運用層級貝氏方法建構台灣兒童遊具產品安全評估模式－滑梯產品爲例

「遊戲」是兒童生活中重要的成長與學習，是兒童生活的重心。而「遊戲事故傷害」又是我國一歲以上兒童死亡的的首要原因，也是兒童急診花費最多的健康問題。近年來，由於兒童對遊具之使用方式不當、兒童本身的疏忽或遊具設計維護之不良，造成很多兒童在遊戲過程中受傷的案例。平均而言，幾乎每天都有小朋友死於事故傷害，而每兩分鐘就有一名小朋友因事故傷害而去看急診。而為了要避免校園運動或遊戲傷害，需特別注意校園內運動空間及遊戲設備的安全管理。根據研究調查，遊樂設施的設計和使用將會影響傷害的發生。因此本研究選擇滑梯常見的遊具做為研究標的物，並以國內各縣市之7-12歲國民小學學童作為研究對象，期望建立出一個遊具設計因子和兒童受傷程度（例如輕傷、重傷等）之間的模式。而在模式建構的過程中，主要是以貝氏層級法做為資料分析的方法。本研究將建立出一個模型以描述產品設計因子和傷害之間的因果關係，再利用一個更一般化的資料分析技術－層級貝氏模式(Hierarchical Bayesian Model)來推估出模型中我們欲求的參數。藉此以找出影響兒童傷害的關鍵產品設計因子，提供玩具業者將來設計生產遊具，或作為加強目前已存遊樂設施之參考依據，希望提高兒童在使用遊樂設施之安全性。本研究結果建議遊具廠商在設計滑梯時，要考量對不同年齡層兒童其適合的平台高度，以及「滑出段高度」與「體重」間的安全測試，並在滑梯生產完成後，特別標明適合遊玩的兒童年齡範圍與體重範圍，以減少因體重或年齡不符安全範圍的兒童遊玩滑梯而發生的意外傷害。

關鍵字

層級貝氏模型；安全評估；遊具設計

並列摘要

”Playing Games,” a main activity of children's life, is important for their growing and learning. However, ”injury during play” mainly not only causes deaths for children over one year old, but also produces health issues that cost the highest expenses in our country. In recent years, children's negligence and misuse of children's commodities, many with defects in the designs, have caused many cases of injury during play. On average, each two minutes a child is sent to the emergency room due to injuries. Children die from injuries almost every day. Nevertheless, to avoid injuries from sports or play at school, we should pay more attention to the activity space, playground facilities and safety management. According to research, the design and usage of playground facilities could affect occurrences of injuries. Therefore, wishing to construct a model: the level of children injuries (i.e., flesh wound, serious wound) relating to designs of children's commodities, the most used playground facilities: slide, as target, and elementary students from each city in Taiwan between seven to twelve years old were selected as research objects. During the constructing period, Hierarchical Bayesian Approach is the main reference when analyzing data. This research will construct a model that describes the relationships between the designs of children commodities and the level of children's injuries. Then, by using a simplified data analyzing technique-Hierarchical Bayesian Model, we calculate the parameters from the research model. After finding the main factors of product designs that affects children injuries, and hoping to insure the safety of children when using these facilities, the model will be provided as a reference to the toy industry to design or enhance current or future children's commodities.