The COVID-19 pandemic has ignited a secondary pandemic of mental health challenges across the globe. Reports suggest that rates of depression during the pandemic quadrupled among young adults (Ettman et al., 2020), who are particularly vulnerable to the effects of pandemic-related stress (Varma et al., 2021). The COVID-19 pandemic added an external and largely uncontrollable stressor to the lives of already stressed-out college students. During normative times—whether due to internal or external pressure to succeed, concerns about academic performance or career planning, or heavy workload—college students experience a significant amount of stress (Beiter et al., 2015; Jain & Verma, 2016). According to the American College Health Association (2018), a staggering 87.4% of college students, prior to the COVID-19 pandemic, reported feeling overwhelmed at some point during the past 12 months. Stress also has been related to depression and suicidality in U.S. college students (Adams et al., 2016; Coiro et al., 2017). It is not surprising, then, that during the pandemic, college students reported heightened levels of stress and more difficulty coping (Clabaugh et al., 2021), with well-established negative consequences for their psychological well-being (Batra et al., 2021; Clabaugh et al., 2021; Fruehwirth et al., 2021; Lee et al., 2021a). A meta-analysis of over 90,000 college students in 15 countries revealed significantly higher rates of anxiety, depression, stress, and sleep impairment during the pandemic compared to rates reported pre-pandemic (Batra et al., 2021). Furthermore, this negative impact of pandemic-related stress on mental health appears comparable for students at different levels of their education, from first-year college students (Fruehwirth et al., 2021) to those closer to graduation (Lee et al., 2021b).

Of course, the experience of stress in one domain, such as during the pandemic, rarely occurs in isolation. It is well-documented that stressors in one area may spark or interact with stressors in another, resulting in an accumulation of stress. Bakker and Demerouti (2013) developed the spillover-crossover model to describe how stress is transmitted across domains within an individual (spillover) as well as across individuals within a domain (crossover). Typically conceptualized in reference to work and home arenas, these are obviously not the only two breeding grounds for spillover or crossover stress; as previously noted, school can be especially stressful. Flook and Fuligni (2008) found an interconnectivity of family and school stress among an international sample of high school students, with one often spilling into the other, leading to what they referred to as a “spillover loop” (p. 784). Among college students, although specific stressors may differ from student to student, the core academic stressors regarding coursework, deadlines, and performance may exacerbate the stress of family and work demands outside of school, and vice versa. Such a spillover loop may be especially relevant for nonresidential, first generation, and nontraditional college students, who often work at least part-time and thus have additional demands outside of their schooling (Calderwood & Gabriel, 2017).

Given the growing percentage of these diverse student categories (Choy, 2002; Iloh, 2018), surprisingly few studies have examined spillover stress among college students. One study found that college students’ school spillover stress has been related to poorer mental health and less sleep (Pedersen, 2012). Studying the association between spillover stress and college students’ mental health and functioning is warranted, given the changing demographics of our college student population (Choy, 2002; Iloh, 2018) Even more so, to our knowledge, no study has examined spillover stress among college students during the COVID-19 pandemic. As the world went into lockdown, the pandemic facilitated an initial pivot to online learning from home, which may have exacerbated students’ stress, as school and home life became inherently inseparable. Examining the spillover stress during the pandemic thus is especially warranted, although a pre-pandemic comparison would be ideal. The present study was designed accordingly.

Given the growing evidence showing amplified stress and worsening mental health for college students during the pandemic, it is important to identify protective factors that might mitigate against the effects of stress and foster adaptive functioning and resilience. Extremera (2020) has argued that emotional intelligence may be a key factor for coping with stress during the COVID-19 pandemic, although the author noted that more research is needed. As conceptualized by Mayer et al. (2000), emotional intelligence (EI) involves a set of skills and the ability to perceive, utilize, understand, and manage or regulate emotions in oneself and others. Findings suggest a direct association between EI and better mental health for a wide range of samples during the pandemic. For example, higher EI has been associated with less depression among medical students in Greece (Skokou et al., 2021); less depression, anxiety, and stress among frontline nurses in China (Sun et al., 2021); less intense negative emotions among Polish adults (Moroń & Biolik-Moroń, 2021); better general well-being among Lebanese adults (Sfeir et al., 2021); and more resilience among Chilean teachers (López-Angulo et al., 2022). Prior to the pandemic, researchers found that college students with higher EI used more adaptive, less avoidant coping (Noorbakhsh et al., 2010) and that coping mediated the link between EI and academic performance (MacCann et al., 2011).

