Examining Predictive Factors For Student Success In A Hybrid Learning Course

In the United States, online education has become a significant part of higher education in the last two decades and has proliferated with the development of new technologies and access to the Internet. In the fall of 2014, over 5.8 million students, which is 28% of total postsecondary students, were enrolled in at least one online course. Among them 2.85 million students were fully online, taking all of their courses online, and the other 2.97 million students were taking some of their courses online (Allen & Seaman, 2016). Earlier studies

often examined whether online learning is equal or better than traditional face-to-face courses, and comparison studies were frequently conducted between online and face-to-face classes or class sections, comparing students’ average grades, final course grades, retention, or other measures of student learning. Research settings such as the sample size, single or multiple instructors, different course structures, and control of the experiments were varied among studies (Bowen, Chingos. Lack, & Nygren, 2013). Consequently, the findings were as varied as the research settings and a good number of studies found “no significant difference” in student learning between the online and face-to-face format. A longer range meta-analysis study also found that students who are in online courses outperformed the students in face-to-face courses (Shachar & Nuemann, 2010). On the other hand, many studies report a higher rate of attrition in online learning for various reasons: different characteristics of individual students, students’ preference to choose the platform (online or face-to-face), or self-discipline to study independently. Consequently, 44.6% of chief academic officers across the country expressed concern in retaining students in online courses, much more than those who are enrolled in face-to-face courses (Allen & Seaman, 2015). The questions about why more students in online courses would drop out than those who are in face-to-face courses and what factors would affect student learning in online courses would offer an important insight to support students for their success (Jenkins, 2011; Xu & Jaggars, 2011).

In addition to the dichotomy of fully online and face-to-face courses, another approach that researchers have been investigating is combining the portions of in-class and online instruction in a form of hybrid learning or mixed delivery. The mixed delivery method used in hybrid learning course could allow students to take the benefits from both approaches (He, Gajski, Farkas, & Warschauer, 2015; Owston, York, & Murtha, 2013; Wu, Tennyson, & Hsia, 2010), help students learn better (Horn & Staker, 2011), and increase access for more students (Legon & Garrett, 2018) as an alternative mode of course delivery method. Also, by controlling the potential factors that may have negative effects on student learning in online courses, students may be better guided toward effective leaning, instead of being pushed to overcome the difficulties on their own. As students enroll and progress in hybrid learning courses, they may also develop their own perceptions regarding the online content and online activities, which could affect their motivation to persist, progress, or, in some cases, to drop the course.

Following review of literature surveys the existing research on hybrid learning format, effects of student demographic and academic backgrounds, and student participation and persistence in online and in-class activities in a hybrid learning course.

The concept of hybrid learning refers to the integrated use of face-to-face and online instruction, structured on the foundation of instructional design for organic structure and plan of the course. A typical hybrid learning course would include a substantial proportion (30%–79%) of course content delivered online, combined with a reduced number of inclass meetings (Allen & Seaman, 2016), and specific instructional plans for face-to-face and online portion of the course. As a result, a course with this mixed mode (Huang, Lin, & Huang, 2012) would include online delivery of content, online activities, and in-class activities, at varying proportions and weights, and sometimes in specific order of presentation, such as flipped or inverted order.

As the main intention of the hybrid learning approach is to capture the benefits of both face-to-face and online instruction (He, Gajski, Farkas, & Warschauer, 2015; Wu et al., 2010) in order to help students learn more effectively (Horn & Staker, 2011), the expectations for the approach in facilitating student learning are high, and a growing number of institutions are implementing hybrid learning initiatives in place. Also, a meta-analysis of studies on online and hybrid learning reports that students in online and hybrid learning had more gain in their learning compared to face-to-face learning, and students in hybrid learning course had largest gain in their learning among their peers in all delivery formats (Means et al., 2010). However, researchers warn the risk of “course-and-a-half’ phenomenon, where a simple mix of online and face-to-face instruction without appropriate integration would end up as an ineffective course (Diaz & Brown, 2010).

Researchers continue examining the effects of hybrid learning, and report students’ academic performance, as well as their persistence. Lopez-Perez, Perez-Lopez, and Rodriguez-Ariza (2011) reported that students in hybrid learning course had increased final grades and reduced dropout rates, and Al-Qahtani and Higgins (2013) reported increased performance among students in hybrid courses when compared to students in traditional and online courses. Students in hybrid learning courses also had improved exam passing rates (Deschacht & Goeman, 2015). On the other hand, other studies reported that students achieved lower grades (Drysdale, Graham, Spring, & Halverson, 2013; Xu & Jaggars, 2011) and that there was an increased attrition rate (Ashby, Sadera, & McNary, 2011) in hybrid learning courses.

