Adaptability to Online Learning: Differences Across Types of Students and Academic Subject Areas Di Xu and Shanna Smith Jaggars February 2013 CCRC Working Paper No.. Abstract Using a da
Introduction
One of the most pronounced trends in higher education over the last decade has been a strong growth in distance education through online coursework (Allen & Seaman,
The growth of online distance education has significantly broadened learning opportunities, especially for nontraditional students who often juggle employment and family responsibilities that hinder their ability to attend conventional in-person classes This trend has led to a notable surge in online learning enrollments, particularly at two-year colleges.
2008), where a large proportion of the population are nontraditional students (Kleinman
Many college students, having primarily experienced face-to-face education, face challenges when adapting to online coursework Research comparing student performance in online versus traditional settings has yielded mixed results, with some studies indicating positive outcomes for online learning while others report negative findings (e.g., Bernard et al., 2004; Zhao et al., 2005; Sitzmann et al., 2006; Jahng et al., 2007; U.S Department of Education, 2010).
The variation in study results may stem from differences in student populations and course contexts Students with greater exposure to technology or training in time-management and self-directed learning tend to adapt more easily to online learning Additionally, certain academic subjects may facilitate higher-quality online learning experiences, thereby better supporting students in their adaptation efforts.
The National Center for Education Statistics (2002) identifies a nontraditional student as an individual who exhibits one or more of the following seven risk factors: part-time attendance, full-time employment, delayed enrollment in postsecondary education, financial independence, having dependents, and being a single parent.
Students without a high school diploma may face unique challenges in adapting to online coursework Research indicates that various student characteristics, including gender, age, ethnicity, and prior academic performance, significantly influence online learning outcomes Understanding these factors is essential for improving the online learning experience for diverse student populations.
In terms of gender, while several studies have found no differences between males and females in terms of their learning outcomes in online courses (e.g., Astleitner
Research indicates that women tend to outperform men in various academic settings, particularly in online courses Studies by Chyung (2001), Gunn et al (2003), Price (2006), Rovai and Baker (2005), Sullivan (2001), and Taplin and Jegede (2001) support this finding, highlighting the significant advantages women demonstrate in these environments.
McSporran and Young (2001) analyzed course observation and student survey data, finding that female participants demonstrated higher motivation, better online communication skills, and more effective learning scheduling In contrast, male participants accessed fewer course website pages and discussion forum posts, exhibited poorer time management skills, and often displayed overconfidence in their ability to complete learning tasks and assignments.
Research indicates that women often excel in online courses, reflecting their overall stronger educational performance in various settings Notably, women have a higher likelihood of graduating from high school, as highlighted by studies from Swanson (2004) and Heckman & LaFontaine.
Research indicates that women in college are more likely to earn degrees than men A pertinent question arises: Are women better at adapting to online courses compared to men? This leads to an inquiry about whether the performance gap between genders is wider or narrower in online learning environments versus traditional classrooms However, the impact of gender on students' adaptability to online learning remains largely unexamined.
Research indicates that Black and Hispanic students often underperform in online courses compared to their White peers (Newell, 2007) This trend may be linked to their overall lower college performance, which is influenced by systemic disadvantages in the quality of their primary and secondary education (Feldman, 1993; Allen, 1997; DuBrock, 2000).
Research has yet to investigate the impact of ethnicity on student adaptability to online courses, particularly regarding whether the performance gap between ethnic minorities and White students is worsened by online learning Concerns have been raised that online education may exacerbate the postsecondary access gap for students of color due to disparities in home technology access For instance, in 2009, only 52% of African Americans and 47% of Hispanics had high-speed Internet at home, which could hinder their performance in online courses.
In terms of student age, some studies have found no relationship between age and satisfaction or performance in online learning (e.g., Biner, Summers, Dean, Bink,
Anderson, & Gelder, 1996; Osborn, 2001; Wang & Newlin, 2002; Willging & Johnson,
Research indicates that older students are more likely to successfully complete online courses compared to younger students For instance, a study by Dille and Mezack (1991) found that the average age of successful online learners was 28, while non-successful learners averaged 25 Colorado and Eberle (2010) suggest that this success may stem from older students' enhanced skills in rehearsal, elaboration, critical thinking, and metacognitive self-regulation, all of which are crucial for excelling in online education.
Older students may excel in online courses despite generally poorer academic outcomes, potentially due to family and work commitments that hinder their success in traditional settings Research indicates that older community college students are less likely to earn credentials or transfer to four-year universities If these students adapt well to online learning, promoting this mode of education could enhance their access to postsecondary opportunities and provide them with a significant academic advantage.
In contrast to the large volumes of studies examining gender, ethnicity, and age as predictors of online success, very few studies (e.g., Hoskins & Hooff, 2005; Figlio, Rush,
Research indicates that students with weaker academic backgrounds often struggle with time management and self-directed learning, which are essential for success in online education Consequently, it is anticipated that these students will perform worse than their better-prepared peers, particularly in online settings A study comparing online and face-to-face economics courses revealed that while high-GPA students performed similarly in both formats, low-GPA students scored significantly lower in online classes, highlighting their challenges in adapting to this mode of learning.
Research indicates that student characteristics significantly influence online success, with women and White students generally performing better than their peers However, existing studies often treat these characteristics as simple predictors of performance, neglecting their role in students' adaptability to online learning This oversight raises questions about whether women adapt more readily to online environments than men, potentially widening the gender gap in online courses compared to traditional face-to-face settings Consequently, there is a lack of evidence regarding how the growth of online learning affects various student demographics differently.
