Edited by: Joanna Sokolowska, University of Social Sciences and Humanities, Poland
Reviewed by: Adam Roark Cobb, Medical University of South Carolina, United States; Jason Bendezu, University of Minnesota Twin Cities, United States
This article was submitted to Emotion Science, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
The COVID-19 crisis has introduced a variety of stressors, while simultaneously decreasing the availability of strategies to cope with stress. In this context, it could be useful to understand issues that people find most concerning and ways in which they cope with stress. In this study, we explored these questions with a sample of graduate and professional students.
Using open-ended assessments, we asked participants (
Participants frequently reported top problems relating to productivity (27% of sample), physical health (26%), and emotional health (14%). Distraction was the most frequently classified common strategy (43%), whereas behavioral activation was the most frequently identified effective strategy (50%). Participants who reported a common strategy classified as an EBP reported lower depressive and anxiety symptoms. In contrast, there was no evidence of an association between symptom levels and whether or not participants’ effective strategy was an EBP. Participants who reported the same strategy as both their common and effective strategy (29%) reported lower depressive symptoms than those whose common and effective strategies were different.
Our findings highlight stressors that students are experiencing and ways they are coping during the COVID-19 crisis. We discuss how these findings can inform mental health promotion efforts and future research on coping with stressors.
Coronavirus disease 2019 (COVID-19) has had an enormous public health impact. In addition to its serious physical health consequences, the virus and the resulting societal changes have had major impacts on the mental health of society (
We thought it would be useful to assess how people are
Although there are many measures of coping styles and common psychological problems, open-ended measures may be especially valuable. Many standardized measures of coping ask participants to respond to a set of predetermined items with predefined response options. In contrast, open-ended measures allow participants to freely report on their experiences without restriction. Closed-ended questionnaires have several strengths, including quantitative interpretations of scores, norms and benchmarks for comparison across different samples, and often well-documented psychometric integrity (
Taken together, this logic suggests that the information acquired from open-ended measures could be especially useful for researchers, policymakers, public health officials who are trying to understand responses to stressful situations. Therefore, we employed open-ended questions prompting participants to identify, without restriction, the coping strategies that they perceive as most useful (i.e., “effective strategies”), coping strategies that they engage in most frequently (i.e., “common strategies”), and problems that they consider most important (i.e., “top problems”).
We also wanted to examine the extent to which peoples’ coping strategies mirrored treatment components in evidence-based psychotherapies. For several decades, scholars have tested mental health interventions, often in the form of published treatment manuals. Some scholars have identified evidence-based practices and principles (EBPs) that are commonly included within the treatment manuals of empirically supported treatments (
These authors also distinguished between habitual responses (i.e., coping strategies that participants often employ) and perceived-effective responses (i.e., coping strategies that participants perceived as helping them feel better). We reasoned that a similar approach could be helpful in understanding coping strategies during the COVID-19 crisis. Specifically, we were interested in understanding an individual’s most common response to stress (hereafter referred to as an individual’s “common strategy”), the response that they perceived as his or her most effective (hereafter referred to as an individual’s “effective strategy”), and whether or not these strategies match. In a previous study,
We also wondered if individuals employing EBPs as coping strategies during the COVID-19 crisis may be experiencing better mental health outcomes. A diathesis-stress framing suggests that, in non-stressful environments, individuals with and without effective coping strategies may experience similar psychological outcomes (
Because EBPs are commonly included within the treatment manuals of empirically supported treatments and are thought to be active ingredients of change and efficacious means of managing mental health concerns (
In this study, we administered open-ended questions to assess coping strategies and top problems among
Our study has three aims. Our first aim (Aim 1) was to identify the frequencies of each effective strategy, common strategy, and top problem we identified. To that end, we analyzed the open-ended responses to identify commonly reported strategies and problems. Our second aim (Aim 2) was to identify potentially helpful coping strategies by examining associations among coping strategy use and mental health. We had three hypotheses. First, we hypothesized that those who identified EBPs as effective strategies would experience lower depressive symptoms and anxiety symptoms (Aim 2, Hypothesis 1). Second, we hypothesized the same trend for individuals who identified EBPs as common strategies (Aim 2, Hypothesis 2). Third, we hypothesized that matchers (individuals who report that their most common strategy is the same as their most effective strategy) will experience lower depressive symptoms and anxiety symptoms compared to non-matchers (Aim 2, Hypothesis 3). Our third aim was to test whether particular strategies or top problems were associated with higher symptoms (Aim 3). We discuss the implications of these findings for psychologists, higher education leaders, public health officials, and members of the general public.
