by Appleton (1975) might be supportable; but the refuge symbols could be double-faced, being able to provoke both preference and fear, thus its effect on preference being shadowed by fear. The key to decide its effect require further study. The study suggests that eye-tracking researches in dynamic space should use short-operation, large sample size and simple setting; research regarding prospect-refuge theory should be careful with lighting environment and the double-faced characteristic of refuge symbols, and test in multiple settings if possible; on the research of fear, mind those features that provoke both fear and preference, furthermore, try to find the key factor in these features to control fear and provoke preference.

並列關鍵字

eye-tracking ； fear ； prospect ； prospect-refuge theory ； refuge

參考文獻

3. 江彥政，(2009)，自然環境資訊對心理評價反應影響之模式，博士論文，國立中興大學園藝學研究所，台灣：台中。

6. 柯嘉鈞，(2011)，不同景觀空間類型之眺匿涵構比較研究，國立中興大學園藝系博士論文，台灣：台中。

7. 胡毓權，2012，以眼動軌跡檢驗注意力偏誤假設暨不同負性刺激對注意力偏誤修正訓練的效果，國立國防大學政治作戰學院心理系碩士論文，台灣：台北。

13. Andrews, M. & Gatersleben, B. (2010). Variations in perceptions of danger, fear and preference in a simulated natural environment. Journal of Environmental Psychology, 30(4), 473 - 481.

14. Appleton, J. (1975). The Experience of Landscape. London: Wiley.

國際替代計量

恐懼的性質：從視覺注意力到眺匿理論

主題瀏覽