大學生壓力潛在剖面分析並以憂鬱及問題網路使用為同時與預測效度驗證

Abstract

研究背景與目的 研究顯示大學生普遍地感到壓力，甚至進而導致憂鬱與問題網路使用。因此大學生的壓力為值得關注的議題。透過文獻回顧，整體來說大學生主要的壓力來源包含學業壓力、經濟壓力、親密關係壓力、同儕壓力以及與父母關係壓力。過去的研究多為描述單一壓力與其它變項之間的關係，然而現實生活中鮮少存在單一壓力，因此較合理的做法為同時研究多面向的壓力與變項之間的關係。除此之外，壓力的相關研究較少使用以人為中心的統計方法進行分析，也就是依不同的壓力剖面將人分群。基於上述原因，本研究針對學業壓力、經濟壓力、親密關係壓力、同儕壓力以及與父母關係壓力當作分類變項，將大學一年級的學生依所知覺到的壓力進行分群。另一方面，感到壓力可能會為個體帶來負面影響，例如有證據顯示壓力為憂鬱的前導因子、網路成癮為逃避壓力後的結果。此外，在不同的壓力剖面中，學生的背景變項也可能不同。因此，本研究的目的為利用潛在剖面分析(latent profile analysis)將大學一年級新生依壓力分群，接著探討不同的子群體之憂鬱症狀、網路成癮是否有不同，以及比較不同的子群體之性別是否有差異，以證明群體分類的有效性。 研究樣本與研究工具 本研究所使用的資料為長期追蹤資料(panel data)，即針對同一樣本，每半年做一次追蹤，共為期兩年(2012年4月至2014年4月)，累計5次追蹤資料。然而本研究只提取該長期資料庫中2012年4月及2013年4月所收集的兩次資料做分析。研究對象為430位大學一年級新生，其中有262位為男性、168位為女性。參與者需填答背景資訊、自我壓力評估報告(包含學業壓力、經濟壓力、親密關係壓力、同儕關係壓力、與父母關係壓力)、Becks’ Depression Inventory II (Beck, Steer, & Brown, 1996; 包含負面態度、表現困難以及身體因素)，以及研究者所屬研究團隊自行發展的Problematic Internet Use Scale (Lo, 2013; 包含 耐受性、退癮症狀、衝動使用、強迫使用、全神貫注及熱切渴望等因素)。隔年，研究者只針對憂鬱、問題網路使用變項就同一樣本進行採樣。 本研究使用Mplus 7 對自我壓力評分進行潛在剖面分析以分出最適當的子群體。爾後，子群體的憂鬱程度及問題網路使用將以單因子變異數分析(ANOVA)進行比較、性別差異將以卡方百分比同質性檢定進行分析。 研究結果 研究結果發現，透過潛在剖面分析可將大學一年級學生分成三個子群體，分別為「低壓力組」(N = 257, 59.77%)、「高壓力組」(N = 98, 22.79%)，以及「學校生活不利組」(N = 75, 17.44%)。低壓力組自我描述的五種壓力皆低於整體平均數，高壓力組自我描述的五種壓力皆高於整體平均數，不利學校生活組在學業、同儕、經濟壓力相較於其他面向的壓力較高。 進一步分析發現，高壓力組、學校生活不利組的憂鬱症狀顯著高於低壓力組，他們有較負面的態度、學習表現不佳、身體改變，而第二年也有相同的狀況。除此之外，在大學一年級時，高壓力組相較於低壓力組與學校生活不利組有較嚴重的問題網路使用，然而到了大學二年級，高壓力組與學校生活不利組的問題網路使用便無差異了。研究也顯示，男生在高壓力組的百分比顯著高於女生在高壓力組的百分比，且女生在學校生活不力組的百分比顯著高於男生在學校生活不力組的百分比。討論 研究顯示，即便上大學是生命中的轉捩點，此轉捩點對於某群學生來說影響並不大，但也有一群學生感受到多元且龐大的壓力，其憂鬱症狀較高、問題網路使用也越嚴重。然而只對學校生活感到壓力的學生，其憂鬱症狀與低壓力群無異，且問題網路使用日趨嚴重。壓力是可以被改變的，在生活環境與社會支持系統上做出適當選擇，並可運用自我信念調整以正面迎向壓力或保持忍耐等待翻轉的契機。本研究結果顯示當學生知覺到越多壓力，反應出的憂鬱症狀越高、問題網路使用越嚴重。期望此結果提供教育學者、諮商人員輔導學生的建議，以幫助高危險群的大學生遠離不幸。