Empirical Framework and Data
Data and Summary Statistics
A comprehensive analysis was conducted on a dataset comprising 51,017 degree-seeking students who enrolled for the first time in one of Washington State's 34 community or technical colleges during the fall term of 2004 These students were monitored over a span of approximately five years, tracking their enrollment through the spring of 2009 across 19 quarters The dataset, sourced from the Washington State Board of Community and Technical Colleges (SBCTC), contained detailed information on student demographics, the institutions they attended, and their course enrollment and performance data.
3 This sample does not include students who were dual-enrolled during the fall term of 2004 (N = 6,039)
4 There are four quarters in each academic year, which starts in summer and ends in spring We also refer to a quarter as a term
The dataset encompasses various demographic details of students, including gender, ethnicity (Asian, Black, Hispanic, White, or Other), and age (25 or older at college entry) It also captures socioeconomic status (SES) based on the census area, along with academic background variables such as dual enrollment in high school and metrics derived from transcript data, like ever-remedial status, term credits, and GPA Additionally, the dataset features information from Washington State Unemployment Insurance (UI) wage records, detailing individual employment status and working hours for each term.
The transcript data provided detailed information on each course, including course number, subject, delivery format, and grades ranging from 0.0 to 4.0 Course persistence was also analyzed as a measure of student performance, defined as 1 if a student remained enrolled until the semester's end and 0 if they withdrew after the census date but before completion This study aims to explore the relationship between course delivery, persistence, and grades, while accounting for variations across different academic subjects Consequently, courses without valid decimal grades, such as audited courses or those with missing or incomplete grades, were excluded from the analysis.
5 SBCTC divides students into five quintiles of SES status, based on Census data regarding the average income in the census block in which the student lives
The SBCTC offers Classification of Instructional Programs (CIP 2000) codes for every course in the dataset We have organized these courses into broader subject categories, as detailed in Table 2, by utilizing the 2-digit series of the CIP codes.
The SBCTC categorizes courses into three types: face-to-face, online, and hybrid With less than 2 percent of courses offered in a hybrid format, which allows for online technology to replace only 50 percent or less of course delivery, we have merged hybrid courses with face-to-face courses for this analysis A robustness check excluding hybrid courses yielded results nearly identical to those shown in Tables 2 to 5, despite some courses lacking academic subject information.
The 34 Washington community and technical colleges vary widely from one another in terms of institutional characteristics The system comprises a mix of large and small schools, and the institutions are located in rural, suburban, and urban settings Table 1 describes institutional characteristics of the 34 community and technical colleges in fall 2004 based on statistics reported to the 2004 Integrated Postsecondary Education Data System (IPEDS) database Compared to the national sample, Washington community colleges serve substantially lower proportions of African American and Hispanic students and slightly higher proportions of White students The SBCTC system also serves lower proportions of students who receive federal financial aid Compared to national samples, community colleges in the Washington State system are also more likely to be located in urban areas In summary, Washington community colleges seem to more closely represent an urban and White student population than do community colleges in the country as a whole
Table 1 Characteristics of Washington State Community and Technical Colleges Versus a National Sample of Public Two-Year Colleges
Percent of students receiving federal financial aid 43.94 (18.71) 27.94 (10.63)
Percent of full-time students 64.53 (11.87) 64.93 (6.71)
Instructional expenditures per FTE (in dollars) 5,261.52 (20,987.74) 4,848.71 (2,133.11)
Note Standard deviations for continuous variables are in parentheses.
Empirical Models
This study utilizes a basic ordinary least squares (OLS) model to analyze two key course outcomes: student persistence in the course and the final decimal grade achieved The primary explanatory variable examined is the format of course delivery, specifically whether students participated in the course online or in a face-to-face setting.
The equation \( Y_i = \alpha_i + \beta \text{online}_i + \gamma X_i + \mu_i \) illustrates the relationship between various factors and student outcomes, where the variable "online" indicates whether a course was taken online (1) or not (0) The term \( X_i \) encompasses demographic characteristics such as age, gender, race, and socioeconomic status (SES), along with academic preparedness indicators like remedial status and previous dual enrollment, as well as semester-specific data like total credits taken The error term \( \mu_i \) accounts for unobserved factors affecting the outcome.
A significant challenge in assessing the effectiveness of alternative course delivery formats is the issue of student selection bias, as students who choose online courses often differ markedly from those in traditional settings If these differences are not accounted for, the resulting estimates may be skewed In our prior analysis of SBCTC data (Xu & Jaggars, 2012), we employed an instrumental variable (IV) approach to derive a robust causal estimate comparing online and face-to-face coursework Our findings revealed that a simpler ordinary least squares (OLS) method underestimated the negative effects associated with online learning.
To address omitted student selection bias in our analysis, we utilized a data structure featuring multiple course observations per student and applied an individual fixed effects approach This method allowed us to separate the unobserved factors influencing the dependent variable into two components: constant factors, such as gender, and variable factors that differ by course, such as course subject The individual fixed model can be expressed as follows:
Y ic = α ic + β online ic + γ X ic + σ i + υ ic (2)
To ensure a comprehensive analysis of course persistence, which is a discrete outcome variable, we employed logistic regression as a robust verification method The findings align closely with those shown in Table 3 For clarity, we present the results from Ordinary Least Squares (OLS) estimates In this context, \$\sigma_i\$ accounts for all unobserved, course-constant factors influencing course performance, while \$\upsilon_{ic}\$ represents unobserved factors that vary across courses and impact \$Y_{ic}\$ By averaging this equation across courses for each individual \$i\$, we derive meaningful insights into the factors affecting course outcomes.