The present study uses baseline data that were collected as part of an effort to disseminate a mental health promotion program to support graduate and professional students during COVID-19 (for additional details, see
Upon opening the Qualtrics link, participants were directed to a brief introductory screen with information about the study’s purpose and a general description of the activities. Participants then filled out a baseline questionnaire with measures of depressive symptoms, anxiety symptoms, secondary control, perceived ability to handle the COVID-19 crisis (described in further detail below). The questionnaire also included three open-ended questions asking participants to list their most effective coping strategy, most common coping strategy, and biggest problem. The present study uses information from the baseline questionnaire; details about the intervention are presented elsewhere (
The Patient Health Questionnaire-2 (PHQ-2;
The Generalized Anxiety Disorder 2-item scale (GAD-2;
Informed by idiographic approaches to measurement (
We want to understand how you deal with negative emotions or stress. Please list your most
Then, participants received a write-in text box to list their most effective strategy and a separate box to list their most common strategy. This order was deliberate, so that the participants would report general coping strategies, rather than those that may be specific to the top problem they described.
Informed by previous research on open-ended assessments of problems (
We want to understand problems that are causing you stress or discomfort. Please list your biggest problem or concern below. Try to be as specific as possible.
Then, participants received a write-in text box to list their biggest problem or concern.
Our codebook of coping strategies was guided by our two main goals: (a) To examine the frequency of EBPs and (b) To identify commonly reported non-EBPs.
We developed a list of EBPs by drawing from several sources. First, we reviewed a previous study which had applied a coding scheme of EBPs to coping strategies identified by middle school students (
One code, distraction, could not be neatly conceptualized as an EBP or as a non-EBP. For our distraction code, we used the definition applied by
Next, we identified coping strategies that were commonly reported but did not match EBPs. To identify these codes, we applied thematic analysis guidelines (
Our final codebook for effective and common strategies included 29 codes that match EBPs and 6 codes that do not (see our
To develop our codebook of top problems, we applied thematic analysis (
Next, we coded all responses according to whether the problem was definitely related to the COVID-19 pandemic (e.g., “my family becoming ill”), likely related (e.g., “loss of jobs/income”), or unlikely to be directly related (e.g., “the stability of my romantic relationship”). The third and fifth author applied these codes and obtained high agreement (
To address our first aim, we assessed the frequency of each coping strategy and each top problem. We were especially interested in identifying strategies that were frequently reported as effective though not common (and vice-versa). Because our data were paired (i.e., each participant provided both a common and effective strategy), we performed an omnibus McNemar-Bowker chi-squared test with strategies that were listed by at least 5% of participants (i.e., behavioral activation, distraction, social support, and “other,” a category which consisted of the remaining responses). Then, we performed follow-up 2 × 2 McNemar tests to compare pairs of strategies (e.g., comparing the proportion of participants who listed behavioral activation as effective and distraction as common to the proportion who listed distraction as effective and behavioral activation as common).
To address our second aim, we tested three hypotheses related to coping strategies. First, we tested whether participants who identified an EBP as their most effective strategy reported lower depressive and anxiety symptoms. Second, we tested whether participants who identified an EBP as their most common strategy report lower depressive and anxiety symptoms. Third, we tested whether participants who reported the same strategy as their most effective and their most common reported lower depressive symptoms and anxiety symptoms. To test each of these hypotheses, we performed one-tailed
Finally, to address our third aim, we examined if specific strategies and specific problems were associated with depressive symptoms and anxiety symptoms. To reduce the number of tests performed, we only ran tests that were adequately powered to detect a between-group effect size of
Hypotheses were stated prior to data analysis. Analyses were performed in R, and our code is available as
From 3/30/20 to 4/6/20, our survey received 561 clicks. Our sample for this present study consists of 305 individuals who began the baseline questionnaire and provided a response to our open-ended question about top problems and coping strategies. Demographic characteristics were collected at the end of the entire survey, so demographic characteristics are only available for participants who completed the survey. Demographic characteristics for these participants are reported in
Sample demographics.