In the context of this growing evidence of the importance of EI for emotional functioning, the current study sought to add to the literature by examining EI as a potential buffer against stress, or moderator in the association between stress and depression, among a highly diverse sample of U.S. college students, specifically during the pandemic. Additionally, given that no publications have studied whether EI might protect against depression in the context of spillover stress, the current study intended to fill that void. It would be expected for the pandemic to elicit heightened spillover stress, given college students’ increase in learning and working from home; thus, spillover stress may be a particularly important variable to study in the context of the pandemic. Even if not in the context of a pandemic, however, knowledge about the potential for EI to buffer against spillover and other types of stress for college students would add to the literature on both EI and coping with stress among college students. To that end, capitalizing on previously collected, unpublished data, we sought to examine the links between stress, depression, and EI among students both prior to and during the COVID-19 pandemic. Such a comparison would add to our knowledge of whether EI’s potential buffering effect would be unique to the stress during the pandemic or whether it would buffer more universally under normative stressful circumstances for a diverse sample of college students.

Study aim and hypotheses

The primary aim of the current study was to examine the potential buffering effect of EI against stress in general and spillover stress in particular in connection with college students’ depression symptoms during the COVID-19 pandemic. A secondary aim was to determine whether EI serves as a buffer for college students during a non-pandemic time. The following specific hypotheses were made:

  • Hypothesis 1: Due to the impact of the pandemic, college students’ stress and depression levels would be higher in 2021 than in 2013, and COVID-specific stress would be directly linked with depression in 2021.

  • Hypothesis 2: EI would buffer against the stress-depression link by serving as a moderator at both time points.

Method

Below we report all details of our study’s methodology, including any data exclusions, all data manipulations, and all measures used. The study’s sample size was based on a convenience sample, with the aim to collect data in a given time frame from as many participants as possible to allow for analysis of more complex models.

Participants

Before COVID

Ninety college students participated in the study in 2013, ranging in age from 18 to 52 years (M age = 23.4 years, SD = 7.2 years). Participants were predominantly female (78%) and racially/ethnically diverse: 52% White, 26% Black, 8% Asian, and 14% other, with 17% identifying as Latino. Two-fifths (40%) of the sample were first-generation college students for whom neither parent had obtained a bachelor’s degree. In addition, 31 of the 90 students (34%) were classified as nontraditional based on their endorsement of at least one of the following criteria: being older than 24 years, being married, having children, or being employed full-time.

During COVID

In 2021, data were collected from a much larger sample of 382 college students, ranging in age from 17 to 68 years (M age = 22.6 years, SD = 6.8 years). Participants were predominantly female (76%) and more racially/ethnically diverse than in 2013: 36% White, 33% Black, 14% Asian, and 19% other, with 25% identifying as Latino. There was a higher percentage of first-generation college students in 2021 than in 2013 (45% vs. 40%, respectively), although the percentage of nontraditional students was similar at both time points (35% and 34%, respectively).

Procedure

The procedures for both data collection time points were similar. IRB approval was obtained, and participants were recruited from the same, moderately-sized, public, access undergraduate institution in the southeastern United States. Identical self-report measures of EI, stress, and depression were completed at both time points, along with additional questionnaires as part of a larger study. In 2021, participants also completed measures unique to stress during the COVID-19 pandemic.

The two primary differences in procedure across time points pertained to recruitment and method of data collection. In 2013, participants were recruited by flyers on campus, word of mouth, and Introductory Psychology courses, resulting in primarily psychology majors. Interested participants arranged a time to meet with researchers in-person at the campus psychology laboratory to sign the informed consent form and complete study measures. In contrast, in 2021, participants were recruited via an online Qualtrics link sent to faculty across the college to encourage student participation, resulting in a larger sample size representing more diverse majors. Participants during the pandemic gave consent and completed all measures online. A majority of participants in 2013 (70%) and 2021 (86%) received course credit for taking part in the study.