𝑌� 𝑖 = α� 𝑖𝑐 + β 𝑂𝑛𝑙𝚤𝑛𝑒��������� 𝑖 + γ X� 𝑖 + σ 𝑖 + υ� 𝑖 (3) where 𝑌� 𝑖𝑐 = T -1 ∑ 𝑌𝑖𝑐 , and so on Because σ i is fixed across courses, it appears in both equation (2) and equation (3) Subtracting (3) from (2) for each course yields:
The equation \( \overset{ }{Y}_{ic} = \overset{ }{\alpha}_{ic} + \beta \overset{ }{O}_{nline_{ic}} + \gamma \overset{ }{X}_{ic} + \overset{ }{\nu}_{ic} \) illustrates the course-demeaned data on course outcomes, where \( \overset{ }{Y}_{ic} = Y_{ic} - \overline{Y}_{i} \) A key aspect of this equation is that the unobserved effect \( \sigma_{i} \) is eliminated through the within-individual transformation, allowing for the removal of potential unobserved bias when using the individual fixed effects model, provided this bias remains constant across courses This model effectively compares online and face-to-face courses taken by the same student, with the online coefficient \( \beta \) reflecting student adaptability to online learning A negative coefficient indicates that the student performs worse in online courses compared to face-to-face courses, while a positive coefficient suggests better performance in online settings.
While we have successfully eliminated course-invariant biases, biases that vary by course may still persist in equation (4) One potential source of this bias is specific course-level attributes that affect both online enrollment and course outcomes For instance, if online courses are more frequently offered in later years or specific subjects, the estimates from equation (4) could be biased if the academic subject or timing of enrollment correlates with course outcomes To mitigate the issue of varying probabilities of online enrollment across different subjects and time periods, we incorporated time and academic subject fixed effects into the individual fixed model.
In addition to variations in the likelihood of enrolling in online courses based on timeframes or subjects, we identified three key sources of selection bias One significant factor is the differing levels of difficulty among courses within the same subject; for instance, advanced courses often present greater academic challenges than introductory ones If introductory courses are offered online at different rates compared to advanced courses, this could skew our estimates To mitigate this issue, we conducted a supplementary robustness check, focusing exclusively on courses taken during each student's initial term, where first-time students are restricted to introductory courses.
The strategy also addressed concerns about students sorting course modalities based on prior performance For instance, among 2,765 students who enrolled in an online course during their first term, those who did not achieve a grade of C or higher were 18 percentage points less likely to attempt another online course later, even when controlling for other individual factors Consequently, adaptability estimates for online courses taken in subsequent semesters may be positively biased By focusing on courses taken in the first term, we can mitigate this selection bias, as students are less likely to make modality choices based on their initial online course performance, given their limited experience and understanding of online learning within the college context.
A significant source of course-variant bias is the individual characteristics that evolve over time, particularly working hours, which can directly affect online enrollment and course outcomes Fluctuating work schedules among students may influence their course-taking patterns and performance The dataset utilized for this analysis included quarterly employment data for 60 percent of the sample To ensure robustness, we performed an individual fixed effects analysis, incorporating academic subject and time fixed effects, along with individual working hours as a covariate; the findings are detailed in Table 3 (in section 3).
Empirical Results
Online Course Enrollments Across Different Subjects
In a sample of 498,613 course enrollments, around 10 percent were conducted online, with significant variation across different subjects Table 2 illustrates the enrollment patterns across all subject areas, ranked by the proportion of online enrollments Notably, online courses were most favored in the humanities, accounting for over 19 percent of enrollments among the 14 subject-area categories analyzed.
Between 2004 and 2009, social science emerged as the second largest category for online enrollments, accounting for 18 percent, while education and computer science followed closely with approximately 15 percent each Other notable subjects with above-average online enrollments included applied professions at 13 percent, English at 12 percent, and mass communication at 11 percent Conversely, engineering saw extremely low online enrollments, with less than 1 percent, alongside developmental education and English as a second language, which had 4 percent.
The online enrollment data across various subject areas indicate three key trends Firstly, online courses are predominantly favored in the arts and humanities, while they are less popular in natural sciences, despite high enrollments in specific fields like astronomy and geology, which represent a small fraction of overall science courses Secondly, the proportion of online enrollments remains relatively stable within each subject-area category, with social sciences such as anthropology, philosophy, and psychology showing enrollments between 18% and 24% Lastly, online enrollments are significantly higher in college-level courses compared to pre-college courses, including developmental and ESL education.
Table 2 Proportion of Online Enrollments by Subject
Subject Area Proportion of Enrollments
Note Please refer to footnote 6 for information on how the subject areas were classified.
Students’ Online Adaptability Overall
The average persistence rate for students across courses was 94.12 percent, with online courses at 91.19 percent and face-to-face courses at 94.45 percent Among the 469,287 students who completed their courses, the average grade was 2.95 on a 4.0-point scale, showing a disparity between online courses (2.77) and face-to-face courses (2.98) Table 3 illustrates the online coefficients for course persistence and grades, revealing consistently significant and negative estimates across all models, highlighting the challenges students faced in adapting to the online learning environment.