M (SD) or N (%) | |
N | 305 (100%) |
PHQ-2 | 2.04 (1.69) |
GAD-2 | 2.68 (1.87) |
Age | 31.04 (8.91) |
White | 114 (66.67%) |
Asian | 41 (23.98%) |
Hispanic/Latinx/Spanish Origin | 12 (7.02%) |
Black | 11 (6.43%) |
Middle Eastern or North African | 3 (1.75%) |
Other | 2 (1.17%) |
Missing | 134 |
Female | 127 (72.99%) |
Male | 42 (24.14%) |
Other | 2 (1.15%) |
Prefer not to answer | 3 (1.72%) |
Missing | 131 |
Heterosexual or straight | 140 (81.40%) |
Bisexual | 16 (9.30%) |
Queer | 10 (5.81%) |
Fluid | 6 (3.49%) |
Gay or lesbian | 5 (2.91%) |
Pansexual | 5 (2.91%) |
Asexual | 4 (2.33%) |
Demisexual | 3 (1.74%) |
Questioning | 3 (1.74%) |
Prefer not to answer | 5 (2.91%) |
Missing | 133 |
Poor | 5 (2.89%) |
Working class | 27 (15.61%) |
Middle class | 111 (64.16%) |
Affluent | 30 (17.34%) |
Missing | 132 |
Yes | 72 (41.62%) |
Unsure | 22 (12.72%) |
No | 79 (45.67%) |
Missing | 132 |
Top problems reported during the COVID-19 pandemic.
Top problem | Percentage of people endorsing the problem (%) |
Productivity/Work | 27 |
Academic problems | 15 |
Loss of productivity | 10 |
Health | 26 |
Loved ones | 16 |
Personal health | 10 |
World health | 2 |
Frontline workers | 1 |
Emotional problems | 14 |
General uncertainty/anxiety | 9 |
Existential crisis | 2 |
Mental illness | 1 |
Lack of control | 1 |
Economic problems | 13 |
Job | 6 |
Economy | 2 |
Social distancing/Travel restrictions | 12 |
Loss of daily routine | 5 |
Isolation/Loneliness | 3 |
Far from home | 1 |
Changes to plans/Goals | 8 |
Miscellaneous | 3 |
Altruism | 3 |
Relationship problems | 1 |
Other relationships | 1 |
Roommates | 0 |
News | 0 |
No problems | 0 |
Common and effective coping strategies reported during the COVID-19 pandemic.
Coping strategy | Percentage of people endorsing the strategy as a common strategy (%) | Percentage of people endorsing the strategy as an effective strategy (%) |
43 | 15 | |
Behavioral distraction | 42 | 14 |
TV | 18 | 3 |
Food | 9 | 1 |
Productivity | 4 | 7 |
Social media | 3 | 0 |
Reading | 3 | 2 |
Music | 2 | 0 |
Cognitive distraction | 0 | 1 |
27 | 50 | |
Physical activity | 19 | 40 |
Going outside | 5 | 14 |
Social activities | 3 | 4 |
Routine | 0 | 2 |
9 | 12 | |
Friend | 5 | 9 |
Family member | 3 | 4 |
Significant other | 1 | 2 |
Help | 0 | 1 |
Feelings | 0 | 1 |
24 | 26 | |
Any EBP (distraction included) | 85 | 89 |
Any EBP (distraction excluded) | 42 | 74 |
An omnibus McNemar-Bowker chi-squared test suggested that some strategies were more likely to be listed as effective though not common, while others were more likely to be listed as common though not effective (
As mentioned, we also assessed whether or not participants’ responses matched EBPs in empirically supported interventions. Given that some scholars have conceptualized distraction as an EBP (e.g.,
We hypothesized that individuals who listed EBPs as effective strategies or common strategies (i.e., “EBP endorsers”) would report fewer depressive symptoms and fewer anxiety symptoms than individuals who did not list an EBP as their effective strategy or common strategy (i.e., “non-EBP endorsers”). For each test, we performed a sensitivity analysis removing “Distraction” from our EBP list.
Relationship between evidence-based practice endorsement and mental health.
