本研究是由動力性觀點，探討大學生的一般生活事件、重大事件、連續事件三者與生活壓力、身心健康二者之間的關係。目的在於了解各個人面對生活變動的狀況，以及面對生活變動時，其心理與生理的反應，因應生活變動的方式，及生活變動經驗累積的情形。研究對象主要是臺大二年級學生73人，以身心症狀指標將之區分成：生理組（生理症狀指標高）、生心組（生理及心理症狀指標均高）、心理組（心理症狀指標高）及健康組等四個組別，再以訪問方式進行資料收集。結果顯示：1.在「生活事件總數」、「不受歡迎的事件數」、「非獨立事件」以及「影響廣度」等四個變項，四組學生有顯著差異：生心組與心理組二者均高於健康組，而生理組與健康組較為接近。在「影響持續度」與對事件的「喜歡程度」二個變項，則無差異。2.就個人的「自我功效預期」而言，生心組及心理組在「預期自己的應付能力」方面低於生理組及健康組；而在「發生的可控制度」方面四組學生無顯著差異。3.對事件的因應行為，只有「尋求協助」此一變項，四組學生有顯著差異：生心組及心理組高於健康組。四組學生的「解決」或「逃避」行為傾向並無差異。對於結果的歸因，亦未見差異。4.重大事件對個人造成的影響主要為人際態度與人生觀方面；並以心理組受到重大事件負向影響的比率較高。5.生心組在相似事件之連續經驗方面，其實際的應付能力與預估再次經驗的應付能力無顯著變化。經歷兩次相似的負向事件經驗後，生心組與心理組傾向認為若該事件再次發生，個人的情緒將更惡化。6.解釋「症狀困擾指標」的變項，以「預期應付能力」、「非獨立事件」、「發生的可控制度」、「影響廣度」及「歸因個數」等五者爲最適當之解釋組型，其解釋變異量達47%。

The purpose of this research is to use a dynamic P-E fit model of life stress to understand the perception of life events, such as individual's psychological and physiological reactions, coping behaviors and cumulative effects of life events. Seventy three college students are divided into four groups: Group A: highest in the physiological symptom index and lowest in psychological symptom index among 2965 students; Group B: both indices are the highest; Group C: lowest in physiolosical symptom index and highest in psychological symptom index; Control group: both indices are the lowest. The results show that group differences in numbers of life event, undesirable event, dependent event and influenced extent of life event are significant. Group B and Group C scored higher than the control group on all four measures; Group A do not differ from the control group. For self-ability expectation, Group A and Group C are scored lower than Group A and the control group. For help seeking behavior, Group B and Group C are higher in frequency than the control group. The most influential life events are interpersonal attitude and life philosophy. Group C perceive more negative impact of a life event. After experiencing two or more similar negative events, Group B and Group C tend to feel more negatively than the other two groups when the event occures again. Five variables, selected from 26 variables in stepwise regression analysis, are identified to be most effective predictors of PSDI (account for 47% of the total variance): self-ability expectation, dependent events, controllability, influenced extent and number of attribution.