請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/3987
標題: | 學習策略與工作記憶廣度對快速型態辨識學習的影響:以圍棋棋型辨識為例 The Influence of Learning Strategy and Working Memory Capacity on Fast Pattern-Recognition Problems: Using Go Game as an Example |
作者: | Ching-Chia Lin 林京佳 |
指導教授: | 連韻文(Yunn-Wen Lien) |
關鍵字: | 棋型辨識,自我解釋,問題導向策略,工作記憶廣度,圍棋,認知負荷, pattern recognition,self-explanation,problem-oriented strategy,working memory capacity,game of Go,cognitive load, |
出版年 : | 2016 |
學位: | 碩士 |
摘要: | 不同於分析與搜索解題路徑,快速型態辨識能力被認為是偏重直觀的,也是專家解決問題與決策時優勢所在。過去發現問題導向型策略(例如自我解釋策略)可促進問題解決的表現並有遷移效果,但是該策略是否可提升快速型態辨識能力則尚待檢視。本研究以圍棋棋譜佈局棋型的型態辨識(以下簡稱棋型辨識)為例,探討問題導向程度不同的學習策略對初學者解決棋型辨識問題的效果。由於問題導向型策略需要較多的認知資源投入,本論文從認知能力的個別差異(個體的工作記憶廣度)與學習情境認知負荷來檢視其成效。實驗一與二隨機分派59名和62名圍棋初學者(18到29歲)至三組,分別使用自我解釋(問題導向程度高)、預測(問題導向程度中等)和範例觀察(問題導向程度低)三種策略進行棋譜學習四十分鐘,學習階段前後各進行一次棋型辨識測驗。棋譜學習指擺放專家棋士對局棋步的圍棋學習法,兩個實驗分別採用兩種常用的棋譜學習情境:實驗一為電腦擺譜情境(低認知負荷),實驗二為傳統擺譜情境(高認知負荷)。結果發現在電腦擺譜情境下,自我解釋策略較觀察策略更有助於棋型辨識的遷移效果。此外,個體工作記憶廣度只對預測組學習遷移有所影響。而在傳統擺譜的情境下,各組棋型辨識的學習遷移都有顯著進步,且無差異,也不受工作記憶廣度影響。本論文首次發現自我解釋策略有助於快速型態辨認的學習,且不受個體工作記憶廣度與學習情境認知負荷的影響。同時也顯示低工作記憶廣度者棋型辨識的學習受到學習策略和情境設計的影響較高工作記憶廣度者大。本論文結果也支持認知負荷理論(Sweller, 2010)對問題導向型策略可引導認知資源分配的說法,但亦有所修正。 Experts are known to be better at fast pattern-recognition in problem solving and decision making than the novices. This intuitive ability is the results of massive practice of a skill. However, the differences in learning strategies that facilitate a beginner in establishing this ability have never been examined. This thesis is using pattern recognition of Go game in a playbook as an example to investigate this issue. In particular, whether problem-oriented strategies, such as self-explanation and prediction, are better than repeated observation during a short-term practice in facilitating learning transfer of fast pattern-recognition of Go was focused. Traditionally, when one practices of the Go game through a playbook, learners examine each piece placed by two professional players step by step. They can 1) search for a right piece on a playbook and repeat the move on the Go board (the observation strategy); 2) predict the next steps before search for the answer (the prediction strategy); or 3) give an explanation for that moves after prediction (self-explanation strategy). Prior studies have found that the last two problem-oriented strategies can improve participants’ performance on analytic problems. However, whether learners using these strategies can outperform those with repeated observation on fast pattern-recognition ability is still unclear. In addition, since problem-oriented strategies are relatively cognitive-demanding than simply observation, participants’ working memory spans were also measured to clarify whether cognitive resources mediate the efficacy of strategy. In experiment 1, 59 young grown-ups who had never learned how to play a 19x19 Go were randomly assigned to three strategy groups in a relatively low-load computer-aided circumstance, in which a computer program automatically showed each move of the game on the screen just by a click. Participants were taught to practice a playbook of Go for forty minutes with each of the assigned strategy. Before and after playbook learning stage, they did a rapid go-pattern test. In that test, participants had to choose a best move from five options for each items in the test within 8 seconds. Half of the items are from the learned playbook (the memory items), and half from an unknown playbook (the transfer items). In Experiment 2, 62 young grown-ups went through the same procedures and tasks in a high-load traditional playbook practice circumstance, in which they had to search for each move from a printed playbook. It was found that only self-explanation group showed significant transfer effect on Go pattern recognition task and outperformed the observation group in computer-aided (low cognitive load) situation. In addition, the transfer effect of prediction strategy depends on participants’ working memory capacities. In traditional playbook circumstance (high cognitive load), all three groups showed significant transfer effects regardless of learning strategies and working memory spans. This research, for the first time, shows that using self-explanation strategy is beneficial for beginners’ learning of Go pattern-recognition, and its effect is robust regardless with learners’ mental capacities and the cognitive load of learning circumstance. In addition, the results also revealed that participants with relatively low mental capacities were more likely to be influenced by learning strategies and circumstances on learning Go pattern recognition than those with high mental capacities. The results also in line with Sweller’s cognitive load theory (2010) with a new perspective of individual differences in cognitive ability. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/3987 |
全文授權: | 同意授權(全球公開) |
顯示於系所單位: | 心理學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-105-1.pdf | 746.95 kB | Adobe PDF | 檢視/開啟 |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。