Measures

Recent and life stress

At both data collection time points, participants completed two measures of stress specifically assessing stressful life events and internal responses to recent stress. First, stressful life events were measured with the widely-used Holmes and Rahe (1967) Life Stress Inventory, a 42-item scale of possible major life events that respondents may have experienced in the last year, each associated with predetermined stress levels that are weighted accordingly, such that, for example, “death of close family member” is deemed more stressful than “change in sleeping habits.” A total score for stressful life events was based on the sum of the scale values of the events endorsed, with higher scores indicating more overall life stress, not necessarily a higher number of stressful life events. Generally speaking, scores below 150 are interpreted as relatively low life stress, scores between 150 and 300 represent a moderate amount of life stress, and scores above 300 suggest a substantial amount of life stress. The Life Stress Inventory has demonstrated strong psychometric properties in the form of strong predictive validity for its association with the effects of stress on illness (e.g., Noone, 2017). Given that it is an inventory of events that are weighted differently—in contrast to a scale measuring a stable internal state—this measure does not lend itself to conventional measures of internal consistency like Cronbach’s alpha, which assumes each item in a scale has equal weight. Second, recent stress was measured using the stress subscale of the Depression Anxiety Stress Scale (DASS; Lovibond & Lovibond, 1995), which consists of seven items reflective of internal feelings of stress and an inability to relax in the past week. Sample items include “I found it hard to wind down,” “I felt that I was using a lot of nervous energy,” and “I found it difficult to relax.” Participants indicated on a four-point Likert scale whether that item did not at all apply to them (scored 0) to whether that item applied to them very much or most of the time (scored 3); thus, raw scores on this subscale range from zero to 21. A wide body of literature has established the strong psychometric properties of the DASS with diverse samples (e.g., Lee, 2019). In the current study, the stress subscale had high internal consistency (Cronbach’s α = 0.84). The DASS stress and Life Stress Inventory scores were combined into a global index of past and present stress by standardizing each variable and averaging the two scores (M = 0, SD = 0.86). To facilitate the interpretation of results, all other predictor variables were mean-centered by subtracting the variable mean from individual scores (Hayes, 2018).

Family and school spillover stress

Family and school spillover stress were assessed via two six-item scales that respondents rated on a five-point Likert scale from strongly disagree to strongly agree (Pedersen, 2012), with high internal consistency scores of 0.83 and 0.85, respectively. Sample family spillover items include: “Worrying about family interferes with work and school life” and “My family members regularly call and disrupt me when I am at work or school, or when I am studying.” Sample school spillover items include: “I feel overwhelmed by all the things that need to be done for my classes” and “I might have better relationships if I didn’t have so many school obligations.” These scales were combined to reflect an overall level of spillover stress, with scores ranging from 12 to 60. Higher scores indicate more spillover stress.

Depression

Depression symptoms both before and during COVID were measured using the depression subscale of the DASS (Lovibond & Lovibond, 1995), which consists of seven items assessing recent feelings of depression. As with the recent stress subscale above, participants noted on a four-point Likert scale whether that item did not at all apply to them (scored 0) to whether that item applied to them very much or most of the time (scored 3); thus, scores range from zero to 21. The full 21-item DASS measure has been used in prior research and shown to have good construct validity (Ciarrochi et al., 2003). In the current sample, the depression subscale had strong internal consistency (Cronbach’s α = 0.90).

Emotional intelligence (EI)

At both time points, participants’ emotional intelligence (EI) was measured using the Emotional Intelligence Scale (EIS; Schutte et al., 1998), which has been widely used and has demonstrated strong psychometric properties. The EIS contains 33 items specifically pertaining to emotional intelligence qualities like emotion awareness and accuracy and mood regulation. Sample items include: “I easily recognize my emotions as I experience them,” “I am aware of the non-verbal messages other people send,” “I have control over my emotions,” and “I motivate myself by imagining a good outcome to tasks I take on.” Participants responded to each item on a five-point Likert scale, ranging from strongly agree to strongly disagree. Scores range from 33 to 165, with higher scores indicating higher levels of EI. The EIS had strong internal consistency in the current sample (Cronbach’s α = 0.92).