EBP endorsement |
Non-EBP endorsement |
|||||||
Mental health symptoms | Mean | SD | Mean | SD | Mean difference | Effect size ( |
Confidence interval (CI for Cohen’s |
|
Depressive symptoms | 1.88 | 1.55 | 2.91 | 2.17 | [0.30, 0.94] | |||
Anxiety symptoms | 2.53 | 1.77 | 3.54 | 2.20 | [0.23, 0.87] | |||
Depressive symptoms (without distraction as EBP) | 1.63 | 1.47 | 2.34 | 1.78 | [0.20, 0.66] | |||
Anxiety symptoms (without distraction as EBP) | 2.37 | 1.69 | 2.92 | 1.97 | [0.07, 0.53] | |||
Depressive symptoms | 2.00 | 1.65 | 2.30 | 2.01 | 0.004 | 0.18 | [−0.19, 0.54] | |
Anxiety symptoms | 2.69 | 1.88 | 2.67 | 1.83 | 0. |
−0.01 | [−0.37, 0.35] | |
Depressive symptoms (without distraction as EBP) | 1.96 | 1.66 | 2.26 | 1.76 | 0.18 | [−0.08, 0.43] | ||
Anxiety symptoms (without distraction as EBP) | 2.72 | 1.88 | 2.58 | 1.87 | −0.08 | [−0.34, 0.18] |
Contrary to our hypothesis, we did not find that individuals who listed an EBP as their most effective coping strategy reported fewer depressive symptoms [EBP endorsers:
Consistent with our second hypothesis, we found that individuals who listed an EBP as their most common coping strategy reported fewer depressive symptoms [EBP endorsers:
We hypothesized that individuals who listed their most effective strategy as their most common strategy (i.e., “matchers”) would report fewer depressive symptoms and anxiety symptoms than those who did not list the same strategy for both questions (i.e., “non-matchers”). In our sample, 29% of participants were matchers and 71% were non-matchers. Consistent with our hypothesis, individuals whose common strategy matched their effective strategy reported fewer depressive symptoms [matchers:
As exploratory analyses, we examined the relationship between specific strategies (with at least 18% endorsement as either common or effective) and mental health problems (
Relationship between coping strategies and mental health during COVID-19.
Common strategy |
Effective strategy |
|||||||||
Strategy endorsed |
Strategy not endorsed |
Strategy endorsed |
Strategy not endorsed |
|||||||
Mean | SD | Mean | SD | Effect size ( |
Mean | SD | Mean | SD | Effect size ( |
|
Depression | 1.59 | 1.52 | 2.20 | 1.72 | 2.09 | 1.72 | 1.98 | 1.66 | −0.07 [−0.29, 0.16] | |
Anxiety | 2.31 | 1.70 | 2.82 | 1.92 | 2.81 | 1.92 | 2.56 | 1.81 | −0.14 [−0.36, 0.09] | |
Distraction | ||||||||||
Depression | 2.11 | 1.58 | 1.98 | 1.77 | −0.08 [−0.31, 0.15] | 2.22 | 1.58 | 2.00 | 1.71 | −0.13 [−0.45, 0.19] |
Anxiety | 2.68 | 1.84 | 2.68 | 1.90 | 0.002 [−0.23, 0.23] | 2.51 | 1.91 | 2.71 | 1.87 | 0.107 [−0.21, 0.42] |
Depression | 1.19 | 1.19 | 2.23 | 1.73 | 2.08 | 1.64 | 2.01 | 1.73 | −0.05 [−0.28, 0.18] | |
Anxiety | 2.09 | 1.71 | 2.82 | 1.88 | 2.88 | 1.96 | 2.55 | 1.80 | −0.17 [−0.40, 0.06] | |
Depression | 2.11 | 1.57 | 2.02 | 1.72 | −0.05 [−0.35, 0.24] | 2.10 | 2.28 | 2.03 | 1.67 | −0.04 [−0.67, 0.59] |
Anxiety | 2.58 | 1.72 | 2.70 | 1.91 | 0.07 [−0.23, 0.36] | 2.60 | 2.17 | 2.68 | 1.86 | 0.05 [−0.59, 0.68] |
Individuals who reported behavioral activation (BA) as their common coping strategy reported fewer depressive symptoms [BA-endorsers:
We did not find a statistically significant difference in depressive symptoms or anxiety symptoms based on whether participants endorsed distraction as a common strategy (
As exploratory analyses, we examined the relationship between specific top problems (with at least 18% endorsement) and mental health problems. We did not find a statistically significant difference in depressive symptoms or anxiety symptoms based on top problem endorsement (
We administered open-ended assessments to survey graduate and professional students about the top problems they are encountering during the COVID-19 pandemic and the coping strategies they find effective and use commonly. The majority of problems (81.6%) were coded as explicitly related to COVID-19 or likely related, due to widespread changes to daily life in response to the virus. We found that most participants were concerned about problems related to productivity and work-related stressors, health concerns, and emotional problems in this new context. Furthermore, many of the coping strategies that participants reported as being their most effective or most common strategy frequently corresponded with components of evidence-based interventions. We hypothesized that reporting an EBP as an
Behavioral activation was the most frequently reported effective strategy, whereas distraction was the most frequently reported common strategy. Behavioral activation is a core component of many empirically supported interventions for depression, and treatments targeting engagement in enjoyable activities and reward sensitivity through behavioral activation are effective for depression and anxiety (
