以書目計量學方法探討行為經濟學跨領域程度之研究

Abstract

行為經濟學是經濟學發展歷史上的一次重大突破，早期是心理學結合行為經濟學的跨學科研究，從起源至今已發展超過半個世紀，行為經濟學研究的範疇與關注的主題一直以來不斷的在變動，但針對行為經濟學跨領域程度的研究卻非常少，故對於行為經濟學進行跨領域程度的研究有其必要性。 本研究收錄Web of Science 1967年至2018年的書目紀錄，以多樣性和凝聚性兩個方面著手，研究行為經濟學跨領域的程度，藉由關鍵字、領域分類的數量分布了解行為經濟學領域主題的組成與多樣性，以及共字、共著、書目耦合等書目計量方法繪製網絡圖，計算網絡凝聚指標，觀察行為經濟學書目網絡凝聚情形。 研究結果顯示：（一）行為經濟學以商學和心理學為主，逐漸包含更加多樣的領域，領域分類數量持續增加；（二）行為經濟學領域的跨領域多樣性逐漸提高，涵蓋的領域數量增加，領域分布平均程度也增加；（三）在共著網絡中可以看到行為經濟學者在近幾年凝聚成一個主要的大群體，顯示領域在近年應有整合的趨勢。另外，本文根據不同網絡採取不同的篩選方式，並提出新的指標代表性（representative）作為繪圖時網絡節點文字大小變化的依據，凸顯群體重要的節點，在方法上做出創新的嘗試。

Behavioral economics is a major breakthrough in the history of economic development. It is a interdisciplinary study of psychology combined with behavioral economics. It has been developed for more than half a century since its origin, and the scope and focus of behavioral economics research are always changing. There are very few studies on interdisciplinary of behavioral economics, so it is necessary to conduct interdisciplinary research on behavioral economics. This study includes the bibliographic records of Web of Science from 1967 to 2018. It deals with two aspects of diversity and cohesion, and studies the degree of interdisciplinary of behavioral economics. The composition and diversity of behavioral economics is understood by the quantitative distribution of keywords and domain categories. The bibliographic measurement methods such as co-words, co-authoring, and bibliographic coupling are using in realized cohesive of the filed. Drawing network graphs and calculate network indicators to observe the behavioral economics bibliography network cohesion. The results show that: (1) Behavioral economics is dominated by business and psychology, and gradually includes more diverse fields, and the number of domain classifications continues to increase; (2) The diversity in the field of behavioral economics is gradually increasing. The number of covering the fields has increased, and the average degree of domain distribution has also increased. (3) In the co-authored network, it can be seen that behavioral economists have condensed into a major large group in recent years, indicating that the field should have an integration trend in recent years. In addition, this paper adopts different filtering methods according to different networks, and proposes a new Indicator calls 'representative' as the basis for the difference of the text size of the network nodes when drawing the graph, highlighting the important nodes of the group, and making innovative attempts in the method.