COVID stress

Participants in 2021 also completed an additional measure of stress pertaining specifically to the pandemic. The COVID-19 Family Stress Screener (CFSS), an unpublished measure developed during the COVID-19 pandemic (Huth-Bocks, 2020), consists of 10 items which respondents rated on a five-point Likert scale of strongly disagree (1) to strongly agree (5) as to their level of stress. Items include “food running out or being unavailable,” “tension or conflict between household members,” “physical health concerns for me or a family member,” and “loss of social connections, social isolation.” Scores range from 10 to 50, with higher scores indicating more pandemic-related stress. The internal consistency in the current sample was high (Cronbach’s α = 0.87) and comparable to recent studies utilizing the CFSS, which had reliabilities of 0.90 (Bates et al., 2021) and 0.85 (Haskett et al., 2022).

Demographic information

Participants completed an identical measure of basic demographic information at both time points, including their age, gender, race/ethnicity, marital status, whether or not they had children, and employment status.

Results

Data analytic plan

Basic bivariate analyses compared students’ stress levels in 2013 and 2021 and determined the correlations between the stress and depression variables (hypothesis 1). Regression analyses were performed to investigate the moderating effect of EI and sample year on the relationships between the measures of stress—(1) combined recent and life stress and (2) spillover stress—with a measure of depression (hypothesis 2). A supplemental regression analysis was conducted using only the 2021 participants to examine the impact of COVID-related stress on depression (hypothesis 1) and EI’s potential moderating effect on this relationship (hypothesis 2). To enhance the interpretability of the results, all regression analyses were conducted using mean-centered predictor variables, with the variable mean subtracted from each predictor score. An examination of the regression residuals found no egregious violations of normality or homoscedasticity. All analyses were performed using SPSS 28.

Descriptive statistics

Table 1 shows the means and standard deviations for the measures of stress, EI, and depression. The means reported are for the raw variables, prior to centering. In addition, given that the sample was predominantly female at both time points, we ran analyses initially using gender as a variable and found no significant gender effects. We also found no significant association between age and depression in our sample. Thus, both gender and age were excluded from the model.

Table 1 Descriptive statistics for stress, emotional intelligence, and depression
Full size table

Pre- versus during-pandemic comparisons

The first hypothesis stated that college students’ stress and depression levels would be higher in 2021 than in 2013 due to the pandemic’s impact and, further, that stress specific to the COVID-19 pandemic would be associated with depression in 2021. The pattern of differences between the 2013 and 2021 stress levels was not consistent across all measures of stress. As predicted and shown in Table 1, significantly higher scores were found in 2021 than 2013 for spillover stress, \(t\left(470\right)=2.15, p=0.032, d=0.252),\) and depression, \(t\left(470\right)=2.81, p=0.005, d=0.329)\). In contrast and counter to expectations, the mean for the Life Stress Inventory was significantly lower in 2021 than in 2013, \(t\left(470\right)=2.31, p=0.021, d=0.271\). This finding can be attributed to fewer major life events endorsed in 2021 (M = 9.64) than in 2013 (M = 11.21). During the COVID-19 pandemic, many activities assessed by the Life Stress Inventory were restricted due to lockdowns and quarantining, resulting in lower overall scores. Also contrary to the first hypothesis, there were no significant differences pre- and post-pandemic on participants’ reports of recent stress on the DASS.

Correlations

The zero order correlations of the measures of stress, EI, and depression symptoms are given in Table 2. All measures of stress—recent stress, life stress, combined stress (a standardized and averaged combination of recent stress and life stress explained below), and spillover stress—were significantly correlated with higher levels of depression. EI was significantly negatively correlated with recent stress and depression but was not significantly correlated with the measures of life stress or spillover stress. As predicted in hypothesis 1, COVID stress and depression symptom scores were strongly correlated, \(r\left(380\right)=0.340,p,<0.001\).

