The Impact of Online Education During the Pandemic On Students’ Interpersonal Relationships: A Systematic Review
Article information
Abstract
Purpose
This study investigated the impact of online education during the pandemic on students’interpersonal relationships.
Methods
In this systematic review, we analyzed2,238 articles fromdatabases such as DBpia, PubMed, EMBASE, and CINAHL, ultimately including12 articles in the final analysis.
Results
The findings revealed that male students were negatively affected by an increasein online activities, while female students were more impactedby psychological factors. Forlower-grade students, parental involvement was crucial as parents soughtto compensate for their children'sdisconnection frompeers. Consequently, educational activitiesshould be designed to promote interaction with parentsthrough play. Forupper-grade students, attempts to mitigatedisconnection occurred through online activities, but they experienced significant stress when their studieswere disrupted. Therefore, an educational support system mustbe established to alleviate anxietyand stress related totheir academic responsibilities. During this period, the social inequality faced byvulnerable groupswas exacerbatedby various factors.
Conclusion
To prepare for future crises, it is essential to addresstechnical and institutional issues that can bridgethe care gap and enhance the quality of hybrid education.
INTRODUCTION
The coronavirus disease 2019 (COVID-19) that emerged in 2019 left us with many scars. However, from another perspective, it has become a catalyst that has boldly led the technological trends of the fourth industrial revolution that various social and cultural barriers had blocked. It is also evaluated as a new paradigm shift that has spread the awareness that we can experience and feel more than we can without meeting in person through social distancing [1].
Due to the social distancing implemented in each country, the 'education' field had to shift from the traditional face-to-face method to a non-face-to-face one online. Contrary to concerns expressed in the early stages of the change, learners have indicated that online education is free from time, economic, and spatial constraints, allowing for repeated learning, which improves self-directed learning ability, and 87.3% of all learners responded that they were satisfied with online classes [2]. Educators who had taught classes for a semester also acknowledged the potential of online education, noting that the academic efficiency of students struggling with learning improved and that opportunities for individual feedback increased. There was an increase in positive responses to whether they would continue online classes in the future [3]. In the case of Korea, the 'digital new deal' was proposed as a national development strategy after COVID-19, and various plans such as building online and offline facilities, fostering digital talent, and building education platforms and e-contents are being explored [4]. In this way, the COVID -19 period has become an opportunity to recognize the potential of online education, and by utilizing its advantages, a wide range of digital capabilities and availability will be emphasized to prepare for another infectious disease outbreak in the post-COVID. Accordingly, the non-face-toface area is gradually expanding, the proportion of homeschooling is increasing, and in the future, it is expected that the utilization plan as a standard system will be systemized, such as a hybrid system that combines face-toface and online classes or for supplementary and in-depth learning for students [5].
However, despite these advantages, online education has various concerns and disadvantages. Specifically, in the case of non-face-to-face online education, there were opinions that two-way communication, such as immediate questions and answers, was limited. That concentration and participation were reduced depending on the learner. The quality and content of the class were significantly affected by the educator's enthusiasm and ability to handle digital devices [2,3]. Since the class is entirely dependent on digital devices, there are opinions that online education can further deepen existing social inequalities and individual learning deficits among vulnerable groups. In vulnerable families, it is challenging to obtain relatively expensive digital devices, the time spent at home increases, and there is a high risk of exposure to conflict or violence at home [5,6]. In addition, during the quarantine period, schools are functionally closed as an institution, so the risk of various diseases related to the decrease in students' physical exercise increased, and it is known that they suffer from extreme physical and emotional effects such as vision impairment, depression, and helplessness [6].
Furthermore, as the regular daily routine collapsed, the time spent solely on the Internet or smartphones increased, and this vicious cycle of life stress events and isolated lifestyles gradually made it challenging to direct interpersonal relationships [6-9]. As schools closed, many students were forced to stay home and deprived of opportunities to meet and socialize with peers. They began to miss their friends and realized nothing could replace the happiness they felt from human relationships in their daily lives [10]. Education in schools does not simply mean the provision of knowledge. Interactions within the school's physical space are complex relationships among students receiving education, teachers, textbooks, and various activities and symbols through which individual experiences are formed and developed [11]. In addition, it also aims to experience psychological stability and a sense of belonging through interactions with others and to develop sociality and cooperative skills required for group life [12]. Therefore, the reason for concern about online education for children and adolescents who have left school and their mental health and interpersonal relationship difficulties is that it is expected to have a significant impact in the future from the perspective of human lifespan development [13]. However, as mentioned above, the future form of education is expected to increase the proportion of non-face-toface online education. The proportion of homeschooling is increasing, and a hybrid paradigm that improves and develops existing problems is being pursued [5]. Therefore, we who live in the post-corona era have reached a point where we must anticipate the possibility of the next pandemic through this experience and prepare countermeasures. Accordingly, this study uses the systematic review [14], which is described as a research methodology that systematically identifies, evaluates, and synthesizes highquality evidence to obtain answers to specific research questions to examine the impact of online education during the pandemic on students' interpersonal relations and sociality and to think about its meaning and future direction.
1. Purpose
This study is a systematic review designed to collect and analyze literature on the impact of online education on students' interpersonal relationships. It aims to provide data on the research question and consider future directions and various alternatives in the mental health and education field.
METHODS
1. Data Search Strategy
The research questions were designed according to the criteria in the Systematic Review Manual [14]. The key research question was set according to the PICO framework (Population, Intervention, Comparison, Outcome). (P) refers to 'Students' and includes elementary to college students in the regular education curriculum. (I) refers to 'Online education,' and this also refers to the period of non-face-to-face education conducted according to school regulations, which is a lockdown due to COVID-19. (C) refers to any form of education other than non-face-to-face education, excluding traditional face-to-face education or hybrid education that combines online and face-to-face education. (O) refers to 'Interpersonal relationships.' The data collection was performed according to the PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-analysis) [15] (Figure 1). The literature search that analyzed the content from 2020 to 2024 was set to ascertain the impact of the pandemic. Moreover, the data search was conducted from April 9, 2024, to May 7, 2024. Data was collected using DBpia for domestic literature, and data was collected for foreign literature through Pub-Med, EMBASE, and CINAHL. For each search engine, 2,238 data were identified by searching 927 DBpia, 228 PubMed, 655 EMBASE, and 428 CINAHL. Among these, duplications were excluded, and the appropriateness of the data was evaluated by examining the title and abstract. When the full text was reviewed, and data were excluded according to the researcher's inclusion criteria, 12 articles were finally selected (Table 1). The two researchers who participated in the study independently conducted the selection process and had sufficient discussions until a consensus was reached. All data collected were peer-reviewed academic articles. For this reason, all grey literature, such as theses and reports, were excluded.
PRISMA flow diagram of the study selection process.
Characteristics of Included Article (N=12)
2. Search Terms
The search terms were devised by combining natural language with each word whose meaning was confirmed in MeSH Terms and EMTREE. Therefore, the final search terms for databases were composed by combining terms such as 'student' or 'adolescent' corresponding to (P) of the key question and terms that can describe online education corresponding to (I). The detailed search query for this is attached separately (Appendix 2).
3. Inclusion and Exclusion Criteria
The data was reviewed according to the following inclusion and exclusion criteria. The inclusion criteria were as follows: only students from elementary school to college who were in the regular education course and online education was implemented due to COVID-19 were accepted as data. In addition, even if the full text included mental health, it was included if the content dealt with interpersonal relations and social interaction. On the contrary, the data was excluded if the article dealt with online education but was unrelated to the situation due to COVID -19. In addition, the data was also excluded if the age was too high (>age: 35), even if the participants were enrolled in a university. In addition, data that followed a research design that did not match the purpose of this study, such as a review or a systematic review, was excluded. The language was limited to English and Korean.
4. Quality Assessment of Data
The 12 articles included 1 cohort study and 11 cross-sectional studies. The quality assessment of the articles was performed according to the NIH Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies [16]. The tool evaluates the data by reviewing each item with three answers: 'yes,' 'no,' and 'other.' If a 'yes' answer is obtained for 14 items, the highest score is 14 points (high quality). Two researchers independently evaluated this process, and the results were reviewed later.
RESULTS
1. General Characteristics of the Data
A total of 12 articles were included in the data analysis. The research design includes 1 cohort study and 11 crosssectional studies. By country, they were Turkiye (n=3), Korea (n=2), India (n=1), Thailand (n=1), China (n=1), Rumania (n=1), Portugal (n=1), UK (n=1), and Poland (n=1). As for the publication years, since the research was conducted after the COVID-19 lockdown, they were relatively recent: 2023 (n=3), 2022 (n=4), and 2021 (n=5). Since the participants were children and adolescents, data were collected directly from the students. However, there were also two articles in which data were collected through parents who observed them. In one of these articles, the number of parents who participated in the survey was confirmed, but it was unclear how many children were targeted.
Among the included data, the article that analyzed the most significant number of participants was 10,199, and the article that analyzed the smallest number of participants was 104. When excluding one article whose exact number of participants was unknown, the average number of participants recruited from the 11 articles was approximately 2,112. Although no distinction was made between diseased and non-diseased groups, one article deals with the social function of children with ADHD (Attention Deficit Hyperactivity Disorder). By grade, students enrolled in elementary school to university were included, and the ages ranged from a minimum of 6 to a maximum of 34 years. The 34-year-old participant may be considered heterogeneous data, but the average age of the data was 20.2 years, so it was included as meaningful data. One of the included data specifically stated that the participants were recruited from elementary, middle, and high schools, but the exact age of the participants was not described separately.
Regarding sex, two articles did not indicate and could not be confirmed. However, there were two articles (17%) in which males were more prevalent and eight (67%) in which females were more prevalent, confirming that the proportion of females was higher. The outcome of interest, 'interpersonal relations,' was measured with different assessment tools in 12 articles. This is also a limitation because it is challenging to synthesize the highly abstract concept. However, the researcher analyzed the full text of each article and tried to classify the content that fits the purpose of the study, such as interaction with peers and parents and social functions. The data collection period and the time point after the lockdown were described. The national lockdown action described or known in the literature mainly was in March 2020, thus varying from the earliest after 1 month to the latest after 22 months. However, the study design limitations made it difficult to describe the differences.
2. Characteristics of Primary Variable (Interpersonal reLa-Tionships)
When examining the measurement of interpersonal relationships, which is the primary variable of the study, it was confirmed that they were all measured using different assessment tools. In a total of 12 articles, the assessment of the variable was classified into two. In 9 of these data, students evaluated their own interpersonal relationships aspects [17-25], and in the other 3 data, parents who observed their lifestyles evaluated them [26-28]. Each assessment tool and main concepts can be confirmed through the table (Table 2). If we look at each specific variable, it can be described as social-emotional competency before and after COVID-19 [20], satisfaction with social interaction [23], adjustment to school [17], social dysfunction [19], psychological responses [25], cognitive ‧ emotional ‧ behavioral difficulties or problems in social interaction [18,22, 26-28], prosociality [24,26], perceived sociability & social intelligence [21], relationship between children and their parents or peer relationship [17,18,24-26]. As we can see, the researcher subjectively interpreted each variable in the study as being able to explain interpersonal relationships. However, there was no agreement or consensus on the tools, so unfortunately, this deviation in definition can be considered a bias of the study.
Measurement of Primary Variable (Interpersonal Relationships)
3. Result of Syntheses
When synthesizing the study results, it can be inferred that online education significantly affected their interpersonal relations. The proportion of those who reported difficulties in face-to-face interactions with others increased from 59.8% to 95.5%, and the proportion of those who experienced restrictions in physical contact and social activities increased significantly from 63.4% to 97.35% [21]. Children experience negative emotions due to isolation from their peers, resulting in problems such as irritability, fussiness, and tantrums, further aggravating the difficulties [26,27]. In addition, they were very dissatisfied with the limited interactions with teachers and peers after implementing online education [22]. They expressed that they experienced increased feelings of loneliness, depression, and anxiety, decreased happiness, social isolation, lack of emotional interaction, and feelings of separation [18,21,23,27]. In one study, it was reported that 77% of the total was in the moderate or higher range based on the GHQ (The General Health Questionnaire), which indicates the level of mental health [19], and it was found that dissatisfaction with online education had a significant impact on these negative emotions [19,21,23,24]. Therefore, it was confirmed that students in childhood and adolescence, which are important periods when peer interaction is emphasized and social skills are learned through experiences of interacting with others, suffer significant disadvantages [12,13]. This result supported concerns that the pandemic, which was a concern in the early stages of the COVID-19 period, would harm children's acquisition of social skills [6-10]. In particular, the main variables that showed notable results were sex, grade, and vulnerability. The specific details for each variable are as follows.
1) Sex
Sex differences in interpersonal relations showed inconsistent results. Both sexes showed significant changes, but they showed different positions on which sex was more vulnerable. Some data explained that females experience more withdrawal and dissatisfaction in interpersonal relations, but this was explained as having a more significant negative impact on females who are vulnerable to psychological changes such as depression or anxiety [17-19]. However, another data reported that males experienced greater dissatisfaction, which was attributed to the fact that during the pandemic, males spent more time online or playing games than females. Their daily lives became irregular, exposing them to more risks [20]. In addition, there was data that male who are relatively clumsy in expressing their emotions suppress them, experiencing greater loneliness and emotional and behavioral difficulties [21,26]. Other data did not show any differences by sex, so additional research seems necessary.
2) Grade (or age)
The vulnerable period presented in each data was reported differently, as middle school, high school, or first graders are moving up a grade or adjusting to a new semester, so the results were inconsistent. However, it is noteworthy that the adaptation patterns of lower grades (at the elementary school level) and upper grades (at the middle school level) differ regarding interpersonal relations. First, as the educational place of lower grades (at the elementary school level) shifts from school to home, the target of interaction with them changes from peers to parents. At this time, parents play the role of peers who play with their children while also playing the role of educators. Since these children are not yet fully accustomed to keeping rules and promises at school, many attention problems have been reported since the transition to online education [25]. Therefore, the role of their parents as educators was emphasized, and the younger the child who has difficulty adapting independently, the more the parents support their children's online learning [26]. One of these data reported that children with internalization problems had fewer problems when their fathers worked from home. Therefore, parents' occupations, residence, and whether they work from home influence students' interpersonal relations [28]. As such, children felt dissatisfied with the limited interaction with peers, but on the contrary, the younger they were, the more positively they perceived the increased time with family [22]. The more positive the interaction between parents and children during COVID-19, the more it relieved children's psychological dissatisfaction. It restored their mental health, and this change led to satisfaction with interpersonal relations [22, 24,25]. Therefore, the home's function and parents' role play an important bridging role in online education and interpersonal relations for lower grades. Therefore, the appropriate solution for them is to view education from the concept of play to promote interaction between parents and children and to provide socialization opportunities [18,27].
In contrast, in the case of middle school students and older, the time spent on online activities increased significantly, and they showed a new adaptation pattern of trying to fill the gap in interpersonal relations that was created by not being able to interact with their peers or teachers in the online environment. They were satisfied with the current level of interpersonal relations and adapted well to the online environment. In other words, even if they received online education, they did not experience significant changes if they maintained their daily routines well [20,28]. However, it is important to note that the lower their self-perceived interpersonal abilities, sociability, and social intelligence were reported, the more they were immersed in online activities [17,21,25]. In addition, as online activities increased and they were unable to manage their daily routine well, conflicts with their parents at home deepened, and mental health problems and stress increased significantly [17]. Another coping mechanism involved decreased social and increased learning activities, which were more prevalent among the older grades. However, although they are using their leisure time to study on the surface, they are experiencing a high stress level proportionally to the increased study time, so further research is needed [17]. Therefore, it was found that when the stress situation peaks, they give up on their studies [19]. What is notable about this type is that following the transition to online classes due to COVID-19, middle and high school students and older cited 'worries about dropping out of school' as the main stress factor, which includes concerns such as not being able to understand class content well due to academic stress or low-quality classes, worries about grades, and difficulty building qualification requirement for university admission [18,19]. In addition, the higher the dissatisfaction with this educational system and limited social interaction, the more negatively it affected psychological indicators such as depression and anxiety, which also showed a negative evaluation of the overall interpersonal relation status [23]. This can be seen as a difference from the lower grades, where dissatisfaction with social interaction was the most significant concern. Therefore, an appropriate solution for upper grades is to provide high-quality education and establish a systematic educational support system that does not cause concerns about dropping out of school [18].
3) Vulnerability
Another point of note is the deepening of social inequality for vulnerable groups. Many students who could not attend school due to restrictions on outdoor activities and became independent returned to their homes, and the time spent with their parents increased [19,22]. As a result, children are inevitably affected by their parents' parenting style. For example, risk factors such as single-parent families, low economic status, or a family atmosphere that harms children, such as lack of supervision by parents, have been shown to have a negative impact on children and adolescents [17,20,26]. In addition, parents who have to care for these children at home instead of at school and feel burdened also experience negative emotions, intensifying conflicts between them. Reports have shown that the negative impact on children living independently worsened when they returned home [19,28]. In addition, the economic level in this problem means living in a small city, not a large city, and other issues such as problems with the Internet connection or insufficient electronic devices, so it has been known that these problems cause dissatisfaction with online education, which in turn harms emotional problems and interpersonal relations [19,21, 23,24]. In particular, the more severe the disease group, the more these difficulties are experienced. Children with ADHD required more support from their parents, and it was also shown that they had difficulty concentrating in the online class environment at home, which led to an increase in distraction and carelessness [28]. In addition, in the case of children and adolescents with ASD (Autistic Spectrum Disorder) or Down's syndrome, there were a small number of people who were reluctant to contact people due to previous experiences and perceived this period positively, but this was found to be a barrier that further deepened their interpersonal relationship difficulties by depriving them of the opportunity to interact with various people and develop sociality [12,29,30]. Therefore, there is no clear evidence that the pandemic itself worsens the symptoms of children with the disease, but it is clear that it takes away their social learning opportunities in the long term [28]. In addition, a small number of data have shown some differences by ethnicity, such as those who are ethnic minorities, mixed race, or who are reluctant to disclose their gender when participating in research tend to be psychologically unstable [18]. It was also shown that sexual minorities, referred to as LGBT (lesbian, gay, bisexual, transgender), were more adversely affected as they returned home and their conflict with their families deepened [19]. Ultimately, the opinion was supported that the global pandemic called COVID-19 would deepen not only socioeconomic inequality but also inequality between classes in terms of opportunities to develop education and interpersonal relations [5,6].
4. Result of Quality Assessment
The results of the quality assessment of the 12 articles using the NIH Quality Assessment Tool [16] are shown in Appendix 3. Suppose each question is evaluated as 'yes,' the highest score is 14 points. The items consisted of 14 questions; the highest-scoring article was evaluated as 10 points, and the lowest was evaluated as 3. The average score was 7.33 points, and since the quality rating decreases as the risk of bias increases, it was found that the articles had some degree of bias. However, there was no cut-off criterion for excluding data from the quality assessment tool, and the researchers discussed that considering the Special circumstances of COVID-19, data with a high risk of bias would also be considered sufficient significance. Therefore, all 12 articles were included. Looking into this in detail, except for one cohort study, all were designed as cross-sectional studies, so question 13, asking about the dropout rate in follow-up studies, was evaluated as 'Not applicable (NA).' The sample size was large, ranging from a minimum of 104 to a maximum of 10,199. However, the regrettable point in many articles is that very few articles presented specific data, such as the appropriate sample size and the effect size estimate. In addition, since the study was conducted as part of the national lockdown, it was difficult to conduct a preliminary investigation, which is a weakness of the design. In addition, although most articles specify the data collection period, they do not explain how long it took for online education to be implemented in each country. Therefore, if the various exposure levels of online education had been investigated and repeated studies had been conducted by dividing each period after the implementation of online education, the quality of the study could have been improved. However, since this detailed classification was not divided, there was a point where the bias was created. In this study, online education was evaluated as having a definition as it refers to a period of non-face-to-face education that occurred during the COVID-19 lockdown. However, this was also a limitation of the research design that made it impossible to blind each evaluator.
DISCUSSION
The nightmare of the pandemic has passed for a while, and we are catching our breath. We, who suffered terrible wounds because we were not prepared, must prepare for the future, the new normal. Therefore, although this article deals with the past as a review, its orientation is toward the future. In this study, online education refers to the change in education due to the lockdown of educational institutions due to COVID-19. This study synthesized 12 articles and examined the impact of online education on students' interpersonal relations.
In conclusion, some students were found to have been negatively affected in their interpersonal relationships during the pandemic. The interpretation of the result is cautious, but some speculations are as follows: Due to the restrictions on physical contact during the lockdown, they were deprived of the opportunity to learn social skills [17,27]. As the school disappeared, the communication space where they could cooperate with others and exchange signals with each other disappeared [11,12]. In addition, during lockdown, they experienced negative emotions such as depression, anxiety, and loneliness [17-19,21, 24,26], which may have been a catalyst for them to perceive encounters and interactions with others more negatively. Their daily lives were disrupted by being engrossed in the Internet [20,25], and their interpersonal skills were impaired. They experienced dissatisfaction and discouragement [20,23]. As this situational stress and the time of the collapsed daily life accumulated, their interpersonal difficulties were presumed to have worsened [6-9]. Accordingly, the role of the family, the only direct communication partner, significantly increased [22,27]. However, it was difficult to expect this protective effect when the family was vulnerable [19,26]. It became clear that some students faced more significant difficulties due to exposure to conflict and violence at home and difficulties acquiring the digital infrastructure necessary for education [5,6,17,20]. We obtained two key points to prepare for another pandemic through this.
The first is the necessity for a 'solution to the care gap.' As previously discussed, some students who maintained their daily lives and demonstrated resilience could resolve and overcome dissatisfaction with interpersonal relations and loneliness [17,20]. In addition, children who positively perceived their relationships with parents were shown to function as positive protective factors in various mental health indicators [24]. However, in cases where there was no adult to supervise their lives, vulnerable groups such as single-parent families or those with family discord were shown to have a higher rate of experiencing greater dissatisfaction with interpersonal relationships and psychological loneliness due to the breakdown of their daily lives and their immersion in the Internet [19-20,26]. As the isolation period extended, children adopted irregular lifestyle patterns, such as going to bed late and waking up late, and experienced various side effects, including decreased physical activity [27]. Therefore, even if online education is implemented, it may not be significantly affected if a regular lifestyle pattern can be maintained [28]. Moreover, since lower grades are more dependent on family interaction, and through this, they overcome their peer deficiencies and learn social skills, the lack of proper care for them implies an even greater risk [17,18]. Suppose they have to take on this role again within the family. In that case, ironically, this will create a structure that puts a heavy burden on the child and the parents of the family to simultaneously fulfill the roles of caregiver and educator [27,28]. Therefore, in modern times, where females are entering the workforce and the number of dual-income couples is increasing psychological and institutional support should be provided for these various vulnerable groups because they may experience fear and anxiety about leaving their children in this situation once more. Therefore, to prepare for the future change in the educational paradigm and the new normal, it is important to focus on a holistic perspective that looks at a family and society rather than simply focusing on individuals [28].
The second keyword is that a change in educational methods is necessary. Since we have experienced the benefits of online education, it is expected that hybrid education combining traditional and online methods will become mainstream [5,21]. However, as emphasized earlier, emergency remote teaching stole opportunities for insufficient socialized children. Therefore, future education must address the problems sufficiently and consider how children can acquire social skills online [17,20]. The education provided in existing schools influenced the curriculum and the sociality they learned through interactions with others and the manners, discipline, and moral standards necessary for living [27]. Feeling each other's bodysensory dimensions of presence in sociality was natural. However, it can be understood that the opportunity and experience to recognize each other's language and eye contact are limited, and each person's joy, delight, anxiety, or tension signals are limited. Therefore, the lower grades who did not have sufficient experience and children with diseases (ADHD, ASD) who are vulnerable to social interaction suffered even more significant deficiencies. Therefore, for these vulnerable groups, each session of online education should be set to a time that allows them to pay sufficient attention [27], and there needs to be an in-depth discussion on how to teach these sensitive signals even without physical sensory contact [17,18,27]. Therefore, some literature recommends play-based education as a suitable method for them [18]. In addition, it was found that in the case of middle school students and older, the lower their satisfaction with interpersonal relationships and their ability, the more time they invested in learning to relieve their anxiety [20]. The sudden change in educational methods confused teachers who had to teach them, and students who received a lower level of education experienced greater dissatisfaction and greater psychological loneliness, depression, and anxiety [18,20,21,23]. Therefore, sufficient educational resources should be secured in future educational changes, and the quality of e-learning should be improved so they can enjoy it. Sufficient infrastructure and open access should be secured, and efforts should be made to create and operate suitable platforms. Also, digital literacy should be emphasized more, as it is essential for maintaining healthy social networks and obtaining correct information online [19]. This alternative is expected to be a more effective solution, especially for middle school students and older.
In conclusion, the core of the discussion for educational changes should begin with considering how to provide opportunities for socialization and high-quality communication in an online environment to the upcoming hybrid education revolution and another pandemic [17,20]. The limitations of this study are as follows: Firstly, data collection and recruitment in the collected data were conducted online due to quarantine due to the COVID-19 lockdown. Therefore, problems with sample selection or data reliability may be expected, and the participants' medication or substance use was not strictly reviewed. There are some limitations in inferring and generalizing the results because there is a high dropout rate due to online recruitment. Secondly, this study focuses on online education implemented due to the pandemic. However, since data from 9 countries were used, the lockdown and forms of online education implemented in each country may be different. Therefore, since the situations of each country are different, there is a risk of bias because there is no explicit agreement on arbitration. Thirdly, the final inclusion of data was made based on the researcher's subjective judgment. Therefore, there is a risk of bias because there is no precise operational definition of the abstract concept of interpersonal relations. Finally, most of the data is cross-sectional research. Therefore, It does not provide strong evidence to infer causality because of various confounding variables. Therefore, follow-up studies such as longitudinal or RCT data analysis are required. In addition, the data of included articles has a weakness in that there is a significant difference in the timing of data collection, from 1 month to 22 months after the lockdown. Therefore, follow-up research is needed on the differences between the early and late stages of the pandemic and the differences in the duration of online education.
CONCLUSION
This study is a systematic review of the impact of online education on students' interpersonal relations. Among the main conclusions obtained through 12 articles, the notable variables were sex, grade, and vulnerability. Depending on sex, all students were negatively affected by the transition to online education. However, females were found to be more affected by psychological factors such as depression and anxiety, and males were more absorbed in online activities, which made the situation worse. When classified by grade, in the case of lower grades, much meaning was found in the roles of parents and family. In the case of upper grades, they were significantly affected by it as it was closely related to learning efficiency and interpersonal relation satisfaction. Finally, in the case of vulnerable groups, due to various factors such as family atmosphere and the presence of illness, it was found that they had to shoulder this burden entirely within the family, which led to a fatal impact. For this reason, it is thought that they lost the opportunity to develop interpersonal skills, and various inequalities have gradually deepened. Therefore, this study suggests that in preparation for such pandemics or crises in the future. First, the care gap between classes should be resolved so that the negative impact on vulnerable groups at risk of severe damage to interpersonal relationships is not aggravated. Second, in preparation for changes in the hybrid education system, a system should be established to enable sufficient socialization online. However, these results require caution in interpretation because the definition of interpersonal relations, the timing of online education, and the policies of each country are different. Therefore, additional research is needed to address such biases.
Notes
The authors declared no conflicts of interest.
AUTHOR CONTRIBUTIONS
Conceptualization or/and Methodology: Hwang, DW & Han, KS
Data curation or/and Analysis: Hwang, DW & Han, KS
Funding acquisition: None
Investigation: Hwang, DW
Project administration or/and Supervision: Han, KS
Resources or/and Software: Hwang, DW & Han, KS
Validation: Hwang, DW & Han, KS
Visualization: Hwang, DW
Writing: original draft or/and review & editing: Hwang, DW & Han, KS