Table 3 Coefficients for Online (Versus Face-to-Face) Learning
Full Course Sample Initial Semester Only
Individual FE No Yes Yes Yes No Yes
Subject FE No No Yes Yes No No
Time FE No No Yes Yes No No
Individual FE No Yes Yes Yes No Yes
Subject FE No No Yes Yes No No
Time FE No No Yes Yes No No
All models in this study account for standard errors clustered at the student level and incorporate several covariates, including gender, race, socioeconomic status, federal financial aid status, limited English proficiency, prior dual enrollment, total credits taken in the term, enrollment in remedial courses, and full-time college enrollment during that term.
***Significant at the 1 percent level
Estimates from the individual fixed effects model indicate that the impact of online course enrollment on academic performance is 20 to 40 percent greater than those derived from the OLS model Incorporating time and academic subject fixed effects, along with working hours, yields similar or even larger estimates This suggests that students inclined to take online courses generally exhibit stronger academic performance compared to their peers Consequently, OLS estimates may underestimate the negative effects of online course enrollment by failing to account for essential individual variables, leading to an overestimation of students' ability to adapt positively to online learning.
Table 3 focuses on courses taken during a student's initial term to examine how prior online learning experiences influence course format selection This period typically limits students to introductory courses, allowing for an analysis of the relationship between course difficulty and the likelihood of online offerings The negative estimates associated with online learning outcomes remain significant in this first-term-only analysis, reinforcing the findings from the full sample by demonstrating that these negative effects persist even after accounting for potential student-level and course-level selection biases.
Adaptability Across Different Types of Students
To investigate the differences in outcomes between online and face-to-face learning, we analyzed the moderating effects of gender, age, previous academic performance, and ethnicity on specific student subgroups The findings are detailed in Table 4, where we first incorporated an overall interaction term between each individual attribute and the course format in our heterogeneity analysis.
For this robustness check, the sample size was limited to 297,767 for course persistence and 279,073 for course grade, as students without a valid Social Security Number, such as international students, or those in special employment situations like self-employed individuals, were excluded due to missing values for a given quarter.
The results exclude a model with time or academic subject fixed effects due to minimal variation by term and subject when individual fixed effects are applied Additionally, working hours are not included, as they do not vary across courses within a single term, leading to their automatic exclusion from the individual fixed model focused on one term.
The last row of each panel presents the p-value for each interaction term in Equation 2 To clarify the significance of these interactions, we performed separate analyses for each subgroup while maintaining the same model specification Additionally, to interpret the main effects of student characteristics, we conducted supplementary analyses using Equation 1 when necessary.
Table 4 Individual Fixed-Effects Estimates for Online Learning, by Student Subgroup
Course Persistence Course Grade Gender
Male (N = 225,775) −0.054 (0.003)*** −0.288 (0.013)*** p-value for the interaction term < 001 051
Other (N = 73,253) −0.046 (0.005)*** −0.224(0.019)*** p-value for the interaction terms 484 < 001
Below 25 (N = 376,448) −0.049 (0.002)*** −0.300 (0.009)*** p-value for the interaction term < 001 < 001
Took any remedial courses (N = 305,091) −0.045 (0.002)*** −0.272 (0.010)*** p-value for the interaction term 078 017
GPA in 1st Term Face-to-Face Courses
Below 3.0 (N = 170,219) −0.058 (0.003)*** −0.314 (0.015)*** p-value for the interaction term < 001 < 001
In this analysis, \(N\) denotes the total number of courses undertaken by the subgroup Each cell in the table corresponds to a distinct regression employing an individual fixed effects approach All equations incorporate both time fixed effects and academic subject fixed effects, particularly for subjects encompassing multiple disciplines, as detailed in Table 2 Additionally, standard errors for all models are clustered at the student level.
***Significant at the 1 percent level
Equation 2 accounts for individual fixed effects, which means the main effects of student characteristics, such as gender, on face-to-face course performance are controlled for and excluded from the model Our research focuses on course-varying effects, specifically the performance gap between online and face-to-face formats, allowing us to include interactions between the online format and student characteristics These interactions can be interpreted similarly to those in models that include their main effects To discuss the main effects of student characteristics for a broader understanding of the results, we must refer to Equation 1.
The study revealed that all student subgroups experienced negative impacts from online learning, with varying degrees of severity Notably, male students exhibited stronger negative effects on course persistence and grades compared to female students, although the interaction for course grades was only marginally significant (p = 051) This suggests two interpretations: first, that men struggled more with the transition to online learning than women, and second, that while females generally performed better than males across all courses, the performance gap was more pronounced in online settings than in traditional face-to-face environments.
Students from various ethnic backgrounds were more likely to drop out of online courses compared to face-to-face courses, with no significant variation in dropout rates across ethnic groups However, among those who continued, there were notable differences in grades based on ethnicity in online learning Specifically, Black students exhibited a negative performance coefficient nearly double that of Asian students, indicating a wider performance gap in online courses compared to traditional classroom settings.
Both older and younger students exhibited significant negative coefficients for online learning, but the impact was notably weaker for older students regarding course persistence and grades While older students tended to achieve higher course grades, they were also more likely to drop out compared to younger students To clarify the moderating effect of age, we analyzed course persistence rates for both age groups across different course delivery formats In face-to-face courses, the adjusted probability of persistence was 95% for younger students and 94% for older students, whereas in online courses, younger students had a predicted persistence rate of 90%, indicating a reversal in trends.
Older students showed a 91 percent performance rate in online courses, indicating they performed worse than in face-to-face settings However, the decline in their performance was less significant compared to younger students This suggests that older students possess a better adaptability to online learning, giving them a slight edge over younger learners in virtual courses.
To explore how lower academic skills may influence online learning outcomes, we examined whether students had ever enrolled in a remedial course The F test revealed a significant interaction effect on course persistence (p = 078) and course grades (p = 017), suggesting that students with lower academic preparedness struggle more with online courses However, using remedial enrollment as a measure of academic skill is problematic, as many students assigned to remediation do not actually complete the courses Consequently, the "non-remedial" group may include students who are academically underprepared but opted out of remediation Additionally, a significant number of students placed in remediation tend to drop out in their first or second semester, leading to a student population that is increasingly motivated and better equipped for success Therefore, the findings may underestimate the interaction effects between initial academic preparedness and the format of course delivery.
We conducted an additional analysis to explore the impact of academic capacity by utilizing students' GPA from their face-to-face courses during the initial term This approach was chosen for two main reasons: first, it allowed us to eliminate the influence of varying course formats on GPA outcomes, and second, it accurately reflected academic performance in the majority of courses taken by students in their first semesters, as only 7 percent of students enrolled in online courses during that time.
The main limitation of this indicator is the exclusion of students without a valid first-term face-to-face GPA, which led to a 13 percent reduction in the overall course sample This group may have withdrawn from all courses, earned only remedial credits, or completed only online courses during their first semester We were concerned that this reduced sample might differ from the original in terms of the online format's impact on course outcomes To address this, we re-analyzed the online impacts on this subsample, finding results nearly identical to those in Table 3, particularly regarding coefficient persistence.
The analysis revealed a significant negative coefficient of -0.046 (p < 01) and -0.275 (p < 01) for grade, indicating a strong relationship between academic capacity and performance Notably, only 3 percent of students completed all their courses online during that term As illustrated in Table 4, the interactive effect of academic capacity was more pronounced when assessed using GPA, with significant p-values for the interaction terms (p < 01) related to both course persistence and course grade Furthermore, the disparity in coefficients between the two groups was greater than that observed in the ever-remedial model.
Research shows that students with higher academic abilities experienced less negative impact from online courses compared to their lower-skilled peers Additionally, the performance gap between high- and low-skill students was more pronounced in online learning environments than in traditional face-to-face classes.
Differences in Online Adaptability Across Course Subject Areas
To investigate the effectiveness of online learning across different academic subjects, we incorporated interaction terms between subject areas and online course formats into our analysis The results revealed a strong and significant interaction effect on both course persistence (F = 6.01, p < 001) and course grades, indicating that students adapt differently to online learning depending on the subject area.
The analysis revealed a significant variation in student adaptability to online learning across different academic subject areas, with an F-value of 13.87 and a p-value less than 001 To further explore these interaction effects, we calculated the coefficients for online learning specific to each subject area using Equation 3.
All models incorporate time fixed effects and fixed effects for academic subjects, particularly for those with multiple sub-disciplines, as detailed in Table 2 Table 5 illustrates the results, with each cell representing a distinct regression that utilizes both individual and time fixed effects, alongside the fixed effects for academic subject areas with various sub-disciplines.
Table 5 Individual Fixed-Effect Estimate for Online Learning, by Course Subject
(restricted to academic subjects with at least 5 percent online enrollment)
Subject Course Persistence Course Grade
Math −0.065 (0.016)*** −0.234 (0.056)*** p-value for the interaction terms < 001 < 001
All models account for standard errors clustered at the student level and incorporate both time fixed effects and academic subject fixed effects, particularly for subjects with multiple disciplines, as detailed in Table 2.
***Significant at the 1 percent level **Significant at the 5 percent level *Significant at the 10 percent level
Research indicates that all academic subjects exhibited negative coefficients for online learning regarding course persistence and grades Nonetheless, certain subjects displayed weaker negative coefficients, with three areas showing insignificant effects on persistence Notably, students adapted better to online learning in computer science, applied professions, and natural science, where the negative impact on both course persistence and grades was less pronounced.
The variation in student adaptability across different subject areas may be influenced by the characteristics of students enrolled in online courses Although we accounted for overall student traits, we did not consider how these traits affected performance differences between online and face-to-face formats By incorporating interaction terms between course delivery format and key individual characteristics—such as gender, ethnicity, first-term face-to-face GPA, and age—we found that the significance of these interaction terms persisted Specifically, the results indicated that differences in course persistence (F = 2.55, p = 004) and course grades (F = 5.55, p < 001) remained significant, suggesting that the effectiveness of online courses varies across subject areas, even after adjusting for student characteristics and their adaptability to online learning.
Variation in online learning impacts across academic subjects may stem from peer effects related to the overall composition of students in each area While individual characteristics are accounted for, the influence of peers on performance remains unexamined Analyses reveal that students with higher first-term face-to-face GPAs tend to enroll in subjects with less negative online learning outcomes For instance, the average first-term GPA was 2.95, but it was notably higher in natural sciences (3.02), computer science (3.02), and applied professions (3.03) In physics, where the first-term GPA reached 3.12, the negative impacts of online learning on course persistence and grades were insignificant Conversely, subjects like English and social sciences, with lower GPAs (2.89 and 2.82, respectively), exhibited stronger negative estimates for online learning These findings indicate that students are influenced by the performance levels of their peers, which may affect their adaptability to online courses in different subjects.
To assess the influence of peer effects on students' adaptation to online courses, we developed an indicator called "online-at-risk," which identifies students who are academically underprepared (with a first-term face-to-face GPA below 3.0) and possess demographic traits associated with a higher likelihood of poor online performance, such as being male, younger, or Black We calculated the proportion of online-at-risk students for each course and examined its interaction with the course delivery format The results showed that the interaction terms were significantly negative (p < 01) for both course persistence and grades, suggesting that students' performance in online courses deteriorated more when their peers struggled to adapt to the online environment.
We analyzed the peer effect interaction by estimating the online learning coefficient for two distinct groups of courses: those with 75% or more students online-at-risk and those with 25% or fewer In the former group (N = 25,128), we observed significant negative coefficients for online delivery, with values of −0.064 (p < 01) for course persistence and −0.359 (p < 01) for course grade Conversely, in the latter group (N = 201,539), the negative impacts were reduced, showing coefficients of −0.035 (p < 01) for course persistence and −0.231 (p < 01) for course grade.
After accounting for various student characteristics, including peer effects, significant differences were found in the effectiveness of online courses across different academic subjects, with p-values less than 01 for both course persistence and grades To clarify these findings, we focused on a specific group of students—female, older, non-Black individuals with a GPA of 3.0 or higher in face-to-face courses—totaling 39,614 enrollments In this adaptable subgroup, any persistent negative online coefficients for specific subjects suggest that those areas may be inherently challenging to transition to an online format.
In this subsample, the online coefficients were largely non-significant for course outcomes across most subject areas; however, they exhibited a significantly and substantially negative impact specifically in the field of social science (N = 3,136).
Coefficientpersistence= −0.050, p < 01; Coefficientgrade = −0.195, p < 01) and applied professions (N = 12,924; Coefficientpersistence= −0.020, p = 0.01; Coefficientgrade = −0.135, p < 01).
Discussion and Conclusion
This study examined student performance in online and face-to-face courses using a statewide community college dataset to determine the adaptability of different student subgroups to online coursework The findings revealed a significant negative impact of the online format on both course persistence and grades, suggesting that most students struggled with online learning Although this negative trend was consistent across all subgroups, the extent of the impact varied notably among them.
Our research indicates that male students, Black students, and those with lower academic preparation face greater challenges in online learning, reflected in lower course persistence and grades compared to their peers This highlights the variability in students' adaptability to online education, suggesting that outcomes can differ significantly based on demographic and academic factors.
Research indicates that performance disparities among key demographic groups, such as between male and female students and between White and ethnic minority students, are intensified in online learning environments (Hoskins & van Hooff, 2005; Jun, 2005; Stewart et al., 2010) This trend raises significant equity concerns, suggesting that the growth of online education may worsen, rather than alleviate, existing educational inequities across various states and sectors.
Older students demonstrated a greater adaptability to online courses compared to their younger counterparts, despite generally experiencing poorer academic outcomes This is noteworthy, as older students often juggle work and family responsibilities, making the flexibility of online learning essential Although they performed slightly worse in online settings than in traditional classrooms, this minor decline in performance may be a reasonable compromise, allowing them to enroll in more courses each semester.
Our research indicates that the effectiveness of online learning differs not only among various student types but also across different academic subjects Certain subjects may inherently lend themselves better to online formats, yet the overall makeup of student enrollments in a specific subject area also plays a crucial role in determining the success of online courses.
Different types of students cluster into specific academic subject areas, with some excelling in online coursework while others struggle A student's performance in an online course can be negatively impacted by classmates who do not adapt well, particularly in subjects like English and social science, which attract less-adaptable students This can lead to challenges in interpersonal interactions and group projects, ultimately affecting overall course performance Additionally, instructors may focus more on students who are struggling, leaving others with less support Future research should explore the mechanisms of peer effects in online courses to better understand these dynamics.
Research indicates that certain academic fields, particularly the social sciences and applied professions, pose greater challenges for students in online learning environments Subjects like anthropology, philosophy, psychology, business, law, and nursing often necessitate hands-on demonstrations and practical experiences, complicating the development of effective online materials and assignments Additionally, these disciplines typically rely on intensive interactions between students and instructors, as well as peer discussions, which are harder to facilitate in an online setting.
Our research reveals that while many students struggle to adjust to online courses, there is a significant variation in adaptability among individuals To enhance student success in these courses, colleges can implement four key strategies: screening students for potential challenges, providing scaffolding to support learning, establishing early warning systems to identify at-risk students, and pursuing comprehensive improvements in course design and delivery.
Colleges could redefine online learning as a privilege by implementing screening measures that require students to demonstrate their ability to adapt to online coursework, such as maintaining a GPA of 3.0 or completing a workshop on online learning skills However, this approach may disadvantage older students who need the flexibility of online classes and could lead to decreased enrollments as students seek institutions without such restrictions Additionally, the demographic variation among students affects academic departments, as more adaptable students tend to cluster in certain areas As an alternative, colleges might limit the availability of online courses in subjects where many students struggle to adapt, a strategy already seen in the limited online offerings for developmental education, where many students are academically underprepared.
One effective strategy for enhancing online learning is scaffolding, which involves integrating the teaching of online learning skills into courses where less-adaptable students often enroll, such as English composition This approach necessitates collaboration between colleges and instructors to create materials and assignments that foster these essential skills However, a potential downside is that students may take multiple scaffolded courses, leading to boredom and frustration with repetitive online learning exercises.