Table 2 Correlations for measures of stress, emotional intelligence, and depression
Full size table

Moderation analyses

The second hypothesis predicted that EI would moderate the link between stress and depression, with higher levels of EI associated with a reduced impact of stress on depression. It was expected that this moderation effect would be found in both the pre- and during-pandemic samples. The primary focus is thus on the two-way interactions of EI with measures of stress and on potential three-way interactions between the two moderator variables of EI and sample year. Additionally, to determine whether EI moderated the effects of COVID stress, a separate analysis was performed using only the 2021 sample. Table 3 presents the results of the moderation regression analyses for the two models.

Table 3 Moderation models predicting the effects of stress on depression as moderated by Emotional Intelligence (EI) and year
Full size table

EI significantly moderated the relationships of the combined stress and spillover stress variables with depression for both regression models—the 2013 and 2021 total sample and data from only the 2021 sample. In the 2021-only model, EI also moderated the association between COVID stress and depression. In addition to these predicted moderations, there were several significant main effects in each of the models. Specifically, EI, combined stress, and spillover stress had significant main effects in both models; year had a significant main effect in the model that combined the 2013 and 2021 data; and COVID stress had a significant main effect in the 2021-only data model.

For the regression combining the data collected prior to and during the pandemic, the significant sample year main effect indicated that, as hypothesized, mean depression levels were higher in 2021 than 2013. Main effects for all forms of stress were found, with higher levels of stress associated with higher levels of depression in both models. No significant moderation effects for year were found. In other words, for example, the main effect of EI was constant across the 2013 and 2021 samples; at both time points, individuals with higher levels of EI had lower mean depression scores than those with low levels of EI.

As hypothesized, EI moderated the impact of all types of stress on depression; however, the specific forms these moderation effects took varied for different types of stress. Figure 1 depicts the moderation effects in the model using data from the 2013 and 2021 total sample, and Fig. 2 shows the moderation effects in the 2021-only data model. The moderating effects of EI on the impact of the combined stress and COVID stress variables (Figs. 1 and 2) are similar. As stress increases, depression increases; however, high levels of EI buffered the effects of increasing levels of combined stress and COVID stress. Those low in EI were more strongly affected by higher levels of combined stress or COVID stress than were those with high EI. In other words, the ameliorative effects of EI became stronger as stress increased, with the largest buffering effects of EI on depression found at the highest levels of stress. In contrast, the impact of EI on the relationship between spillover stress and depression revealed a different pattern. When EI was low, depression remained constant and high at all levels of spillover stress. In contrast, those with high EI has low levels of depression when their spillover stress was low; however, their depression levels increased as their spillover stress levels increased. Thus, the ameliorative effects of EI seemed to dissipate as spillover stress increased, with all individuals, regardless of level of EI, converging on the same level of depression when spillover stress was high. In other words, having high EI allowed individuals to manage lower levels of spillover stress in terms of protecting them against depression, but that advantage dissipated when spillover stress levels were high.

Fig. 1
figure 1

Predicted Depression as a Function of Stress and Emotional Intelligence for the Total Sample

Full size image
Fig. 2
figure 2

Predicted Depression as a function of stress and Emotional Intelligence for the 2021-Only sample

Full size image

Discussion

The findings from this study corroborate prior research on the well-established link between stress and depression (Dyson & Renk, 2006; Hammen, 2015) and provide new insight into the role that emotional intelligence (EI) may play in moderating the impact of stress on psychological well-being. These data suggest that EI may be an adaptive quality useful for managing stress among college students during normative times as well as during the heightened, unprecedented stress of the global COVID-19 pandemic. Although no implications can be drawn about changes over time in these cross-sectional data, in separate college student samples studied at time points nearly a decade apart, EI served as a significant buffer against depression in the face of both combined recent and life stress as well as spillover stress.

