適用於台灣教育相關領域研究生之統計焦慮與統計態度、作業延宕量表開發

Abstract

本研究旨在探討並建立適用於台灣教育相關領域研究生之統計焦慮、統計態度與作業延宕量表。經過翻譯、用語修正與加入適用本土文化之題目後形成新修正後量表。以139教育相關研究所學生為施測樣本，將統計焦慮量表(The Statistics Anxiety Rating Scale, STARS)、統計態度調查量表(The Survey of Attitudes Toward Statistics, SATS) 、學業延宕量表(The Procrastination Assessment Scale - Students, PASS)三份量表的100學年初樣本進行探索性因素分析(Exploratory Factor Analysis, EFA)。結果顯示STARS問卷可歸類為8個因素，分別命名為「統計上的自我價值」，「解釋統計資料的焦慮」，「對統計的價值感」，「應用統計資料的焦慮」，「考試和課堂焦慮」，「用英文解釋統計的焦慮」，「害怕統計老師」以及「害怕尋求協助」；SATS問卷可歸類為6個因素，分別命名為「對統計的負面情感」，「統計對我的負面價值」，「統計對我的正面價值」，「努力」，「困難」，「自我效能」；PASS問卷可歸類為5個因素，分別命名為「害怕失敗」，「負向自我價值」，「完美主義」，「厭惡作業」，「延宕行為」。進一步加入100學年末樣本進行驗證性因素分析(Confirmatory Factor Analysis, CFA)之外在效度檢驗，發現各量表結構模式之各項適配度指標皆落在理想範圍內，內部一致性信度皆達.85-.90。而在不同統計程度及不同時間點之樣本施測下，皆發現量表滿足量測恆等性之檢定。 研究結果顯示，對於教育相關研究生而言，此三份量表具有跨樣本的有效性及穩定性。不論是參與初階或是進階課程之研究生來說，統計焦慮為一普遍且穩定存在的情境焦慮特質。

The purpose of this study was to adapt and build the Statistical Anxiety Rating Scale, the Statistical Attitude Scale and the Procrastination scale for Taiwan graduate students in educational filed. After translated the original scales to Mandarin and added the items to meet the special needs of local students, these three revised scales were used to collect responses from 139 graduate students in Education field of a National University. Based on the exploratory factor analysis (EFA) results of 2011 school year students in the beginning of semester, eight factors in the STARS including “Statistical self-concept,” “interpretation anxiety,” “task value of statistics,” “data analysis anxiety,” “test and class anxiety,” “English interpretation anxiety” “fear of statistics,” “fear of asking for help,” were extracted; six factors in the SATS including “negative affect” “negative value” “positive value,” “effort,” “difficulty,” “self-efficacy” were extracted; five factors in the PASS including “fear of failure” “negative of self-value” “perfectionism” “aversiveness of tasks” “procrastination” were extracted. Including the data in the end of semester, the goodness of fit indices of confirmatory factor analysis (CFA) all met the commonly-used criteria. The internal consistency coefficients ranged from .85 to .90. With the inclusion of extra 101 school year data, the measurement model of STARS, SATS, PASS could be seen as statistically equivalent across time points and different class levels. The results indicated that revised scales were useful and valid tools to access students’ anxiety, attitude and procrastination when learning statistics.