聽力聯合模式：說明信念與策略對聽力理解能力影響之實證模式

The Conjugated Model of Listening: An Empirical Model to Explain the Impact of Beliefs and Strategies on Listening Comprehension

倪強(John-Michael Lopez Nix)

指導教授 ： 曾文鐽

國立臺灣師範大學/文學院/英語學系/博士(2015年)

https://doi.org/10.6345/NTNU202205339

摘要

現有研究的基本假設是學習者信念會影響語言學習策略運用的範圍，因此亦間接影響了第二語言 (second language, L2) 學習的結果。赫洛維茲（Horowitz）是這個理論傑出的支持者，且發展了語言學習量表的學說（BALLI: Horowitz, 1985; 1988）。過去數十年變項分析研究已經證實語言學習信念和學習行為之間的聯結，實務上，這些發現也經常被應用於第二語言的學習策略。到目前為止，尚無研究曾經仔細檢視L2聽力學習信念和聽力策略以及聽力熟練度之間的關係。假定語言學習信念會影響L2學生的學習行為，那麼評量L2聽力信念和審視它們和其他獨立變項間的交互影響是非常必要的。本研究採用赫洛維茲的概念模式，將之應用在EFL聽力的特定領域。透過結構方程模式（SEM）、試題反應理論（IRT）以及多向度試題反應理論（MIRT）等方法，本研究提出一個實徵上可驗證的EFL聽力學習的模式以瞭解影響EFL聽力的潛在特質變項與觀察變項間的交互 作用。
研究結果支持赫洛維茲的語言學習信念概念直接影響學生的學習行為，也因此被當作學習策略，並且對聽力理解能力具間接效果。因此，以信念為基礎的語言學習模式當它被應用在EFL聽力學習的特定領域時有了實證的依據。然而，研究結果指出幾個特殊條件必須要運用在這些通論上。所以，本研究亦具體討論並確認了上述所發現之多變項關係。
本研究以經由實證之聽力聯合模式（CML）中認知和非認知各項構念之間的關聯來詮釋臺灣大學生們的EFL聽力學習結果。研究者亦施行多次的前導研究以發展聽力學習信念（BELLA）、聽力策略理解（ELLSI）、運用聽力策略的能力（ISCEL）和被洛斯特（Rost）分類為「詮釋」的更高階、實務性類型的聽力能力等的衡量工具。衡量模式在SEM和CML內的特性與驗證過程之細節在本研究中亦被詳述及討論。
最後，本研究在方法學上、理論層次以及教學應用上有更廣泛的意涵。這些可能引發的結果將按照現有的研究結果加以討論。

關鍵字

聽力信念 ； 聽力策略 ； 聽力理解 ； 試題反應理論 ； 結構方程模式

並列摘要

The fundamental premise underlying the present study is that learner beliefs affect the range of language learning strategies employed, thereby indirectly influencing L2 learning outcomes. The preeminent proponent of this theory is Horwitz, who developed the beliefs about language learning inventory (BALLI: Horwitz, 1985; 1988). Decades of variable analytic research have corroborated the links between language learning beliefs and learning behavior, often operationalized as L2 strategies. To date, no research has specifically examined L2 listening learning beliefs in relation to listening strategies and listening proficiency. Given that language learning beliefs can influence the learning behaviors of L2 students, the imperative to inventory L2 listening beliefs and examine their interaction with other individual difference (ID) variables exists. The present study employs a Horwitzian conceptual model as applied to the specific domain of EFL listening. Via Structural Equation Modeling (SEM), Item Response Theory (IRT), and Multi-dimensional Item Response Theory (MIRT) psychometric measurements, an empirically testable model of EFL listening learning was specified and identified in order to account for latent and observable trait interactions which influence EFL listening.
Results support the Horwitzian notion of language learning beliefs acting directly upon learning behaviors, operationalized herein as strategies, and an indirect effect on listening ability. Therefore, the belief-based model of language learning has support when applied to the specific domain of EFL listening learning. However, the findings indicate that several qualifications must be applied to these general statements. The specific characteristics of the multivariate relations are identified and discussed.
The model identified and validated herein is designated as the Conjugated Model of Listening (CML) to denote the conjunction of cognitive and non-cognitive constructs to explain EFL listening learning outcomes for Taiwanese university undergraduate learners. Multiple pilot studies were used to create measurement instruments for listening learning beliefs (BELLA), listening strategic knowledge (ELLSI), listening strategic competence (ISCEL), and listening ability, i.e., higher-order, pragmatic-type listening abilities, classified by Rost (2005) as “interpretation”. The details of measurement model specification, identification and validation within the SEM, CML, are detailed and discussed.
Finally, this study has far-ranging implications at the methodological, theoretical and pedagogical levels. These implications are discussed in light of the current findings.

並列關鍵字

Listening Beliefs ； Listening Strategies ； Listening Comprehension ； Item Response Theory ； SEM

參考文獻

Andrich, D. (1978). A rating formulation for ordered response categories. Psychometrika, 43, 561-573.

Bacon, S. M. (1992). The relationship between gender, comprehension, processing strategies, and cognitive and affective response in foreign language listening. The Modern Language Journal, 76 (2), 160-178.

Bahrani, T. (2011). How EFL teachers can use TV news to boost listening comprehension and speaking proficiency in their classes. Studies in Literature and Language, 2 (3), 51-55.

Bandalos, D. L. (2002). The effect of item parceling on goodness-of-fit and parameter estimate bias in structural equation modeling. Structural Equation Modeling, 9 (1), 78-102.

Bandura, A. (1986). Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall.