作者與期刊自我引用之研究：以環境工程領域為例

A Study of author self-citation and journal self-citation in environmental engineering

林雯瑤(Wen-Yau Cathy Lin)

指導教授 ： 黃慕萱

國立臺灣大學/文學院/圖書資訊學研究所/博士(2009年)

https://doi.org/10.6342/NTU.2009.00186

摘要

本研究以環工學門二十種重點期刊，於1999至2008年間出版的研究論文為例，採用引用文獻分析與文本分析方法，探討期刊自我引用率、自我引用對影響係數、立即指數、期刊 h 指數之影響，以及作者自我引用與作者數、文獻新穎程度、自我引用的位置、作者 h 指數之關聯等議題。

研究結果顯示，在期刊方面，環工學門重點期刊的次數自我引用率與被自我引用率之間有中度相關，而出版文獻量與自我引用率呈現正相關。被自我引用對影響係數、立即指數的影響皆不顯著，影響係數和立即指數與自我引用率或被自我引用率也均無顯著相關。而影響係數與五年影響係數在包含或扣除被自我引用的條件下，其數值與排名皆為高度相關，且具顯著差異。被自我引用對期刊 h 指數的影響皆不顯著，10年、5年 h指數的數值與自引/被自引率亦皆無顯著相關，但1年、2年 h 指數則在部分年份與自我引用率達中度相關，顯示採計被引用時間區間較短的 h 指數時，應注意自我引用所造成的影響。

在作者方面，具共同作者文章的作者被自我引用率比單一作者文章更高，合著者人數則與作者自我引用率與被自我引用率皆為低度相關，且作者自我引用的文獻確實比引用他人文獻新穎。作者傾向在文章的緒論、結果、方法等區塊引用自己先前發表的作品，在方法、結果、結論三個位置，則是自我引用文獻比引用他人文獻更常見。而無論扣除被自我引用與否，在 h 值或 h 指數的排名方面所造成的影響都很小，作者 h 指數與作者的文章數與總被引次數及其排名之間則均呈現高度相關。

基於上述研究結果，建議環工學門在計算影響係數與期刊 h 指數時，採計影響係數、2年 h 指數即可，且毋須扣除期刊被自我引用，而評量各期刊的自我引用率時應考慮各期刊的文獻出版量殊異。若欲檢視作者自我引用的合理性則可優先從文章的緒論著手，計算作者 h 值時亦毋須扣除作者被自我引用，且採計 h 指數數值或排名皆可。

關鍵字

作者自我引用 ； 期刊自我引用 ； 影響係數 ； 立即指數 ； h 指數

並列摘要

The objective of this study is to identify the author/journal self-citation pattern by citation analysis and context analysis on the environmental engineering key journals which were published in between year 1999 to 2008. For the level of journal comparison, journal self-citation rate, impact factor, immediacy index, and journal h-index were analyzed under the conditions of containing as well as eliminating the self-cited data. On the other hand, the correlation between author self-citation and number of co-authors, age of reference article, location of reference in original articles, and author h-index were examined in the aspect of author difference.

Results suggested that moderate significant correlation existed between the journal frequency self-citing and frequency self-cited rates. Those journals with higher self-citing rate tended to be more productive. The journal self-citation rate had no significant correlation with both impact factor and immediacy index. Under the conditions of containing or eliminating journal self-cited data, a high significant correlation between impact factor and 5-Year impact factor was identified but with a significant difference in the value. Analysis on longer period of h-indices showed similar pattern as on the analysis of impact factor. With shorter citation window of h-indices, self-citation rate and h- values were observed to be moderately correlated in some years.

In the aspect of author difference, the self-cited rate of multi-authored articles was higher than single-authored ones. The correlation coefficient between the number of authors and author self-citation rate was law. Self-citing references were found to be more recent than those citations to others. Authors tended to cite their own previous published works in the introduction, result, and method sections. In the sections of method, result, and conclusion, self-citing references are more common than citing to others. No matter containing the self-cited data or not, insignificant effect was found on the author h-values and h-index. However, the correlation coefficients were strong between the author h-value and both their productivity and the number of total cited times.

Based upon the results, suggestions were made as follows. When evaluating the environment technology related journal publication, using 2-Year impact factor and 2-Year h-index instead of longer period ones would be adequate in the citation window setting. Impact factor, immediacy index, and journal/author h-indices seemed all robust enough against self-citation. Thus, it is not necessary to eliminate the self-cited data when doing environmental technology related journal or author evaluation. The journal productivity should be taken into account when measuring journal self-citation rates. For surveying the self-citing rationality, each article’s introduction section could be a priority target.

並列關鍵字

author self-citation ； journal self-citation ； impact factor ； immediacy index ； h-index

參考文獻

紀佩姍 (2008)。H-index 應用於大學學術評估之探討：以臺灣地區大學為例。未出版之碩士論文， 國立臺灣大學圖書資訊學研究所，台北市。

黃毅志、洪聰敏、黃慕萱、鄭燿男 (2008)。國內教育學門教育學術期刊評比研究 (No. NSC96-2420-H143-001)。 台北市: 行政院國家科學委員會。

McVeigh, M. E. (2004). Journal self-citation in the Journal Citation Reports® - Science Edition (2002): A Citation Study from The Thomson Corporation , The Thomson Corporation.

Testa, J. (2004, January). The Thomson Scientific Journal Selection Process Retrieved October 19, 2006, from http://scientific.thomson.com/free/essays/selectionofmaterial/journalselection/

蔡明月 (2007)。社會科學期刊自我引用數據之分析比較研究 (No. NSC 95-2413-H-004-014-)。 台北市: 行政院國家科學委員會。