序

發表人

指導

老師

研究題目

研究摘要

01

林成璟

陳伊慈

老師

能力信念與學業自我價值後效對 EMI 課程修習之影響:學業焦慮的角色 Ability Beliefs and Contingencies of Self-Worth in EMI Course Enrollment: The Role of Academic Anxiety

本研究探討大學生能力信念、學業自我價值後效與學習焦慮對修習全英語授課（EMI）意願之影響。針對103位大學生進行問卷調查與量化分析。

研究發現：一、學習焦慮是阻礙修課的核心因素，未修習者焦慮感顯著較高；二、能力信念與焦慮及選課行為無顯著關聯；三、學業自我價值後效與焦慮呈極顯著正相關；四、在預測模型中，僅「學習焦慮」對選修EMI具顯著負向預測力。

結論指出，阻礙學生參與EMI的關鍵並非能力心態，而是過高「學業自我價值後效」引發的強烈焦慮。建議教育者未來應著重於學生的情緒調節與焦慮減緩，以提升其修課意願。

This study explores the impact of university students' ability beliefs, academic contingencies of self-worth, and learning anxiety on their willingness to enroll in English-Medium Instruction (EMI) courses. A questionnaire survey and quantitative analysis were conducted with a sample of 103 university students.

The research findings indicate: (1) Learning anxiety is the core factor hindering course enrollment, with students who have not taken EMI courses showing significantly higher levels of anxiety. (2) Ability beliefs have no significant correlation with either anxiety or course enrollment behavior. (3) Academic contingencies of self-worth show a highly significant positive correlation with learning anxiety. (4) In the predictive model, only "learning anxiety" demonstrates a significant negative predictive power regarding the willingness to select EMI courses.

The conclusion suggests that the primary barrier to students' participation in EMI courses is not their mindset regarding their abilities, but rather the intense anxiety triggered by excessively high "academic contingencies of self-worth." It is recommended that educators in the future focus on students' emotion regulation and anxiety reduction to effectively enhance their willingness to enroll in EMI courses.

02

湯竣宇

李昆樺

老師

大學生Instagram使用方式與感知社會支持關聯之研究

本研究旨在探討大學生 Instagram使用、同儕社會支持現況，以及使用方式與感知社會支持之關聯。其研究目的有二：（一）了解大學生Instagram的使用方式與使用現況；（二）探討不同Instagram使用方式，與其感知社會支持之關係。
本研究採量化橫斷式研究設計，以臺灣地區大學一至四年級在學大學生為研究對象，透過網路問卷蒐集資料。研究工具包含「個人背景資料調查表」、「台灣版 Instagram使用模式量表」及「社會支持量表」。所得資料經由JASP統計軟體，以描述性統計、變異數分析、皮爾森積差相關及多元線性迴歸進行分析。
根據研究結果，本研究獲得的發現與結論如下：
一、大學生Instagram互動使用與感知社會支持顯著正相關。
二、大學生Instagram被動瀏覽與感知社會支持顯著負相關。
三、大學生Instagram主動發佈與感知社會支持無相關。
四、大學生Instagram隱私設定、互動使用對感知社會支持具有預測力。
五、不同性別、不同Instagram使用量在感知社會支持上無差異。

03

陽繼安

李昆樺

老師

從衝動與強迫特性試探討電子遊戲玩家的享樂不能與低動機感

在台灣的網路論壇或遊戲社群上，常出現一個玩家自嘲的詞「電子陽痿」，指稱明明買了一堆遊戲或花許多精力培養遊戲角色、資源，卻不再有心力開遊戲玩或登入帳號。本研究將探討電玩玩家在遊戲中失去愉悅感的現象，以先前在強迫症及成癮框架下進行的跨診斷研究為基礎，探討遊戲內的享樂不能是否與衝動及強迫特質有顯著相關。研究以線上問卷進行，並以便利抽樣招募參與者。數據分析採階層迴歸，檢驗在第一層控制壓力、心理困擾等變項後，加入衝動性或強迫性是否顯著提升模型解釋力。
　　研究結果大多為未達顯著水準，或反過來顯著預測愉悅感的提升。以上研究結果指出，進行跨診斷研究時可能面臨的建構定義不一致、一般樣本與臨床樣本間的顯著差異等問題。且基於達顯著的變項關係，本研究亦對既有理論框架，以生理機制、認知與情緒歷程進行更深入的描摹。總體來說，本研究以實徵數據指出，欲進行此類研究需注重的，可能混淆結果的重要因子，並對衝動性、強迫性及享樂不能的跨診斷特徵，彼此間的關聯與區別。

