Online learning in the wake of the COVID-19 pandemic: Mixed methods analysis of student views by demographic group

Abstract

Various aspects of online learning have been addressed in studies both pre- and during the COVID-19 pandemic. However, most pre-pandemic studies may have suffered from sampling selection issues, as students enrolled in online courses were often not comparable to those taking classes on campus. Similarly, most studies conducted during the initial stages of the pandemic might be confounded by the stress and anxiety associated with worldwide lockdowns and the abrupt switch to online education in most universities. Furthermore, existing studies have not comprehensively explored students' perspectives on online learning across different demographic groups, including gender, race-ethnicity, and domestic versus international student status. To address this research gap, our mixed-methods study examines these aspects using data from an anonymous survey conducted among a large and diverse sample of students at a mid-size university in the Northeastern United States. Our findings reveal important insights: (1) Females are nearly twice as likely as males to prefer online asynchronous classes and feel self-conscious about keeping their cameras on during online synchronous (e.g., Zoom) classes. However, gendered views and preferences align in other aspects of online learning. (2) Black students show a stronger preference for Zoom classes compared to online asynchronous classes and emphasize the importance of recording Zoom meetings. Hispanic students are twice as likely to prefer asynchronous online classes, which offer greater flexibility to manage multiple responsibilities. (3) International students value the ability to learn at their own pace provided by online learning but express dissatisfaction with the lack of peer interaction. On the other hand, domestic students are more concerned about reduced interaction with teachers in online education. Domestic students also exhibit a higher tendency to turn their cameras off during Zoom classes, citing reasons such as self-consciousness or privacy. These findings carry significant implications for future research and educational practice, highlighting the need for tailored approaches that consider diverse student perspectives.

1. Introduction

The COVID-19 virus outbreak forced many academic institutions to shift their education from in-person to remote in a matter of days (Dhawan, 2020). Many professors, lacking experience in online education, were left to determine a proper adjustment for their courses during this transition while students, lacking experience in taking online classes, were left to adjust (Gillis & Krull, 2020). As a result, most research on students’ views and experiences with online learning conducted at the peak of the COVID-19 pandemic likely confounds the stress and turmoil caused by the pandemic itself and by the abrupt transition to online learning with the effectiveness of online course delivery.

At the same time, studies of students’ experiences with online learning and its effectiveness conducted in pre-pandemic times suffer from the sample selectivity problem: most students had a choice about whether to enroll in online classes. We can reasonably expect that students who self-select into online courses are more advanced and disciplined and/or more burdened with multiple responsibilities restricting their options of attending classes on campus. This leaves researchers at a likely impasse for finding comparable samples of on-campus and online learners.

As the COVID-19 pandemic wanes and college education returns to normal, it offers researchers a unique opportunity to better understand students’ perceptions of online learning, since almost all college students have experienced both online and on-campus learning. Given this changing landscape of online learning in higher education, the current mixed methods study explores student perceptions of various types of online and in-person learning and how features such as camera use in live synchronous online classes inform student preferences. Our study includes a large, diverse sample of students (n = 690) from a mid-size university at the Northeast coast of the United States (approximately 6000 students total), which allows us to compare experiences of students from different demographic groups:

• •Males and females.
• •Black, White, and Hispanic students.
• •International and domestic students.

2. Literature review

2.1. The origins and types of online learning

Distance or remote learning is a form of education in which the teacher and student are separated geographically and often engage with the content at different times (Kentnor, 2015). The first major online learning experience in the United States happened in 1950 when the University of Houston, TX, began to provide collegiate courses over the nation's first university-licensed radio station, KUHF-FM (Fisher, 2012).

Two major forms of remote learning have since developed, known as asynchronous (async) and synchronous (sync) classes. Synchronous communication, the most akin to in-person learning, requires student and teacher to be logged in at the same time, to review the material. Asynchronous communication offers more flexibility, by allowing the participant to engage with the content at their convenience. The difference between sync and async learning is similar to a difference between a phone call and email conversation. Group discussions have shifted from conversations within the classroom to making comments in an online forum and responding to others without their physical presence. Cameras during sync classes have been introduced to help increase engagement, but some universities have stated that they do not require cameras to be on during classes for various reasons (Castelli & Sarvary, 2021; Kahn, El Kouatly Kambris, & Alfalahi, 2021).

Thus, the pedagogy for online learning is not a new emerging method of teaching and learning; rather, the eruption of the COVID-19 pandemic has worsened the situation and forced educational institutions to shift to online learning abruptly and entirely, to reduce the spread of the COVID-19 virus (Chung et al., 2020; Pokhrel & Chhetri, 2021). This shift has impacted students’ social interactions and performance, caused widespread stress and anxiety, so it is important to review prior studies within the pre-pandemic versus during-pandemic framework.

2.2. General perceptions of online learning before the COVID-19 pandemic

Long before the COVID-19 pandemic, empirical research has highlighted the preferences and issues students have with online learning (see Bygstad et al., 2022; Goldenson et al., 2021). One U.S. Department of Education meta-analysis found students in online learning outperformed students in traditional classrooms, and preferred online conditions (Means et al., 2009). Several studies found flexibility and agency were factors that drove students towards the online format: students felt they had greater control over their learning, experienced enough student interaction, and were able to gain new knowledge through online learning (Eastmond, 1998; Stone & Perumean-Chaney, 2011). Several studies have also found students’ appreciation of flexibility in online learning, particularly in asynchronous courses: the ability to choose their place of study, schedule and timing, and determine the pace of studies (Alexander et al., 2009; Nollenberger, 2015), easy use of online devices like laptops and smartphones (Stanciu & Gheorghe, 2017).

In addition to effectiveness, early research has found online learning environments to foster relationships and interpersonal skills: online networks offer interactions with people who have different social characteristics and opportunities to find shared interests (Wellman et al., 1996). However, subsequent studies are more likely to find a perceived lack of community as a common issue with online learning (Means et al., 2009; Nollenberger, 2015; Song et al., 2004). In addition, the Department of Education's meta-analysis noted that research results did not find online learning to be a superior method, even though it has some advantages over in-person learning (Means et al., 2009). A 13-country European study found that students saw online learning as less effective than in-person learning (Tartavulea et al., 2020). A South African-based study found online-learning students faced isolation and sometimes lacked confidence, especially those who were new to online learning (De Metz & Bezuidenhout, 2018).

2.3. Perceptions of online learning during the COVID-19 pandemic

During the COVID-19 pandemic, empirical research found similar themes of flexibility and convenience motivating students to prefer online learning (Muthuprasad et al., 2021). When Harvard medical students were forced to go remote with their education, student evaluations of the experience revealed a conducive learning environment, with high course ratings (Goldenson et al., 2021). From nursing students who were able to use training videos for active learning (Zhou et al., 2020) to accounting majors where online students overwhelmingly outperformed in-person and hybrid students (McCarthy et al., 2019), online learning benefits continued throughout the pandemic.

However, criticisms were also present. One study found students strongly preferred face-to-face classes (Aguilera-Hermida, 2020) while others found technical issues like internet connectivity were a common problem in developing countries (Muthuprasad et al., 2021). Moreover, the pandemic may have induced additional burdens, and several studies found evidence of increased levels of anxiety and stress, lack of focus, and fatigue in students, which may have affected their perceptions of online learning (Curelaru et al., 2022; Parker et al., 2021).

2.4. Perceptions of camera use in online synchronous learning

An important aspect of live online learning is the use of cameras to connect in the virtual environment. Most teachers make camera use optional but then, few students keep their cameras on, resulting in a disconnected educational experience (Castelli & Sarvary, 2021) and slower pace (Kahn, El Kouatly Kambris, & Alfalahi, 2021). Students have generally reported minimal camera use when not required (Bucholz, 2020; Melgaard et al., 2022; Hariharan & Merkel, 2021), acknowledging self-consciousness, uncomfortable surroundings, feeling insecure (Bashir et al., 2021; Bergdahl et al., 2022; Kahn et al., 2021), and also internet connection and focus issues (Castelli & Sarvary, 2021). Adherence to social norms is also important: Hariharan & Merkel (2021) found 94% of students favored keeping their cameras off for fear of being the only visible student.

Conversely, research suggests that when cameras are on, physical appearance and visible background become the primary focus of students, creating significant barriers to engagement with course material (Castelli & Sarvary, 2020; Neuwirth et al., 2021; Rajab & Soheib, 2021). This problem may be exacerbated for marginalized populations whose race, ethnicity, or gender identity may affect camera use (Oinas et al., 2022; Aresan & Tiru, 2022; Castelli & Sarvary, 2021). Research also finds that students report wanting others to keep their cameras on for increased interactivity and intimacy in the classroom (Schwenck & Pryor, 2021; Sedereviciute-Paciauskien et al., 2022). Additionally, students prefer professors to keep their cameras on (Kahn et al., 2021).

Amidst these mixed findings, students often feel they should make decisions about turning on their camera (Williams & Pica-Smith, 2022). On the other hand, teachers feel a heightened sense of responsibility for their students’ success and, when cameras remain off, they frequently experience the sensation of “teaching into the void” (p. 39). Overall, reduced camera use appears to be common in sync online classrooms but may cause decreased levels of engagement and poorer educational experiences (Melgaard et al, 2022; Cavinato et al., 2021).

2.5. Gendered differences in higher education and online learning

Recognizing gender differences in experiences of students with online learning can help improve student engagement and success. Literature on gender differences in response to online learning has been of significant interest following the pandemic, mainly regarding STEM students due to gender biases in these fields (Nadile et al., 2021; Stolk et al., 2021).

Research has shown that women in STEM were nearly three times less likely to feel comfortable asking or answering questions in class, despite their race, generation status, GPA, or academic year compared to men (Nadile et al., 2021). Women reported decreased confidence in material and increased fear of judgement from peers or instructors. Female STEM majors reported increased external motivations in seeking self-improvement (Stolk et al., 2021). Even in the social sciences, where women make up a larger portion of students, such gendered differences in learning comfort may still exist; in a survey of political science students, females enrolled in more online courses while males had primarily on-campus schedules (Glazier, Hamann, Pollock, & Wilson, 2019).

Evidence suggests the shift to online learning may have also exacerbated pre-existing physical and mental health conditions (i.e. anxiety), particularly for female students (Li & Che, 2022). Considering synchronous learning and camera use, Castelli and Sarvary (2021) found female students more likely to report concerns over their appearance, physical location, or the presence of others in their background, perhaps indicative of symptomology of social anxiety. Conversely, some studies found support for a higher female student satisfaction with online learning (Aresan & Tiru, 2022; Chung et al., 2020; Elkins et al., 2021) and their greater confidence in these settings (Aresan & Tiru, 2022). Female participants were more comfortable speaking during online courses, using their camera, and being satisfied with teachers’ pedagogy (Aresan & Tiru, 2022). Together, these findings suggest that the traditional classroom or spaces in which women must display themselves may provoke a social anxiety associated with gendered expectations whereas online learning, according to Aresan and Tiru (2022), may provide a reprieve from the anxious conditions associated with such emotion work. The experience of social anxiety in these settings may also align with the concerns of female STEM students in the prior paragraph (Stolk et al., 2021), in which women feel far less comfortable asking questions or engaging in the coursework in front of others.

2.6. Educational experiences and perceptions of Black and Hispanic students

Before examining the experiences of Black and Hispanic college students, it is imperative to acknowledge the systemic challenges they incur throughout primary and secondary education.

2.6.1. Structural disparities

Structural limitations in education are a common challenge within both Black and Hispanic communities. Chakraborty and Bosman (2005) found significant inequalities regarding personal computer ownership in Black compared to White households. Despite efforts to close this gap, the “digital divide” is still a prevalent concern (Vigdor et al., 2014). These issues were exacerbated during the COVID-19 pandemic, as the shift to online learning posed a significant barrier to success for students who may not have equal access to technology (Hassan & Daniel, 2020). Friedman et al. (2021) found that 10.1% of children were without adequate access to internet and electronic devices in 2020. Of those, the largest group of individuals who reported insufficient technological access were Black children whose parents had less than a high school education.

During the height of the pandemic, non-White students faced many barriers while navigating the educational system (Allen et al., 2020). Hispanic students were more likely to engage in remote learning and struggle with technical issues in K-12 education (Dorn et al., 2020a, Dorn et al., 2020b). Both Black and Hispanic students are more likely to live in poverty (Gaitan, 2013) and struggle to participate in online instruction, which led to learning loss during the pandemic (Dorn et al., 2020a, Dorn et al., 2020b).

2.6.2. Online learning in college

Hispanic and Black students share challenges within higher education as well: compared to White students, Black and Hispanic students are 2.7 and 1.8 times less likely, correspondingly, to own a laptop (Reisdorf et al., 2020). Students among minority demographic groups report lower satisfaction with online education overall and greater challenges with peer and professor interactions (Arbelo et al., 2019; Ke & Kwak, 2013). When on camera in sync classes, Hispanic students are more concerned about their appearance (Castelli & Sarvary, 2021). However, other studies find that Hispanic students appreciate the convenience and flexibility of asynchronous online classes, citing the need to juggle multiple roles/responsibilities (Means & Neisler, 2021; Kumi-Yeboah, 2018) while also emphasizing self-motivation and individual agency associated with online learning (Kauffman, 2015; Arbelo et al., 2019).

Perceived levels of racism and discrimination amongst non-White students can also have a detrimental effect on student learning and educational achievement (Brown et al., 2005; Johnson et al., 2007). Perceived lack of diversity on campus may also threaten some students’ sense of belonging (Gao & Liu, 2021) and lead minority students to prefer online education, which we assess in our study.

The focus of our study on these demographic groups is particularly important given recent collegiate enrollment trends. In 2021, the Hispanic college student population increased by more than 40% (Krogstad, Passel, & Noe-Bustmante, 2020), while the Black college student population has been steadily decreasing (Miller, 2020). With a booming Hispanic and dwindling Black student population, we need to better understand the perspectives and preferences of these groups regarding online learning.

2.7. Experiences of international students

Research that draws a comparative analysis between international and domestic college students in the U.S. is limited (Han et al., 2022; Li & Che, 2022), though there are some studies that have examined this aspect outside of the U.S. (Cranfield et al., 2021; Demuyakor, 2020). For example, Demuyakor (2020) examined the perception of online learning in China by international (Ghanaian) students: most were pleased with the effectiveness and credibility of online course content, but they mentioned two drawbacks: the lack of social interaction and high cost of internet data.

Han et al. (2022) interviewed 7 international graduate students in U.S. institutions. While professors helped students cope with the pandemic transition, students felt professors were more accessible in person than online. Lack of interactions with peers was also taxing. However, online discussion mode was preferred over in-person one due to ease and comfort (Han et al., 2022). Relatedly, Li and Che (2022) found international students used cameras less often in online classes, especially in classes of larger sizes.

The paucity of studies comparing international and domestic students' perceptions of online learning creates a serious knowledge gap, while the enrollment of international students in U.S. institutions of higher education has been steadily growing in the past few decades (FWD.us, 2021). Thus, it is important to examine the international students’ views and experiences with online learning, which is one of our study aims.

3. Current study: data and methods

Considering the above-noted limitations in previous research, our study helps remedy many of these issues and fill the knowledge gaps by (a) expanding the sample size and versatility to allow comparisons among various demographic groups of students, and (b) timing this study to coincide with the post-pandemic return to normal operations on U.S. college campuses.

3.1. Research questions

Our main goal in this exploratory study is to better understand university students’ experiences and perceptions of online learning (both synchronous and asynchronous) when compared with traditional classes. To make the distinction between different modes of course delivery easier, we use the term “Zoom classes” when referring to live synchronous online classes that students attend virtually at the assigned class time, “async” to indicate online asynchronous courses that have no live class sessions, and “on-campus” for traditional in-person classes.

More specific research questions of our study include comparing different demographic groups of students:

• •What mode of course delivery (Zoom, async, or on-campus) do students prefer and why?
• •What do students see as the main advantages and disadvantages of online learning compared to traditional, on-campus learning?
• •What are the main reasons for students to keep their camera off during Zoom classes?

3.2. Survey development and administration

To investigate these questions, we conducted an anonymous online mixed-methods survey of students at a mid-size private university in the Northeast United States that has a diverse student body in the Spring 2022 semester. No exclusion criteria prevented any students at the University from participating unless they indicated that they were not at least 18 years of age. This study was approved by the University's Institutional Review Board and all participants gave informed consent prior to gaining access to the survey questions.1 Respondents were then asked to reflect on their experiences and share their views regarding advantages and disadvantages of online learning.2 The survey included yes-no, multiple-choice, and Likert-scale questions, as well as free-response questions about students' perceptions of online learning – see the full survey at tinyurl.com/tostosurvey. In the survey, we have partially adapted questions about online learning from Baczek et al. (2021), Castelli and Sarvary (2021), Gillis and Krull (2020), and Murphy et al. (2020) as we found these studies to align closely with our own goals of understanding the effects of recent changes to learning format at the university level.

3.3. Sample characteristics

Over 900 students (out of about 6000) provided responses to the survey. This is a relatively high response rate – about 15% – for this type of online surveys (Pyrczak & Tcherni-Buzzeo, 2018). At the same time, only 690 of 914 survey respondents (just over 75%) also answered the demographic portion of the survey. Demographic characteristics of these respondents are presented in Table 1 , including information on key variables of interest in our study – gender, age, race, and ethnicity – separately for international (n = 148) and domestic (n = 542) students.

Table 1.

Descriptive statistics of the sample, with breakdown by domestic/international student status.

	Overall		Domestic students		International students
	N	%	N	%	N	%
Gender
Male	235	34.1%	157	29.0%	78	52.7%
Female	419	60.8%	352	65.1%	67	45.3%
Transgender	6	0.9%	6	1.1%	0	0.0%
Nonbinary	21	3.0%	18	3.3%	3	2.0%
Other	8	1.2%	8	1.5%	0	0.0%
Race
Native American	2	0.3%	2	0.4%	0	0.0%
Asian	132	19.3%	22	4.1%	110	74.3%
Black/African American	71	10.4%	47	8.8%	24	16.2%
White/Caucasian	434	63.5%	431	80.6%	3	2.0%
Other	43	6.3%	32	6.0%	11	7.4%
Ethnicity
Hispanic/Latinx	98	14.3%	92	17.0%	6	4.1%
Non-Hispanic	589	85.7%	449	83.0%	140	95.9%
Race/ethnicity
Non-Hispanic White	375	54.9%	373	69.7%	2	1.4%
Non-Hispanic Black	62	9.1%	39	7.3%	23	15.5%
Age
18–21	348	51.3%	338	63.2%	10	6.9%
22–30	237	34.9%	126	23.6%	11	77.1%
31+	94	13.8%	71	13.3%	23	16.0%
Class level
Freshman	79	11.5%	68	12.6%	11	7.4%
Sophomore	108	15.7%	105	19.4%	3	2.0%
Junior	116	16.8%	115	21.3%	1	0.7%
Senior	112	16.3%	110	20.3%	2	1.4%
Master's	260	37.7%	130	24.0%	130	87.8%
Ph.D.	14	2.0%	13	2.4%	1	0.7%
Residential Status
On-campus	241	34.9%	233	43.0%	8	5.4%
Commuter	355	51.4%	220	40.6%	135	91.2%
Fully online	94	13.6%	89	16.4%	5	3.4%
Total	690	100%	542	78.6%	148	21.4%

Note: Most demographic variables have some missing data.

3.4. Analytical approach: mixed methods

A combination of closed-ended and open-ended questions in the survey provided opportunities for employing mixed methods research.

3.4.1. Quantitative analysis

Statistical analyses are based on students' answers to closed-ended questions regarding students’ preferences of the type of class (on-campus, Zoom, or async), keeping their camera off during Zoom classes, and their views on advantages and disadvantages of online learning. When comparing categorical responses across demographic groups, we used chi-square tests to determine whether the differences are statistically significant (the results of chi-square tests are listed underneath each table, Table 2, Table 3, Table 4, Table 5 ).

Table 2.

Class type preference by demographic group.

	On-campus/In-Person		Zoom/Online Synchronous		Independent/Online Asynchronous
	N	%	N	%	N	%
Gender
Male	131	66.5%	38	19.3%	34	17.7%
Female	231	60.9%	50	13.4%	110	29.0%
Race/Ethnicity
Black non-Hispanic	30	58.8%	16	31.4%	8	16.0%
Hispanic	52	56.5%	13	14.1%	29	32.2%
White non-Hispanic	234	65.9%	30	8.6%	95	26.8%
Student status
Domestic	322	63.5%	53	10.6%	140	27.7%
International	60	58.3%	39	37.5%	15	15.0%

Note: Group differences statistically significant at the 0.05 level are marked in bold. Below are the chi-square results for specific demographic group comparisons.

Male vs female – preference for async classes: χ² (df = 1, N = 571) = 8.65, p = .003.

Race/ethnicity – preference for Zoom classes: χ² (df = 2, N = 493) = 22.47, p < .001.

Domestic vs international – preference for Zoom classes: χ² (df = 1, N = 604) = 48.25, p < .001.

Domestic vs international – preference for async classes: χ² (df = 1, N = 605) = 7.09, p = .008.

Table 3.

Gender preferences in online learning.

	Male		Female
	N	%	N	%
Advantages of Online Learning
Access to Online Materials	125	53.0%	193	46.1%
Learning at Own Pace	157	66.5%	271	64.7%
Ability to Stay at Home	161	68.2%	307	73.3%
Class interactivity	13	5.5%	19	4.5%
Ability to Record a Meeting	87	36.9%	132	31.5%
Comfortable Surroundings	116	49.2%	254	60.6%
Disadvantages of Online Learning
Reduced Teacher Interaction	166	70.3%	289	69.0%
Technical Problems	96	40.7%	176	42.0%
Lack of Peer Interaction	155	65.7%	245	58.5%
Poor Learning Conditions at Home	37	15.7%	65	13.4%
Lack of Self-Discipline	82	34.7%	171	40.8%
Social Isolation	99	42.0%	173	41.3%
Ever Kept Camera Off?
Yes	168	71.5%	350	83.9%
Reasons for Keeping Cameras Off
Self-Consciousness	44	27.0%	169	49.7%
Social conformity	40	24.4%	80	23.5%
Distractions/Multi-Tasking	46	28.0%	126	37.0%
Technical Issues	18	11.0%	11	3.2%

Note: Group differences statistically significant at the 0.05 level are marked in bold. Below are the chi-square results for male v. female group comparisons.

Advantage – Comfortable surroundings: χ² (df = 1, N = 655) = 8.08, p = .004.

Disadvantage – Distractions/multi-tasking: χ² (df = 1, N = 505) = 3.91, p = .05.

Disadvantage – Self-consciousness: χ² (df = 1, N = 503) = 23.28, p < .001.

Disadvantage – Technical issues: χ² (df = 1, N = 503) = 12.36, p < .001.

Turning off camera during class: χ² (df = 1, N = 652) = 14.25, p = .001.

Camera off due to self-consciousness: χ² (df = 1, N = 503) = 23.28, p < .001.

Camera off due to distractions/multi-tasking: χ² (df = 1, N = 503) = 3.91, p = .05.

Camera off due to technical issues: χ² (df = 1, N = 503) = 12.36, p < .001.

Table 4.

Racial and ethnic differences in online learning.

	Black Non-Hispanic		Hispanic		White Non-Hispanic
	N	%	N	%	N	%
Advantages of Online Learning
Access to Online Materials	31	50.0%	56	57.1%	160	42.6%
Learning at Own Pace	47	75.8%	62	63.3%	229	60.9%
Ability to Stay at Home	45	72.6%	74	75.5%	293	77.9%
Class Interactivity	2	3.2%	8	8.2%	6	1.6%
Ability to Record a Meeting	28	45.2%	32	32.7%	98	26.1%
Comfortable Surroundings	28	45.2%	55	56.1%	223	59.3%
Disadvantages of Online Learning
Reduced Interaction with the Teacher	38	61.3%	61	62.2%	288	76.6%
Technical Problems	30	48.4%	40	40.8%	134	35.6%
Lack of Peer Interaction	33	53.2%	51	52.0%	230	61.2%
Poor Learning Conditions at Home	8	12.9%	22	22.4%	55	14.6%
Lack of Self-Discipline	29	46.8%	46	46.9%	150	39.9%
Social Isolation	24	38.7%	38	38.8%	156	41.5%
Ever Kept Camera Off?
Yes	48	77.4%	81	82.7%	327	87.7%
Reasons for Keeping Cameras Off
Self-Consciousness	21	43.8%	47	59.5%	142	44.7%
Social Conformity	12	25.0%	9	11.4%	83	26.0%
Distractions/Multi-Tasking	16	33.3%	28	35.4%	107	33.5%
Technical Issues	1	2.1%	3	3.8%	17	5.3%

Note: Group differences statistically significant at the 0.05 level are marked in bold. Below are the chi-square results for race/ethnicity group comparisons.

Advantage – Access to online materials: χ² (df = 2, N = 536) = 7.09, p = .03.

Advantage – Ability to record a meeting: χ² (df = 2, N = 536) = 9.92, p = .007.

Advantage – Class interactivity: χ² (df = 2, N = 536) = 11.59, p = .003.

Disadvantage – Reduced interaction with the teacher: χ² (df = 2, N = 536) = 12.14, p = .002.

Camera off due to social conformity: χ² (df = 2, N = 446) = 7.66, p = .02.

Table 5.

Preferences of domestic versus international students in online learning.

	Domestic		International
	N	%	N	%
Advantages of Online Learning
Access to Online Materials	248	45.7%	84	56.8%
Learning at Own Pace	329	60.6%	117	79.1%
Ability to Stay at Home	421	77.5%	78	52.7%
Class Interactivity	14	2.6%	19	12.8%
Ability to Record a Meeting	156	28.7%	75	50.7%
Comfortable Surroundings	319	58.7%	76	51.4%
Disadvantages of Online Learning
Reduced Teacher Interaction	397	73.1%	83	56.1%
Technical Problems	209	38.5%	76	51.4%
Lack of Peer Interaction	313	57.6%	104	70.3%
Poor Learning Conditions at Home	87	16.0%	15	10.1%
Lack of Self-Discipline	228	42.0%	44	29.7%
Social Isolation	221	40.7%	68	45.9%
Ever Kept Camera Off?
Yes	459	85.2%	85	57.4%
Reasons for Keeping Cameras Off
Self-Consciousness	216	48.4%	17	20.5%
Social conformity	104	23.3%	17	20.2%
Distractions/Multi-Tasking	150	33.6%	31	36.9%
Technical Issues	22	4.9%	8	9.6%

Note: Group differences statistically significant at the 0.05 level are marked in bold. Below are the chi-square results for domestic vs international group comparisons.

Advantage – Access to online materials: χ² (df = 1, N = 691) = 5.72, p = .05.

Advantage – Learning at one's own pace: χ² (df = 1, N = 691) = 17.33, p < .001.

Advantage – Ability to stay home: χ² (df = 1, N = 691) = 35.73, p < .001.

Advantage – Class interactivity: χ² (df = 1, N = 691) = 26.92, p < .001.

Advantage – Ability to record a meeting: χ² (df = 1, N = 691) = 25.17, p < .001.

Disadvantage – Reduced teacher interaction: χ² (df = 1, N = 691) = 15.90, p < .001.

Disadvantage – Technical problems: χ² (df = 1, N = 691) = 7.94, p = .005.

Disadvantage – Lack of peer interaction: χ² (df = 1, N = 691) = 7.75, p = .005.

Disadvantage – Lack of self-discipline: χ² (df = 1, N = 691) = 7.32, p = .007.

Turning off camera during class: χ² (df = 1, N = 687) = 54.15, p < .001.

Camera off due to self-consciousness: χ² (df = 1, N = 529) = 22.18, p < .001.

3.4.2. Qualitative analysis: thematic coding

Analysis of qualitative responses to open-ended questions was conducted by all authors. Using principles rooted in grounded theory (Charmaz, 2002; Saldana, 2013; Emerson, 2001), these question responses were coded to uncover participant meaning and shared experiences among university students (Charmaz, 2012; Lofland et al., 2006). The qualitative questions used for coding asked students why they kept the camera off and what are the reasons for their class type preferences (on-campus, Zoom, or async).

For each question, two different authors were assigned who independently coded all responses into broad themes that emerged organically from the data. Once these themes were established, the full group of researchers met to compare findings and finalize the themes for each question. Next, a third researcher independently coded a set of 100 sequential responses to the identified themes to establish the inter-rater reliability and ensure that perceived meanings derived from qualitative responses were accurate (Armstrong et al., 1997; Belotto, 2018). Any discrepancies in the authors' coding were reconciled by the full group of researchers to arrive at the final set of codes for each student's answer. The coded themes are summarized in an online appendix at tinyurl.com/tostocodes.

3.4.3. Integration of methods

Mixed methods study design allows combining the ‘bird's eye view’ of the preferences of large groups of students in different demographic categories with an up-close and personal look into students' feelings and thoughts on the matter. In the Results section, you can see an integration of statistical analyses with students' quotes, which enhances our understanding of reality and offers a paradigm superior to single-method studies (Johnson & Onwuegbuzie, 2004; Pyrczak & Tcherni-Buzzeo, 2018).

4. Results

Among about 690 respondents who provided their demographic information, some interesting differences emerged within the demographic categories of gender, race/ethnicity, and domestic vs international student status (see Table 2 on their class type preferences), and we describe these findings in more detail below.

4.1. Gender differences

Of the 690 respondents who did indicate their gender, about 34% identified as male and 61% identified a female (Table 1). About 5% identified as transgender, non-binary or “other”. Due to the variability and low number of respondents within this latter category, the sections below focus on gender differences between males and females.

4.1.1. Gendered differences in class preference

In their preferences of synchronous online (Zoom), asynchronous online (async), and on-campus courses (see Table 2), both genders rank on-campus courses similarly, as their top choice (66.5% of males and 60.9% of females). Their preferences for Zoom classes differ, but the difference does not reach statistical significance (19.3% of males vs 13.4% of females liked Zoom classes best). Where gendered preferences differ substantially is async classes (17.7% of males vs 29% of females ranked async as their first choice, the difference is statistically significant).

Ability to stay home was selected as the main advantage of online learning by most participants (73.3% of females and 68.2% of males), but gender differences become statistically significant only when respondents are asked specifically about comfortable surroundings as an advantage of online learning (60.6% of women chose this option compared to 49.2% of men).3 The visual representation of the key gender differences in online class preferences is presented in Fig. 1 (a) and (b). “Zoom synchronous is my favorite type of class due to the convenience and interactions. You still get the classroom experience from the comfort of your own home” (Hispanic, female, 26 years old). A close second advantage selected by almost two thirds of respondents of both genders was learning at their own pace (selected by 66.5% of males and 64.7% of females). As one participant noted, “Asynchronous learning allows for me to work at my own pace and have convenience to do schoolwork wherever I want” (White, male, 26 years old).

Fig. 1.

Key gender differences in online learning preferences

(a) preference for independent (online async) classes

(b) advantage of online learning: comfortable surroundings

(c) ever kept the camera off during zoom classes

(d) reasons for keeping the camera off during zoom classes.

Very few respondents listed class interactivity as an advantage of online learning (5.5% of men and 4.5% of women), and many of them consistently acknowledged the ability to interact with peers as a reason for why they ranked synchronous and asynchronous classes differently. “I like getting to ask questions in real-time with in-person classes. I also like seeing my classmates' faces and actually talking to them” (Hispanic, female, 22 years old). When probing the issue of interactivity further, over two thirds of both genders reported reduced teacher interaction as the main disadvantage of online education, with peer interaction being a close second. One participant clarified: “On-campus is more interactive and effective because it's one on one, [an] opportunity to relate with the professor and other students” (Black, male, 32 years old).

4.1.2. Gendered differences in camera use

After students were asked if they ever kept their camera off during Zoom classes, those who said yes (over 70% of respondents) were prompted to explain why. Their responses were then coded by the researchers according to several non-exclusive themes: self-consciousness, distractions/multi-tasking, social conformity, and technological issues.

Women were significantly more likely to keep their camera off during Zoom classes (83.9%) than men (71.5%), and females’ explanations for this often included self-consciousness as a reason to avoid camera use (49.7%) compared to only 27% of men. As one participant indicated, “I am uncomfortable with peers seeing my home environment, I do not have a private space in my home” (White, female, 21 years old). The main reason for men to turn off the camera during Zoom classes had to do with distractions or wanting to multi-task without being noticed (28%). As one participant noted, “I left my camera off due to being at work during some of the class times overlapping with my job” (White, male, 20 years old). Importantly, an even higher proportion of women (37%) listed distractions/multi-tasking reasons as well. Men and women did not significantly differ in their reporting of social conformity (aligning personal camera use with what the rest of the class was doing), but contrary to common gender stereotypes, males were significantly more likely to cite technical reasons (11%) compared to women (3.2%). It is possible that men used technical issues as a more acceptable justification for staying off camera rather than giving more personal reasons related to self-consciousness or privacy. Key gender differences in camera use and reasons for turning it off during online classes are visualized in Fig. 1(c) and (d).

4.2. Racial and ethnic differences

Among the respondents who provided their demographic information, 375 (54.9%) were non-Hispanic White, 62 (9.1%) were non-Hispanic Black, and 98 (14.3%) identified as Hispanic (see Table 1). For convenience, hereinafter we will use “Black” and “White” in the text when referring to “non-Hispanic Black” and “non-Hispanic White”. Even though a large percentage of students identified as Asian (132, or 19.3%), most of them were international students (see Table 1), with origins as wide-ranging as Taiwan, India, Saudi Arabia, and Indonesia, among many other Asian countries. Thus, analyzing these students as a uniform group would have been misleading. The sections below present analyses for Black and Hispanic students, especially in comparison with White students.

4.2.1. Black students’ perspectives

Among the survey respondents, Black students were twice as likely to prefer Zoom classes as their first choice (31.4%) compared to Hispanic (14.1%) or White students (8.6%) – see Table 2. As the main advantage of online learning (see Table 4), Black students especially note the convenience of learning at their own pace (75.8% vs 63.3% of Hispanics and 60.9% of Whites, just below statistical significance). Qualitative analysis revealed this preference to be related to participation in extracurriculars like sports and internships. “Sometimes internships, jobs, or home life come in the way, and it is not as easy to travel to campus.” (Black, female, 21 years old).

Additionally, Black students are significantly more likely to highlight a key feature of the software used for online synchronous classes: the ability to record and upload any class meeting to accessible educational platforms. This allows students to rewatch certain parts of a lecture, pause to take additional notes, or watch entire lectures missed due to athletics, sickness, or another circumstance. “I prefer synchronous classes because most professors will record the course if someone were to miss it, and as a student-athlete, it's hard to avoid missing one or two classes because of games. The synchronous classes also give me the ability to attend class almost anywhere … with on-campus courses, it is easier to fall behind if you miss a class because it is not recorded …” (Black, female, 19 years old). Other students note their preference for Zoom as “getting to class can be difficult” (Black female, 19 years old) or describe their preference for synchronous courses because of the ability to “learn in the comfort of [their] home” (Black, male, 23 years old).

Although Black students like Zoom courses, they seem to dislike asynchronous ones (16% of Black versus 32.2% of Hispanic and 26.8% of White students choose async as #1) – see Table 2. At the same time, these differences do not reach statistical significance at the conventional p < .05 level, and many of the disadvantages of online learning are judged by these racial/ethnic groups similarly (see Table 4). For example, self-discipline necessary for success in async courses is identified as a problem for online learning by almost 47% of Black and Hispanic students and by 39.9% of White students (see Table 4). “For independent work, students must have discipline to complete the work online” (Black, female, 40 years old). Additionally, some students mentioned that asynchronous courses unintentionally encourage students to “just get the work done” for the grade, rather than for education. “Asynchronous is last because I tend to not learn very often … and just complete assignments just to get them done.” (Black, female, 19 years old). Interestingly, the only statistically significant difference among racial/ethnic groups in the disadvantages of online learning was reduced interaction with the teacher: White students (76.6%) are much more likely than Black (61.3%) or Hispanic (62.2%) students to be bothered by this. “Nothing compares to in-person learning; the social environment, the increased opportunity to ask questions or have your confusion noticed by professors … is best in person.” (White, female, 21 years old).

4.2.2. Hispanic students’ perspectives

Unlike Black students, Hispanic students like online asynchronous classes (32.2% of Hispanics rated online asynchronous as their preferred class type, compared to 16% of Black students – see Table 2). Even though these differences did not reach statistical significance, qualitative analyses revealed that, for many Hispanic students, asynchronous classes stood out for convenience-associated factors. “I'm remote and 2 h away from school so having online classes is a cheaper and better option” (Hispanic, female, 22 years old). Juggling multiple responsibilities also seemed to be a common reason for preferring asynchronous classes among Hispanic students, and this extended beyond time needed for extracurricular activities: “As a student and full-time mental health worker, independent work is more sustainable and better suited to my needs” (Hispanic, female, 27 years old). Another student notes: “As a working professional, I need flexibility in the hours I choose to do class and assignments” (White, female, 19 years old).

Other students put an emphasis on individualism for preferring asynchronous learning. “It provides additional time for me to further research if I do not understand something and if I am reading a textbook, I can re-read it as many times as I want without the pressure of having to hurry because of a timeline” (Hispanic, female, 41 years old). Agency over one's learning is a common theme. “At times where I feel more productive, I can do my work at that time rather than having a scheduled meeting time” (Hispanic, male, 23 years old). Interestingly, the individualistic aspect is also reflected in Hispanic students being less likely to worry about social conformity when turning their camera off during Zoom classes: only 11.4% of Hispanics mentioned social conformity versus 25% of Blacks and 26% of Whites, and this difference is statistically significant (see Table 4). We visualized the differences among racial/ethnic categories in online learning preferences in Fig. 2 .

Fig. 2.

Key racial/ethnic differences in online learning preferences

(a) preference for zoom, or online sync, classes

(b) advantages of online learning

(c) social conformity as a reason to keep the camera off during zoom classes.

Continuing with the qualitative analyses of student responses, we see other reasons for Hispanic students to prefer asynchronous classes:

• •“It works better for my needs rather than attending physical classes. I'm also 30 years old, so being surrounded by 18-year-olds in school feels really weird” (Female, 29 years old).
• •“I love [asynchronous] classes because I hate lectures … In the other classes you are forced to interact with professors … Most classes you are talked at and then you leave” (Female, 20 years old).

4.3. Differences between international and domestic students

We see drastically different preferences of domestic and international students for Zoom and asynchronous online class: international students prefer Zoom classes almost 4 times more (37.5%) than domestic students (10.6%), while domestic students prefer async classes almost twice as often (27.7%) as international students (15%), and these differences are statistically significant (see Table 2). However, international students’ preferences in the mode of course delivery (on-campus, Zoom, or async) are difficult to interpret properly since their choice of online classes may be severely limited due to U.S. visa restrictions limiting the number of online (async) courses for international students within the U.S., whereas domestic students are able to enroll in fully online courses with no restrictions in the U.S.

Additional information is provided in Table 5 on how differently international and domestic students perceive advantages and disadvantages of online learning (almost all differences are statistically significant). 79.1% of international students chose learning at one's own pace as the top advantage of online learning (vs 60.6% of domestic students), while domestic students see the ability to stay home as the main advantage of online learning (77.5% of domestic vs 52.7% of international). International students are almost twice as likely as domestic students to emphasize the ability to record a meeting (50.7% vs 28.7%) and almost 5 times more likely to underscore class interactivity (12.8% vs 2.6%) as benefits of online learning.

Among the disadvantages of online learning, lack of peer interaction is the main problem for international students (70.3%) while only 57.6% of the domestic group named it as a problem. Conversely, reduced interaction with a teacher ranked high among domestic students (73.1%) while being less of a problem for international students (56.1%). International students are much less likely to worry about self-discipline required for online learning (29.7% vs 42% of domestic students).

When using camera during Zoom classes, international students are less likely to turn their camera off than domestic students (57.4% vs 85.2%). They are also significantly less likely to mention self-consciousness or privacy as a reason for turning of the camera (20.5% of international vs 48.4% of domestic students). Almost half of domestic students expressed concerns about appearance or privacy. “I am self-conscious of my image and do not want to show myself when I am home” (White, non-binary, 22 years old) and “I don't want people seeing me in my living area. It is a private space that I share with people who I want to share it with, not random classmates …” (White, male, 21 years old).

5. Discussion

Our study examines the preferences and opinions of college students regarding various aspects of online learning, such as preferred class type (on-campus, Zoom, and asynchronous online), decisions to turn the camera off during Zoom classes and reasons behind these decisions, and views on advantages and disadvantages of online learning. Having a large and versatile sample from a mid-size university in the Northeast answering our anonymous survey allows us to use mixed methods to compare students of various demographic groups in their views and preferences.