Incorporating early warning systems into online courses can effectively identify and support students struggling to adapt For instance, if a student does not log into the online platform or fails to submit an early ungraded assignment, the system can alert instructors or the counseling department This proactive approach allows for timely intervention, enabling staff to reach out and discuss potential challenges and solutions with the student While early warning systems are gaining popularity, they may involve significant initial costs and require dedicated faculty or counselor time.
The initial three strategies suggest that most online courses maintain a consistent quality, while students enhance their online skills In contrast, the fourth strategy emphasizes the need to elevate the quality of all college online courses to match the learning outcomes of traditional face-to-face classes, irrespective of student demographics Implementing this improvement strategy would necessitate significant investments in course design, faculty development, support for both learners and instructors, and comprehensive course evaluations.
Online coursework is essential in postsecondary education, enhancing flexibility for students and institutions while broadening educational opportunities for those juggling work and family commitments Our findings can guide stakeholders in improving student outcomes in online courses However, this study is limited to community colleges in one state, highlighting the need for further research in other states and four-year colleges to better understand how individual characteristics and course subjects affect students' adaptation to online learning.
Allen, D (1997, May) The hunger factor in student retention: An analysis of motivation
Paper presented at the Annual Forum of the Association for Institutional
Allen, I E., & Seaman, J (2010) Class differences: Online education in the United
States, 2010 Needham, MA: Babson Survey Research Group
Aslanian, C (2001) You’re never too old: Excerpts from adult students today
Astleitner, H., & Steinberg, R (2005) Are there gender differences in web-based learning? An integrated model and related effect sizes AACE Journal, 13(1), 47–
Bambara, C S., Harbour, C P., Davies, T G., & Athey, S (2009) Delicate engagement:
The lived experience of community college students enrolled in high-risk courses
Bailey, T., Jeong, D W., & Cho, S W (2010) Referral, enrollment, and completion in developmental education sequences in community colleges Economics of
Bernard, R M., Abrami, P C., Lou, Y., Borokhovski, E., Wade, A., Wozney, L.,…
Huang, B (2004) How does distance education compare with classroom instruction? A meta-analysis of the empirical literature Review of Educational
Biner, P M., Summers, M., Dean, R S., Bink, M L., Anderson, J L., & Gelder, B C
(1996) Student satisfaction with interactive telecourses as a function of demographic variables and prior telecourse experience Distance Education,
Calcagno, J C., Crosta, P., Bailey, T., & Jenkins, D (2007) Stepping stones to a degree:
The impact of enrollment pathways and milestones on community college student outcomes Research in Higher Education, 48(7), 775−801
Choy, S (2002) Findings from the condition of education 2002: Nontraditional undergraduates (NCES 2002–012) Washington, DC: U.S Department of
Education, National Center for Education Statistics
Choy, S., & Premo, M (1995) Profile of older undergraduates, 1989–90 (NCES 95–
167) Washington, DC: U.S Department of Education, National Center for
Chyung, S Y (2001) Systematic and systemic approaches to reducing attrition rates in online higher education The American Journal of Distance Education, 15(3), 36−49
Colorado, J T., & Eberle, J (2010) Student demographics and success in online learning environments Emporia State Research Studies, 46(1), 4−10
Didia, D., & Hasnat, B (1998) The determinants of performance in the university introductory finance course Financial Practice and Education, 8(1), 102–107
Dille, B., & Mezack, M (1991) Identifying predictors of high risk among community college telecourse students American Journal of Distance Education, 5(1), 24–
DiPrete, T A & Buchmann, C (2006) Gender-specific trends in the values of education and the emerging gender gap in college completion Demography 43(1), 1–24
DuBrock (2000) conducted a five-year longitudinal study examining the relationship between financial aid and college persistence among freshmen students who began their studies in 1993 and 1994 The findings were presented at the Annual Forum of the Association for Institutional Research in Cincinnati, OH.
Eisenberg, E., & Dowsett, T (1990) Student drop-out from a distance education project course: A new method of analysis, Distance Education, 11(2), 231−253
Ehrman, M (1990) Psychological factors and distance education, American Journal of
Feldman, M J (1993) Factors associated with one-year retention in a community college Research in Higher Education, 34(4), 503−512
Figlio, D N., Rush, M., & Yin, L (2010) Is it live or is it internet? Experimental estimates of the effects of online instruction on student learning (NBER Working
Paper No 16089) Cambridge, MA: National Bureau of Economic Research
Gladieux, L., & Swail, W S (1999) The virtual university and educational opportunity:
Issues of equity and access for the next generation Washington, DC: The College Board Retrieved from http://www.educationalpolicy.org/pdf/Virtual%20University.pdf
Goldin, C., Katz, L., & Kuziemko, I (2006) The homecoming of American college women: The reversal of the college gender gap (NBER Working Paper
No.12139) Retrieved from: http://www.nber.org/papers/w12139
Gunn, C., McSporran, M., Macleod, H., & French, S (2003) Dominant or different?
Gender issues in computer supported learning Journal of Asynchronous Learning
Heckman, J J., & LaFontaine, P A (2007) The American high school graduation rate:
Trends and levels (NBER Working Paper No 13670) Retrieved from http://www.nber.org/papers/w13670
Horn, L., & Carroll, C (1996) Nontraditional undergraduates: Trends in enrollment from
1986 to 1992 and persistence and attainment among 1989–90 beginning postsecondary students (NCES 97–578) Washington, DC: U.S Department of Education, National Center for Education Statistics
Hoskins, S L., & van Hooff, J C (2005) Motivation and ability: Which students use online learning and what influence does it have on their achievement? British
Hyllegard, D., Deng, H., & Carla, H (2008) Why do students leave online courses?