Furthermore, even though half of our sample reported behavioral activation as their most effective strategy, only a quarter reported it as their most common strategy. This “common-effective gap” suggests that much of our sample may benefit from implementing behavioral activation strategies that they already view as effective. In contrast, about half of our sample reported distraction as their most common strategy, yet only 15% reported it as their most effective strategy. While behavioral activation might be underutilized, distraction might be overutilized in the context of the COVID-19 pandemic. Specific forms of distraction may be especially overutilized: watching television (18%) and eating (10%) were the most commonly reported kinds of distraction, yet very few participants listed these strategies as their most effective strategy (<1 and 3%, respectively). It is possible that behavioral activation is viewed as relatively effortful, while distraction is viewed as easier to implement (albeit less effective for addressing distress). Furthermore, certain kinds of behavioral activation (e.g., going outside, performing in-person activities with friends) may have been limited by measures designed to stop the spread of COVID-19, whereas certain kinds of distraction (e.g., watching TV, eating food) might have remained accessible. Among a cross-national survey of 551 adults, 73.7% self-reported increases in “binge watching” behavior as a result of the pandemic (
The heterogeneity in the types of distraction reported may contribute to the literature on whether this form of coping confers psychological benefits and for whom. In some therapy modalities, such as acceptance and commitment therapy, distraction is not always considered maladaptive (
Finally, some types of distraction may be adaptive for some individuals during acute stress (
Our findings suggest that
Our findings also offer suggestions that can inform efforts to help people cope with stress during the pandemic. Importantly, individuals who are commonly employing EBPs reported better mental health. While not conclusive, this finding supports the idea that teaching people to use EBPs in daily life could prepare them to cope effectively in stressful situations; such skills may be particularly valuable for buffering risk in stressful environments. Additionally, it is notable that 13 of our 29 EBP codes were not reported by any participant as a common or as an effective strategy, including exposure, finding meaning, and self-monitoring. Others were mentioned rarely, such as reframing (1% listed as effective, 0% as common) and relaxation (4% as effective, 1% as common). Interestingly, cognitive coping strategies were extremely rare relative to behavioral strategies. This is especially surprising, given that reappraisal is a highly studied emotion regulation strategy and cognitive restructuring is a well-studied tool in several mental health interventions (
Future research is needed to understand why these specific EBPs are uncommonly used. One possibility is that strategies like reframing and relaxation are often subsumed under other strategies in our codebook. For example, if a person listed “talking to a friend” as a coping strategy (coded as “social support”), we would not be able to identify if these conversations involved changing one’s beliefs, making meaning out of a difficult situation, distracting oneself from a problem, relaxing, or several other coping strategies. It is also possible that these strategies are not considered helpful in everyday coping or are more difficult to implement without guidance from a therapist or self-guided intervention. Finally, it is possible that these strategies would be helpful, but most people are not aware of them. In this case, disseminating information about these strategies and including them in interventions might be especially important. Additional research is needed to understand whether mental health professionals should prioritize teaching people new coping strategies or training people to use existing coping strategies in new ways. Such research could inform a related body of work, examining whether clinicians should focus on amplifying clients’ strengths or working on their weaknesses (
Our findings also suggest that behavioral activation, and especially physical activity, may be particularly important during the crisis. Although physical activity was one of the most frequently reported effective coping strategies, it was less frequently listed as a common strategy. During the pandemic, unfortunately, individuals’ options for physical activity have been limited. In order to safely practice social distancing, many gyms have closed and many individuals are limiting the time they spend outside; some participants even listed this as their top problem (e.g., “Haven’t been able to participate in my main stress reducing activities: gym and Brazilian Jiu Jitsu”). Such strategies, even if psychologically helpful and even necessary for minimizing viral transmission in the short-run, may lead to important health consequences in the long-run. For example, many of our participants reported common coping strategies that involve being sedentary (e.g., watching television, using social media, and refraining from activity) or consuming food or alcohol, trends which have been reported in other articles (