Background. Feelings of stress are prevalent among college students and are sometimes accompanied by depression, thus generating great concern in society. As the literature has noted, students frequently experience academic, financial, intimate relationship, peer relationship and parent-adolescent relationship stressors. Although many researchers have investigated possible causes and outcomes of stress, most of them focus mainly on a single stressor, rather than multiple stressors. In real life, it is rare to see only one stressor; therefore, adopting a constellation of stressors in a study is more reasonable. Moreover, very few stress studies have adopted the person-centered approach that groups people who have different stress profiles. For these reasons, one of the purposes of this study is to classify college freshmen into different profiles based on the aforementioned five stressors. Research has also noted that perceiving stress could lead to adverse outcomes. Evidence has shown that stress precedes depression. Problematic Internet Use (PIU) is also one of the consequences due to poor management of stress. Additionally, college students with different backgrounds have different levels of stress. The relationships between stress profile, depression, PIU and students’ background are worthy of investigation. Objective. The present study is aimed to identify the stress profiles among college freshmen. A series of analyses would help to confirm the classifications’ validity. First, concurrent and predictive depression levels were compared among the different profile groups. Second, the researcher tested whether different groups have different concurrent and predictive PIU levels. Third, gender difference was also investigated between the groups. Sample. Panel data was collected from the same sample every half year for 2 years (April 2012 to April 2014). Based on the purpose of the present study, the researcher only extracted two specific waves’ data: April 2012 and April 2013 to analyze. The participants were 430 freshmen (261 males and 168 females) with an average age of 19.91 years (SD = 1.61). Four main kinds of data were collected from these students, including background, stress, depression and PIU. After one year, the researcher tracked the same students, and obtained depression and PIU data from 387 students (237 males and 149 females). Method. Participants completed surveys including background, self-reported stress rating, Beck’s Depression Inventory II (Beck, Steer, & Brown, 1996), and a self-developed Problematic Internet Use Scale. In the self-reported stress questions, participants rated their self-perceived academic, financial, intimate relationship, peer relationship and parent-adolescent relationship stressors. Beck’s Depression Inventory II measures negative attitude, performance difficulty, and somatic elements. The PIU scale was used to measure PIU behaviors such as impulsive use, withdrawal, tolerance and related problems. Mplus 7 was adopted to conduct latent profile analysis (LPA) which determined latent classifications for students based on their self-reported stress from five stressors. ANOVA was performed to examine whether there are differences in depression and PIU. Chi-square test of homogeneity of proportion was conducted to find out if gender difference is related to classification. Results. We found that a 3-group model was the most appropriate. The three groups were labeled as carefree (N = 257, 59.77%), all-stressful (N = 98, 22.79%) and college-life disadvantageous (N = 75, 17.44%). The carefree group exhibited low amounts in all stress indicators, while the all-stressful group displayed high scores in all stressors. The college-life disadvantageous group perceived more stressors that were related to college campus experiences, such as academic, financial and peer relationship stressors, but perceived fewer stressors in parent-adolescent and intimate relationships. Furthermore, the three groups were significantly different in the three subscales of depression. Compared with the carefree group, the all-stressful group and the college-life disadvantageous group displayed significantly higher scores in negative attitude, performance difficulty and somatic element. Such differences were also observed in the rating of depression in the next year. Moreover, the all-stressful group had more serious PIU than the college-life disadvantageous group and the carefree group. However, PIU in the second year was not significantly different between the all-stressful group and the college-life disadvantageous group. A difference between genders was also found; the percentage of all-stressful group for males is higher than that for females, and the percentage of college-life disadvantageous group for females is higher than that for males. Conclusion. The contribution of this study is to utilize LPA to identify different stress profiles among college students. Although freshmen year is a transition period, there are a group of students who do not feel stressed. In addition, individuals who perceive higher stress have a higher level of depression and PIU. However, those who are low in intimate relationship stress, but high in academic stress and peer relationship stress, are still likely to have depressive moods and even having serious PIU. Understanding the stress profiles and psychological well-being of students may provide college educators and counselors insight in identifying and offering help to high-risk students.