We find that men and women have different preferences for asynchronous online classes, 28% of women but only 17% of men ranking it as their first choice. This is likely related to the camera use in synchronous class: women prefer to turn the camera off during online classes, mentioning self-consciousness as a common reason. In other ways, both genders are very similar in their views of online learning.

Class type preferences differ significantly among Black, White, and Hispanic students, with Black students preferring Zoom classes at a higher rate (31.4%) than online asynchronous ones (16%), and Hispanic students preferring asynchronous courses (32.2%) to Zoom classes (14.1%). These findings echo existing research, which finds that Hispanic students much more often than non-Hispanics cite challenges managing responsibilities at home with their educational obligations and find online options more convenient (Means & Neisler, 2021; Kumi-Yeboah, 2018). Overall, flexibility and convenience of online learning is a prominent factor identified in previous studies to be important for students (Alexander et al., 2009; Muthuprasad et al., 2021; Nollenberger, 2015), and our study adds to this body of evidence.

Lack of interaction with professors was the only significant difference observed among the racial and ethnic groups, with 76.6% of White students noting this as a disadvantage of online learning compared to 62.2% of Hispanic students and 61.3% of Black students. In fact, perceived lack of interaction or sense of community among online students has been noted as one of the main online learning challenges by previous researchers (De Metz & Bezuidenhout, 2018; Nollenberger, 2015; Song et al., 2004).

International students' preferences in terms of the mode of course delivery may be limited due to visa restrictions, so the exact interpretation of these differences between international and domestic students is unclear. In terms of online learning advantages, international students prefer learning at their own pace (79.1%) and the ability to record a meeting (50.7%), while domestic students prefer being able to stay at home (77.5%) as the main advantage of online learning. The main disadvantage of online learning for international students is lack of peer interaction (70%) while for domestic students, it is reduced interaction with the teacher (73%). It also emerged in our study that international students are much less likely to turn off their camera during Zoom classes or to mention self-consciousness/privacy as a reason for doing so.

Our findings echo the results of some prior research, although studies that draw comparisons between international and domestic students are scarce (Han et al., 2022; Li & Che, 2022). Previous studies reported that international students struggled with a lack of social interaction generally (Demuyakor, 2020) and peer interaction more specifically (Han et al., 2022). Another study, though conducted outside the U.S., found that students (in Wales, South Africa, and Hungary) reported a lack of peer interaction as a significant disadvantage of online learning during the COVID-19 pandemic (Cranfield et al., 2021).

5.1. Study limitations and directions of future research

Since our survey relies on a convenience sample, the question naturally arises whether the students who answered the survey are different from those did not (self-selection bias). A relatively high response rate and the fact that the demographics of our sample closely resemble the demographics of the university overall assuage these concerns. Like any self-report survey, our study relies on people's own opinions about themselves, and these opinions may be distorted by social desirability. However, this concern is minimized by the anonymity of the survey. In addition, the survey itself was not empirically validated in either the prior studies from which questions were adapted or the current work. While this poses a minor limitation, the questions were adapted to apply specifically to the potential experiences of students at the studied University, making our concern over this issue quite minimal.

It is possible that the experiences of students are unique to the university from which the sample was drawn since the faculty at this university had extensive preparation and training for online learning and course hybridization well before the start of the COVID-19 pandemic (so the transition to online learning likely went more smoothly than it could have gone otherwise). However, enough time has passed from the initial hectic stage of the pandemic switch to online learning, and extensive training in online delivery methods has been conducted in many colleges since then. Future studies at other universities and colleges will be able to answer the questions of generalizability more fully.

Another limitation is that we have not been able to meaningfully incorporate some gender categories (for example, non-binary and transgender students) and races (for example, Asian students). This is an important direction of future research with more targeted samples. While exploratory analyses were conducted in these areas, it was also beyond the scope of this single paper to review findings related to class differences, experiences of graduate versus undergraduates, and other characteristics that may affect these class-based experiences. However, we believe that the gendered, racial, and ethnic differences explored here provide particularly salient and timely consideration of the needs of today's student populations.

5.2. Implications for practice

The results of our study, even taking limitations into account, are valuable for informing not only future research but also educational practices at the institutions of higher education. Depending on the demographic makeup of the student body or students taking specific courses, college administrators and faculty can adjust their approaches to meet the preferences and constraints of their students and more accurately anticipate student demand for specific online course modalities.

At the same time, it may be premature to make specific or far-reaching recommendations based on the results of one study. It would be important to first see some replications and extensions of this research on different student populations, to ensure that the demographic differences among college students in their preferences for online learning, which we have uncovered in our study, hold in different contexts. Moreover, we encourage faculty to use the survey questions listed at tinyurl.com/tostosurvey to poll your students on those aspects of course delivery and educational practice that you are considering changing or updating.

6. Conclusions

This is one of the first studies that included a comprehensive assessment of student views on various aspects of online learning and comparisons across several important demographic groups of students: based on gender, race/ethnicity, and international versus domestic student status, while using a large and diverse sample of students at a mid-size university in Northeastern United States. The insights gained in this study are an important post-pandemic update on online learning research and should inform future studies and help make educational practices more sensitive to the needs of specific demographic groups of students.

Credit Author Statement:

Online learning in the wake of the COVID-19 pandemic: Mixed-methods analysis of student views by demographic group

• Samantha A. Tosto: conceptualization, methodology, formal analysis, investigation, writing - original draft, writing - review & editing.

• Jehad Alyahya: conceptualization, methodology, formal analysis, investigation, writing - original draft, writing - review & editing.

• Victoria Espinoza: conceptualization, methodology, formal analysis, investigation, writing - original draft, writing - review & editing.

• Kylie McCarthy: conceptualization, methodology, formal analysis, investigation, writing- original draft, writing- review & editing.

• Maria Tcherni-Buzzeo, PhD: conceptualization, methodology, formal analysis, investigation, writing - original draft, writing - review & editing, supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.