Our findings should be interpreted in light of several limitations. Notably, our findings focus specifically on graduate and professional students; future research is needed to understand if our findings replicate among other populations. Thus, although the mental health of students is essential during the crisis, these findings may not generalize to other groups—especially those who are more proximally affected by the crisis (e.g., healthcare workers). There are also important regional differences in how people are affected by the pandemic. Our sample comes from a university in an urban area of the Northeast and may not fully generalize to other regions. Additionally, the limited range of our depression and anxiety measures (scores on each measure range from 0 to 6), as well as limited variability in our sample, may have reduced our ability to detect effects. We also only administered questionnaires measuring the two most common mental health problems (depression and anxiety). Further research is needed to understand other mental health problems in the context of the pandemic. Future research may also help us understand how people apply coping strategies in response to specific kinds of stressors. It is possible that certain kinds of stressors are more likely to evoke certain kinds of coping strategies (e.g., pervasive stressors may elicit different kinds of coping strategies than acute stressors). Furthermore, our data are cross-sectional, meaning that our inferential statistics are not sufficient to draw causal claims. Finally, we did not restrict participants to list problems or coping strategies that were caused by the pandemic; participants were allowed to list problems and strategies that were present prior to the pandemic. This choice was intentional because we wanted to understand participants’ problems and strategies, regardless of whether or not these problems were caused by the pandemic or these strategies were employed as a result of the pandemic. Thus, future research is needed to understand which kinds of problems and which kinds of strategies are used in direct response to certain stressors.
While our study enabled us to identify coping strategies, future research could probe the quality, frequency, and promoters/limiters of these strategies. Though two individuals report activities that can be classified as behavioral activation, one may be doing so in a way that is more consistent with the ways it would be taught in an empirically supported treatment. Understanding the extent to which individuals are employing these strategies with high or low success could point to opportunities for refining the strategies that individuals are using. Additionally, in future research with the Top Problems Assessment, it could be useful to acquire more information about participants’ problems. Specifically, it may be useful to assess the severity of the problem, when the problem began, and how often the problem occurs. Such information could help researchers understand if participants respond differently to different kinds of problems (e.g., acute vs. chronic). Future research could also include follow-up studies that longitudinally track or experimentally manipulate the use of coping strategies. Intervention studies could be used to support participants in using the coping strategies that are perceived to be effective; this could simultaneously benefit participants during this crisis and test underlying theories about how coping strategies relate to distress beyond the scope of this pandemic.
The datasets for this article are not publicly available because the authors did not receive ethical clearance to share the data collected for this project. The R code associated with this project has been made available as
The studies involving human participants were reviewed and approved by the University of Pennsylvania Institutional Review Board. The ethics committee waived the requirement of written informed consent for participation.
AW conceptualized the idea for the study with RD. RF, SG, JS, and TM applied the codebooks. AW performed data analysis with oversight from RD and support from RF, SG, JS, TM, and RD. AW wrote the initial R script. RF, SG, JS, and TM reviewed the script. All authors contributed to the development of the codebooks, writing, and revising the manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer JB declared a shared affiliation with one of the authors, SG, to the handling editor at time of review.
We acknowledge Angela Duckworth, Katherine Milkman, Joseph Kay, the Behavior Change for Good Initiative, and the deans of the Graduate School of Arts and Sciences for their support.
The Supplementary Material for this article can be found online at: