中學生「不確定性考試壓力」模式之驗證

無論是國內或是國外，中學生的考試壓力均是一項重要的研究議題，然而，關於中學生的考試壓力，國內鮮少進行相關之研究，以致中學生考試壓力的理論發展及實務工作，均無明顯進展。本研究旨在進行一系列中學生考試壓力之研究，以探究考試壓力之現象，並驗證中學生考試壓力與不確定性之關係，以確立「不確定性壓力模式」。研究一檢驗不同學業成就之下的學生，其考試壓力來源，及其補習與唸書之時間，以檢視中成就學生的考試壓力來源是否多為「上不了公立學校」，以及檢視中成就學生是否付出的時間比其他學生更多。研究一以問卷調查，收集5220名9年級學生之資料，其中問卷包含了宋曜廷等(2013)之「中學生考試壓力量表(Examination Stress Scale, ExamSS)」、考試壓力來源、唸書時間、補習天數等資料。研究一同時取得2012年國民中學基本學力測驗之分數，代表學生之學業成就。研究一結果主要有三，第一，學業成就PR30~79學生的考試壓力高於PR90以上及PR29以下的學生，重複驗證學業成就中段的學生，其考試壓力較兩端學生大。第二，不同學業成就的學生，知覺到不同的考試壓力來源：學業成就後段學生的壓力來源多是「再怎麼努力成績都無法提升」、「聽不懂老師的教學內容」，學業成就中段學生則多是「上不了公立學校」及「無法達到自己的標準」，學業成就前段學生則多是「無法達到自己的標準」及「成績無法保持一定的水準」。第三，中成就的學生，唸書時間與學業成就前端的學生差異不大，但是補習天數卻比前端學生多。整體而言，中成就學生可能是因為付出的時間多卻面臨上不了公立學校的風險，因此考試壓力較大，此結果為Sung、Chao與Tseng(2016)提出的不確定性壓力模式提供證據。研究二旨在探討中學生的個人特性與考試壓力與是否有關，檢視是否有特定一群學生，是屬於高考試壓力的中成就學生，並符合特定的個人特性組型，包含不確定性高、神經性高、外向性低、低自尊，以及外控性高。研究二以1858名9年級學生為樣本，以宋曜廷等(2013)之ExamSS做為考試壓力的測量，並以26題自編二元計分的個人特性之題目做為個人特性的測量，包含不確定性、神經性、外向性、低自尊及外控性，並取得2014年國中教育會考之分數做為學業成就的指標。接著，研究二以潛在類別分析(latent class analysis, LCA)，針對考試壓力、學業成就及個人特性的作答反應將學生進行分類。結果發現，研究二的資料最適合將學生分為四類，這四類學生分別為高壓中成就組(佔22%)、中壓高成就組(佔26%)、低壓中成就組(佔32%)及低壓低成就組(佔20%)。同時，由各組之作答機率側面圖(profile)可以得知，高壓中成就組的學生，相較於中壓高成就組及低壓中成就組，有較高的不確定性、較高的神經性、較低的外向性、低自尊，以及較高的外控傾向，此結果除了與研究假設相符，也為不確定性壓力模式提供進一步的證據。同時，研究二亦發現了低壓低成就組的學生，反而會有較高的外控性及低自尊，值得教育工作者付出關切。研究一與研究二分別從社會氛圍及個人特性的角度，說明學生的不確定性與考試壓力可能有關。研究三旨在再次檢視中學生的不確定性與考試壓力之關係，並直接對不確定性壓力模式進行驗證。在研究三之中，我們收集1967名9年級學生的資料，以宋曜廷等(2013)之ExamSS做為考試壓力的測量，並自編一份中學生不確定性量表做為不確定性的測量。在學業成就方面，我們以2017年國中教育會考之分數做為學生學業成就之指標。經由試題分析及因素分析，研究三得到中學生不確定性量表之正式版本，而此量表信度良好，本研究亦有提供建構效度之證據。接著，研究結果發現，考試壓力與不確定性呈正相關，其整體相關為.60。同時，經由多變量變異數分析，高考試壓力的學生在不確定性各分量表，均高於低考試壓力的學生。最後，我們利用徑路分析，驗證無論在學業成就前半或後半的學生，不確定性均部分中介了學業成就對考試壓力的效果；若以不確定性中的「前景不明」做為中介變項，則在學業成就前半段的學生中，達到完全中介的效果，為不確定性壓力模式提供直接的證據。綜合三項研究結果，本研究將以中學生考試壓力之社會氛圍、個人特性，以及不確定性，提供理論上及實務上之建議，期望實際達到降低學生考試壓力之效果。

關鍵字

中學生；考試壓力；壓力來源；個人特性；不確定性

並列摘要

This study investigates phenomena linked to examination stress among secondary school students. In study 1, we consider environmental factors in examination stress, such as different levels of achievement, sources of stress, and the amount of time devoted to study and cram schooling. The responses of 5,220 9th graders’ were collected using a questionnaire comprising Sung et al.’s (2013) Examination Stress Scale (ExamSS) and questions about sources of stress, time spent studying, and the number of days spent in cram school. In addition, student scores on the Basic Competence Test (BCTEST) were used as indices for their levels of achievement. Study 1 yields three main results: First, students between percentile rank (PR) 30 and PR79 on the BCTEST have higher levels of examination stress than both those PR80 or above and those below PR30. This replicates the phenomenon that students of medium achievement experience the greatest levels of examination stress (Sung, Chao & Tseng, 2016). Second, students of different levels of achievement cite different sources of examination stress. The lowest-achieving students gave the most reasons such as “No matter how hard I try, I don’t make progress” and “I don’t understand what the teacher is teaching”. Meanwhile, middle-achieving students cited “I won’t get into a national [high] school”, and high-achieving students cited “I do not reach my own standards”. Third, the amount of time devoted to study varies little between middle-achieving students and high-achieving students; however, the former spend more days in cram school than the latter. Overall, apart from different levels of examination stress, Taiwanese students may experience varying degrees of stress in relation to their academic achievement. It may be that examination stress is greater among middling students because they invest more time in study but still run the risk of not being admitted to a national school. In study 2, we explore whether examination stress among secondary students is related to their personal characteristics, specifically whether students subject to high levels of examination stress have a certain pattern of characteristics. The sample consists of 1,858 9th graders. ExamSS was used as a measure of examination stress, and a questionnaire compiled by the researcher and comprising 26 dichotomous questions was used to measure personal characteristics, including uncertainty, neuroticism, extroversion, low self-esteem, and locus of control. Latent class analysis was used to categorize students according to the relationship between examination stress and personal characteristics. The results show that the data best fit a four-classes model, classifying students into high-stress- medium-achievement (22%), medium-stress-high-achievement (26%), low-stress- medium-achievement (32%), and low-stress-low-achievement (20%). The profiles of these four classes show that, by comparison with low- and medium-high stress groups, the high-stress group has higher rates of uncertainty and neuroticism, lower extroversion and self-esteem, and greater tendency to externalize the locus of control. By contrast, the low-stress-low-achievement group is shown to have high external locus of control and low self-esteem. In study 3, we investigate the relationship between uncertainty and examination stress among secondary students. Data were collected from 1,967 9th graders using Sung et al.’s (2013) ExamSS as a measure of examination stress, and the Uncertainty Scale for Adolescents as a measure of uncertainty. Scores from the Comprehensive Assessment Program 2017 (CAP 2017), the largest standardized achievement test for junior high school students in Taiwan, were used as an index of academic achievement. The results of item analysis and factor analysis confirm the formal version of the Uncertainty Scale for Adolescents and show that the scale’s internal consistency has good reliability, while confirmatory factor analysis provides evidence of construct validity. A positive correlation is found between examination stress and uncertainty, with an overall correlation of .60. Multivariate analysis also shows that students with high examination stress measure higher on all subscales than students with low stress. Finally, path analysis was used to verify the uncertainty-of-stress model which Sung, et al. (2016) proposed. On the basis of these three sets of results, this study provides theory- and practice-based recommendations about examination stress among secondary school students in relation to environmental factors, personal characteristics, and uncertainty, hoping that educational policy can be used to change social atmosphere and provide counselling in schools where it is needed so as to effectively reduce examination stress among adolescents.