2×2成就目標之因素結構與預測效用研究

Abstract

就學習動機的階層而言，成就目標是較具體的動機變項，也是與學習行為最接近、且最直接的影響變項。過去的研究多以西方學生在數學、自然或歐美語言等領域的成就目標對學習的影響為研究主題，較少的研究關注我國學生學習國語文的成就動機。 本研究以台灣國小到高中的學生為樣本，探討2×2成就目標(Elliot & McGregor, 2001)之因素結構、跨一年的測量穩定度及成就目標在國語文學習歷程的預測效用。首先，以2007年的橫斷資料進行驗證性因素分析，檢驗一階四因素結構適配度是否優於各種選替模式(二、三向度)，結果支持2×2成就目標模式的四個因素 (趨向精熟、趨向表現、逃避精熟與逃避表現)是個別存在的。橫斷資料的二階因素結構適配度也優於各種選替模式，證實2×2成就目標符合理論的兩軸因素結構：第一軸為兩個對能力的定義因素(精熟與表現)，第二軸為兩個對能力的定價因素(趨向與逃避)，且此二軸彼此相交，2x2理論的測量模式受到支持。總之，本研究修訂的成就目標量表(AGQ-C)在我國國小到高中學生群中具有良好的信度與效度，亦即2x2成就目標模式能描述我國學生的學習目標。此外，研究結果顯示:四個二階因素個別指向其對應的兩個一階因素的徑路係數值大小相差很多，經由路徑係數的恆等分析檢驗，發現每一對的一階目標對於相對應的二階目標因素有不相同的效果，亦即每一個二階因素(如：表現)受到特定的一階成就目標(趨向表現)所反映。進一步檢驗2x2成就目標在2007、2008兩個時段的測量穩定度，指標檢驗指標包括結構穩定度、差異連續性、平均數穩定度。測量穩定度檢驗結果顯示，兩個時段測量模式的形貌恆等，而模式弱恆等(因素負荷量恆等)、強恆等(截距恆等)模式也都成立，顯示我國學生的成就目標間隔一年後雖然有所變化，但具有合理的連續性與穩定度，而且兩個時段的目標均達正相關、只有趨向表現目標平均數有顯著增加，另三個目標則無顯著差異。預測效用的檢驗以2007年的橫斷資料進行共變結構分析，檢驗成就目標是否能中介國語文自我效能及成績，本研究選擇自我效能，乃因其為社會認知論中重要且穩固的學習動機指標，長久的國語文學習使自我效能內化為近似於內在穩定的特質，而成就目標與外在情境的連結較為密切。結果顯示：趨向精熟、趨向表現、逃避表現三個目標可以中介國文自我效能與國文成績，呼應成就目標理論:成就目標是中介抽象的學習動機與學習行為間的心理建構。同時，成就目標的定價面向(趨向和逃避因素)之中介效果優於成就目標的定義面向(精熟和表現因素)，比起定義面向，成就目標的定價面向:趨向與逃避因素較能反映台灣學生的成就目標；趨向型的目標是預測台灣學生國文成績的最有效因子。本研究發現台灣國小到高中學生的實證資料支持2×2成就目標模式可以描述國語文學習目標，具有跨年的測量穩定性，成就目標對國文自我效能與國文成績具有中介效用。本研究提出未來研究的建議及教室教學的實務建議。

In terms of hierarchical models of motivation, achievement goals are conceptualized as concrete representations of more abstract motivation dispositions and as the proximal, direct regulators of learning behaviors. The major of research have studied achievement goal endorsement in math, science and language subjects among western students. Unfortunately, little research has been published on Taiwanese students’ adoption of achievement goals in their Chinese language class. The aims of the dissertation were to investigate the measurement structure, cross-year goal-pursuit stability and predictive utility of the 2 × 2 achievement goal model (Elliot & McGregor, 2001) in Taiwanese pre-university students while learning Chinese. Factorial/dimensional structures and internal consistencies provided support for the 2 × 2 achievement goal framework. The 2 × 2 achievement goal structure of the achievement goal items was confirmed, and the four-factor goal structure was found to be a better fit to the data than a series of alternative models with dichotomous/trichotomous goal structures. The results further offered evidence for the two-dimensional structure posited by the 2 (definition) × 2 (valence) achievement goal model. Path coefficient invariance of the dimensional model indicated that each pairs of goal has nonequivalent contribution to correspondent achievement goal dimensional factor: each dimensional factor was mainly derived from different achievement goals. No significant decreases in model-fits (compared to the weak invariant model) when constraints were added to various invariant models. Three stability indexes (structural, differential continuity, and mean-level stability) provided evidence for the stability of achievement goal endorsement over time in a panel sample of Taiwan pre-university students. In terms of predictive utility, three of four achievement goals: mastery-approach, performance-approach, and performance-avoidance were found to be effective mediators between Chinese self-efficacy and Chinese performance. The findings support the rationale that achievement goals are viewed as a direct, proximal influence on achievement-relevant behavior and motives are portrayed as direct antecedents of achievement goals (Elliot & Church 1997). Results of dimensional factor predictive utility, the approach-avoidance factors along the valence dimension seemed to be more successful than the mastery-performance factors along the definition dimension. Approach-based goals were observed for significant predictors of Chinese grades in Taiwanese students. Taken together, my data strongly supports that the 2 × 2 achievement goal framework appears to be empirically as well as conceptually sound for Taiwanese students and have mediating utility on self-efficacy and Chinese grades.