Attrition in community college distance learning courses International Journal of
Jaggars, S S (2012, April) Beyond flexibility: Why students choose online courses in community college Paper presented at the American Educational Research
Association Annual Meeting, Vancouver, Canada
Jaggars, S S., & Hodara, M (2011) The opposing forces that shape developmental education: Assessment, placement, and progression at CUNY community colleges
(CCRC Working Paper No 36) New York, NY: Columbia University, Teachers College, Community College Research Center
Jahng, N., Krug, D., & Zhang, Z (2007) Student achievement in online distance education compared to face-to face education European Journal of Open,
Distance and E-Learning Retrieved from http://www.eurodl.org/
Jun, J (2005) Understanding dropout of adult learners in e-learning (Unpublished doctoral dissertation) The University of Georgia, Athens, GA
Kleinman, J., & Entin, E B (2002) Comparison of in-class and distance-learning:
Students’ performance and attitudes in an introductory computer science course
Journal of Computing Sciences in Colleges, 17(6), 206–219
Liu, S., Gomez, J., Khan, B., & Yen, C (2007) Toward a learner-oriented community college online course dropout framework International Journal on E-Learning,
Lu, J., Yu, C.-S., & Liu, C (2003) Learning style, learning patterns, and learning performance in a WebCT-based MIS course Information and Management,
McSporran, M., & Young, S (2001) Does gender matter in online learning? Retrieved from http://hyperdisc.unitec.ac.nz/research/ALTJpaper_9.pdf
Muse, H E (2003) A persistence issue: Predicting the at-risk student in community college Web-based classes (Unpublished doctoral dissertation) Nova
Southeastern University, Ft Lauderdale, FL
Newell, C C (2007) Learner characteristics as predictors of online course completion among nontraditional technical college students (Unpublished doctoral dissertation) University of Georgia, Athens, GA
Ory, J C., Bullock, C., & Burnaska, K (1997) Gender similarity in the use of and attitudes about ALN in a university setting Journal of Asynchronous Learning
Osborn, V (2001) Identifying at-risk students in videoconferencing and web-based distance education American Journal of Distance Education, 15(1), 41–54
Parsad, B., & Lewis, L (2008) Distance education at degree-granting postsecondary institutions: 2006–07 (NCES 2009–044) Washington, DC: U.S Department of
Education, National Center for Education Statistics
Price, L (2006) Gender differences and similarities in online courses: Challenging stereotypical views of women Journal of Computer Assisted Learning, 22(5), 349–359
Rainie, L (2010) Internet, broadband, and cell phone statistics Washington, DC: Pew
Internet Retrieved from http://pewinternet.org/Reports/2010/Internet-broadband- and-cell-phone-statistics.aspx
Roksa, J., Jenkins, D., Jaggars, S S., Zeidenberg, M., & Cho, S W (2009) Strategies for promoting gatekeeper success among students needing remediation: Research report for the Virginia Community College System New York: Columbia
University, Teachers College, Community College Research Center
Rovai, A P., & Baker, J D (2005) Gender differences in online learning: Sense of community, perceived learning, and interpersonal interactions Quarterly Review of Distance Education, 6(1), 31−44
Sierra, C., & Wang, M (2002) Gender, discourse style, and equal participation in online learning In G Richards (Ed.), Proceedings of E-Learn 2002 Conference (pp 2364-2367), Chesapeake, VA: AACE
Sitzmann, T., Kraiger, K., Stewart, D., & Wisher, R (2006) The comparative effectiveness of web-based and classroom instruction: A meta-analysis Personnel
Stewart, C., Bachman, C., & Johnson, R (2010) Students’ characteristics and motivation orientations for online and traditional degree programs Journal of Online
Sullivan, P (2001) Gender differences and the online classroom: Male and female college students evaluate their experiences Community College Journal of
I don't know!
Taplin, M., & Jegede, O (2001) Gender differences in factors influencing achievement of distance education students Open Learning, 16(2), 133−154
U.S Department of Education, Office of Planning, Evaluation, and Policy Development
(2010) Evaluation of evidence-based practices in online learning: A meta- analysis and review of online learning studies Washington, DC: Author
Xu, D., & Jaggars, S S (2011) The effectiveness of distance education across Virginia’s
Community Colleges: Evidence from introductory college-level math and English courses Educational Evaluation and Policy Analysis, 33(3), 360−377
Xu, D., & Jaggars, S S (2012) Examining the effectiveness of online learning within a community college system: An instrumental variable approach Unpublished manuscript New York: Columbia University, Teachers College, Community College Research Center
Wang, A Y., & Newlin, M H (2002) Predictors of performance in the virtual classroom THE Journal, 29(10), 21−25
Wiggam, M K (2004) Predicting adult learner academic persistence: Strength of relationship between age, gender, ethnicity, financial aid, transfer credits, and delivery methods (Unpublished doctoral dissertation) Ohio State University,
Willging, P A., & Johnson, S D (2004) Factors that influence students’ decision to dropout of online courses Journal of Asynchronous Learning Networks, 13(3), 115−127
Willis, B (1992) Effective distance education: A primer for faculty and administrators
(Monograph Series in Distance Education, No 2) Fairbanks, AK: University of Alaska.