This buffering role of EI occurred when stress levels from recent experiences and life events were high and, in the 2021-only sample, when pandemic stress levels were high. In contrast, EI only buffered spillover stress when those stress levels were low. Two factors might help account for this varying impact of EI as a moderator on the combined stress—depression association (and the pandemic stress—depression association) versus the spillover stress—depression association. First, the definition and nature of EI may play a role. Second, the foci of the combined and spillover stress variables may differ in important ways.

As a broad construct, emotional intelligence involves the appraisal of emotions, beliefs about how emotions should be expressed in oneself and recognition of their expression in others, the regulation of emotions, and the use of emotions in solving problems. In this study, EI was measured using the EIS (Schutte et al., 1998), with items that generally focus on the internal perception, regulation, and utilization of emotions, not on direct responses to external stimuli. The EIS’ focus on internal states is echoed in the DASS (Lovibond & Lovibond, 1995) stress scale, one of the two scales standardized and averaged to form the combined stress variable. All items of the DASS start with the pronoun “I” and focus on the frequency of emotions and experiences. Thus, half of the combined stress variable consists of scores from a scale that explicitly focuses on subjective, emotion-oriented responses to stress, precisely the type of responses that are hypothesized to be regulated by low or high levels of EI. A similar pattern was found for the COVID-19 pandemic stress measure, which also assessed the internal experience of stress. This idea of EI as a predicted buffer specifically against internal experiences of stress is supported by research on the improved emotion regulation and decreased cortisol levels found among individuals who received training in EI and related skills versus those who did not (Kotsou et al., 2011).

In contrast to the DASS and pandemic stress measures’ emphasis on subjective, internal experiences of stress, the spillover stress scale focuses on objective, external sources of stress. Even when considering emotional responses to stress, explicit external sources of stress are emphasized on this measure. The differences in the content and focus of the spillover stress measure and the other stress variables provide a viable explanation for the variations in the moderating impact of EI on the relationship between stress and depression. Furthermore, with its focus on external sources of stress, spillover stress may be more effectively dealt with by action-oriented strategies rather than the emotion-focused strategies more directly reflected in EI.

In light of the above discussion, it makes sense that EI was found to have a direct negative correlation with the combined measure of recent and life stress and a moderating effect on the impact of combined stress on depression—but that weaker results were found with the spillover stress variable. These findings suggest that even in the face of increasing combined stress, individuals with higher levels of EI are more protected against depression than those with lower levels of EI; thus, EI acts as a buffer, enabling individuals to cope with their internal responses to higher stress. At lower levels of combined stress, students had low levels of depression, regardless of the EI level. The benefit of higher EI, then, appears to come under conditions of higher internal experiences of stress.

Regarding spillover stress, as previously noted, those with higher EI had less depression than those with lower EI, but only when spillover stress was lower. As levels of family or school spillover stress increased, the buffering effect of EI disappeared, and depression scores converged for all levels of EI. Although EI was less beneficial as a spillover-stress buffer when such stress increased, not having high EI left those with even low levels of spillover stress no buffer against depression. These findings suggest that EI may be less relevant to dealing with external sources of stress and more beneficial when focusing on the internal consequences of this stress. As external sources of stress increase in intensity, they may overwhelm the coping advantages conferred by high levels of EI. Objectively highly stressful situations, then, may overpower the ability of EI to mitigate the consequences of this stress, which is consistent with the observed pattern in our study of decreasing influence of EI as the magnitude of spillover stress increases. These findings fit in the broader context of research supporting the impact of stress overload on physical health (Amirkhan et al., 2018) as well as attrition among college students (Amirkhan & Kofman, 2018). Such effects of spillover stress may be particularly high for female (Pedersen & Jodin, 2016) and African American (Boyraz et al., 2016) college students. Although no gender differences were found in the diverse sample for this study, future research should examine the role of EI as a potential buffer against stress and its health and academic consequences for different and marginalized student populations.