04

李艾珈

李昆樺

老師

童年逆境經驗對大學生社交焦慮之影響：正念特質與反芻的中介作用

本研究旨在探討大學生童年逆境經驗（ACEs）與社交焦慮之關聯，並檢驗正念特質與反芻行為在兩者關係中的中介角色。本研究採橫斷式設計，共回收361份有效問卷。相關分析顯示，ACEs與社交焦慮、反芻呈顯著正相關，與正念特質呈顯著負相關。中介分析指出，正念特質與反芻行為均具有顯著間接效果，且正念特質之間接效果量顯著高於反芻行為；當兩項中介變項同時納入模型後，ACEs對社交焦慮之直接效果未達顯著。路徑分析結果顯示，ACEs、正念特質與反芻可共同解釋社交焦慮48%之變異量（R² = .48）。

05

何若彤

趙軒甫/
張硯評

老師

Nostalgia for Whom? How Nostalgic Targets Shape Status Consumption Through Relational and Self-Concept Pathways

While recalling the moments we spent with our partners, we do not simply retain these shared experiences as episodic memories. Instead, they may reshape our self-evaluations and further affect our behavior, making status consumption one possible outcome. However, the psychological mechanisms linking nostalgia to status consumption remain unclear, especially when different nostalgic targets are involved. This research, therefore, asks whether different nostalgic targets—specifically the self, one’s current partner, and one’s former partner—change the way one makes consumption decisions. The findings show that, compared with the other conditions, when the nostalgic target was one’s current partner, perceived relational status more strongly predicted status consumption tendencies. When the nostalgic target was one’s former partner or the self, lower self-concept clarity and lower self-esteem were more likely to be associated with status consumption tendencies. To apply these findings to a concrete consumption context, this research further examined hotel preferences. To integrate multiple indicators of hotel preferences, exploratory factor analysis was used for dimension reduction, identifying four dimensions: accommodation quality, humanistic and psychological experience, expectation gap management, and functional practicality. The correlational results indicated that status consumption was associated with accommodation quality, humanistic and psychological experience, and expectation gap management, but not with functional practicality. This research contributes to nostalgic consumption literature by showing that the target of nostalgia shapes distinct psychological routes to status consumption. It also extends this perspective to hospitality, suggesting that status consumption is more strongly linked to experiential and evaluative concerns than to basic functional needs.

06

林涵真

詹雨臻

老師

The Use of Machine Learning and Deep Learning Algorithm on Visual Imagery fMRI Data: A Systematic Review

Machine learning (ML) and deep learning (DL) algorithms in analyzing fMRI data have advanced throughout the years, providing insightful understandings of collected data beyond traditional statistics approaches. Previous systematic reviews have offered meaningful outcomes on the analysis of mental disorder fMRI data with ML/DL algorithms. Alternatively, exploring non-clinical themes enhances our understanding of these techniques on a wider scale. The aim of this systematic review is to identify the most commonly used ML/DL algorithms, provide an overview of ML/DL computational framework, highlight how these algorithms were used to reveal the shared neural representation in early visual cortex for perception and imagery in fMRI visual imagery studies. Articles obtained from Scopus, Web of Science, and PsycINFO were screened and reviewed, resulting in a final sample of 17 articles included in the systematic review. The result suggested that the machine learning algorithm is widely adopted across the included studies, with support vector machines being the most common, while the deep learning algorithm is rarely applied. Furthermore, categorical classifiers (e.g. SVM, linear discriminant analysis) effectively decode predefined categorical content of visual imagery, whereas decoding and encoding models reconstruct features or predict neural patterns, and domain adaptation servers as a potential solution to distribution shift between imagery and perception data.

07

陳冠恩

許育光

老師

降低成年早期升學與就業焦慮團體方案設計之行動研究

本研究採行動研究法，探討「接納與承諾療法」（ACT）對處於 Super 生涯「探索期」之大四學生心理彈性與生涯決策之影響。研究發現，受試者在面臨研究所升學與就業轉折時，普遍存在經驗逃避與生涯焦慮。透過 ACT 之認知解離、接納與價值觀澄清等技術介入，能有效協助成員將焦慮視為內在經驗，而非行動阻礙。研究結果顯示，成員在參與團體後，能提升對生涯不確定感的容忍度，並在價值導向引導下，轉化焦慮為具體的承諾行動。本研究實證了 ACT 結合生涯發展理論之輔導方案，能有效支持個體完成探索期之發展任務。

08

09

藍寬驊

許育光老師

（等待更新中）