Further supporting this possible explanation are the findings that EI was significantly negatively correlated with depression and a measure of internal emotional responses to stress, but it was not significantly correlated with the life events and spillover measures. The results suggest that the association between EI and depression is a function of EI’s impact on emotional coping responses to stressors rather than any influence on the respondents’ assessments of the objective magnitude of the stressors themselves. In other words, EI may work specifically through a mechanism of emotion-focused coping, which is especially adaptive when confronted with uncontrollable stressors, rather than problem-focused coping, which typically is more adaptive in response to controllable stressors (Folkman & Moskowitz, 2004). Future studies should replicate these findings and further examine the hypothesis that EI is more beneficial for internal than external stress and for coping with uncontrollable, in contrast to changeable, stressors. In related research, for example, individual differences in emotional processing skills helped moderate links between types of coping and emotional expression (Baker & Berenbaum, 2007). Furthermore, evidence suggests that mindfulness training can improve one’s emotional well-being when faced with uncontrollable stressors (Mutch et al., 2021).

Investigating this pattern of findings is of particular importance in our current global environment, as the COVID-19 pandemic presents a special circumstance of an ongoing, objectively stressful situation. Moreover, the pandemic contains both controllable (e.g., whether I wear a mask) and uncontrollable (e.g., whether people around me wear a mask) stressors, whereby different types of coping may be more adaptive. Thus, the pandemic heightens the need to more fully understand the effects of stress on college students, as well as the potential benefit that EI may have, along with its limitations. In the present study, not surprisingly, depression, recent stress, and spillover stress were higher in 2021 than they were in 2013. The fact that life stress was lower in 2021 is likely a function of there being fewer opportunities during the pandemic for life events such as vacationing, moving to a new residence, and gathering for holiday celebrations. The finding that EI buffered the link between COVID-related stress and depression is encouraging and suggests that higher levels of EI may help college students to reduce the negative impact that large-scale, uncontrollable, stressful situations may have on their psychological well-being. Indeed, the COVID stress scale tapped into uncontrollable aspects of the pandemic, such as food becoming unavailable, income decreasing, and the loss of or reduction in childcare. As postulated above, EI may be particularly beneficial when one is faced with unchangeable stressors and must rely on emotion-focused coping.

To our knowledge, this study is the first of its kind to examine EI as a potential buffer in the stress-depression link comparing pre-pandemic to pandemic time periods in a diverse sample of college students. In our study, EI served as a stress-buffer for a large sample of students who were diverse in terms of race/ethnicity, age, and status as first generation and nontraditional students. Although this study thus adds significantly to the literature, several limitations must be noted. Measures were based solely on self-report, and future research would benefit from multi-method, multi-informant approaches, ideally incorporating biological markers of stress reactivity (e.g., cortisol). Also, additional research should evaluate whether EI does indeed help to regulate the emotional response to stress to a larger degree than it regulates the individual’s perception of the stressors. Finally, sample sizes larger than our sample of convenience would yield the high statistical power necessary for more complex multivariate modelling.

The current findings add to the literature supporting the importance of EI (e.g., López-Angulo et al., 2022; Sun et al., 2021) and highlight the need for effective interventions to enhance it. Taken together, these findings suggest that interventions designed to enhance EI may be an increasingly important tool in helping reduce the impact of stress on one’s well-being. To that end, Persich et al. (2021) found that college students who were serendipitously trained in EI prior to the COVID-19 pandemic exhibited lower anxiety, depression, and suicidal ideation during the pandemic. As we continue to cope with the global pandemic and resulting mental health crisis (Elharake et al., 2022), understanding the ways in which EI can be utilized to protect college students’ and others’ psychological well-being becomes critical, particularly for those who may be most vulnerable and may have lower access to mental health care. Research by Lee et al. (2021a) found that the college students most affected by stress, anxiety, and depression during the pandemic were those who were already struggling academically, who were low-income and lived in rural areas, and who were female. Related, Giangrasso et al. (2022) found that students in lockdown during the pandemic who had a stronger sense of “not mattering” had both higher mental health problems and more difficulty regulating their emotions. The better we are at fostering EI in college students, the more prepared they will be to cope with the stressors they encounter in college and beyond, such as the pandemic. Thus, enhancing EI may have wide-reaching impact on students’ academic functioning, coping with stress, and overall mental health.