While the findings from recent studies on hybrid learning are mixed, it is noteworthy that a common caveat was assumed in several studies—the comparisons of student achievements were made based on the average grades by classes as the unit of measurement, instead of comparing individual student records. Deschacht and Goeman (2015) pointed that students who enroll in hybrid learning courses have more diverse backgrounds than those who are in traditional face-to-face courses. The difference in students’ background may influence on students’ academic performance, acting as a moderating factor (Huang et al., 2012). Hachey, Wladis, and Conway (2015) found that students with low prior GPA and students who were unsuccessful in a prior online course tend to be less successful in subsequent online courses, compared to face-to-face courses. Other studies reported that prior GPA could be a strong predictor of student success (Ary & Brune, 2011) and retention (Boston, Ice, & Burgess, 2012) in online courses. Also, students who had low precourse GPA performed better in subsequent face-to-face classes, but in hybrid learning courses, there was a wider gap in their individual performance. High-achieving students, on the other hand, attained higher grades in hybrid learning course (Asarta & Schmidt, 2017). In another study, students who have a record of high, medium, or low precourse GPA performed differently in hybrid learning courses. For example, students who had low GPA from previous courses were found to make larger gains than those who initially begin with high GPA (Brecht & Ogilby, 2008; Dupuis, Coutu, & Laneuville. 2013; Lambert, Parker, & Park, 2015). This trend of different performance patterns by student groups was also noticed in a previous course redesign project conducted preceding the current research (van Wallendael, Siegfried, & Spaulding, 2011).

As students study using the online course materials, such as e-textbooks, online videos, or external websites, students with different level of previous learning experiences made different level of gain. In Smith-Chant’s study (2010), watching online videos did not impact grades significantly, for “highly resourceful” students or students who already have enough behavioral or emotional skills to handle stressful situations, but if “low resourceful” students who usually lack such skills watched the videos frequently and with sufficient time, they made significant gains.

When there is a significant gap in gain in student learning, between students who had high or low level of achievement prior to taking a hybrid learning course, simply averaging individual students’ gain as a whole class may reduce the size of effect from implementing hybrid learning, as the gap in achievement can level out the effects of hybrid learning format within the same course. In a series of surveys conducted over 10 years, asking “unsuccessful” students who received F or W in online courses at a community college in New York, for the highest ranked reason why students thought they were unsuccessful in the course, 19.7% responded that they “got behind and it was too hard to catch up” (Fetzner, 2013). The study also found that students who are older than 25 years old, and those who have earned more college credit are more successful to get a “C” or better grade than those who are younger and have not taken more college courses.

Additionally, in a large scale longitudinal study of over 40,000 community and technical college students in Washington state, Xu and Jaggars (2013) found that male students who are younger in age, Black, and with low GPA struggled in online courses. Xu and Jaggars’ findings are in line with general findings on male students, as male students found to have weaker determination for completion of their study and weaker study habit (Ruffalo Noel Levitz, 2015). Combined with students’ individual background variables, the type of courses and type of college may add influence on students’ performance in online or hybrid learning course. Students in remedial course find difficulty in online courses (Xu & Jaggars, 2011) and Ashby, Sadera, and McNary (2011) reported that students in community college developmental math courses had a high attrition rate in online or hybrid learning courses, with low success rate (proportion of students who achieve course final grades of 70% or higher) and low scores in tests.

Online video is often used in hybrid learning courses, in order to present the major course topics or concepts to the students. Researchers have found that students’ perception on the value of the online video components may effect their motivation to continue in their study and progress in the course (Merhi, 2015). Also, students’ perceived usefulness of online videos or multimedia presentations is linked to the students’ academic performance (Wei, Peng, & Chou, 2015). For example, when students decide to be more actively engaged in the course and use online videos to learn and review the course content, their decisions can affect their academic performance (Bolt & Koh, 2001) and satisfaction (Shih, 2006). However, while instructor-created videos may have significant value, such as to add a sense of personal engagement among students, not all students used the videos as their main source to enhance their academic performance (Mandernach, 2009). An interesting aspect from Shah, Cox, and Zdanowicz (2013) study is that students perceived the prerecorded lectures that are specifically intended for the class to be most helpful, over classroom exercises and the selected YouTube videos presented to the class as supplemental content.

While some of the earlier studies reported no significant effects of online video content, it is noteworthy that many of the studies reporting no-significant effect (for example, DeVaney, 2009; Kelly et al., 2009) used online instructional videos as supplements to course lectures, often used for one-way delivery of the content. On the contrary, several other studies reported increased learning outcomes when courses integrated various types of interaction combined with online videos. For example, in Dupuis et al.’s (2013) study, students were asked to answer questions at the end of each video lecture, and the students took lecture notes while watching the video. Yilmiz and Keser (2016) found that online videos followed by reflective thinking activities were most effective for student learning. In Tune, Sturek, and Basile’s (2013) study, students learned more in the courses where online quizzes were integrated at the conclusion of lecture videos, compared to the students who were in the courses with standalone online quizzes. Comparing the learning outcomes among students who used interactive videos, noninteractive videos, and no videos, Zhang, Zhou, Briggs, and Nunamaker (2006) found that students who used interactive feature made significant gain, while students with noninteractive videos and who had no videos didn’t. He et al. (2015) also concurred that the interactive mode was most effective.

In hybrid learning courses, a portion of the course contents are specifically designed and delivered through online media, often on the institution’s learning management system. Therefore, to be successful in the course, students need to navigate the online environment, review the course content, and participate in the provided learning activities. For example, students can review the online course materials at their own pace to understand the main concepts and topics discussed in the course and prepare for homework and exams (Brecht & Ogilby, 2008), through active participation in the course learning activities (Huang et al., 2012). Previous studies found that students who access the course materials consistently (Smith-Chant, 2010; Baugher, Varanelli, & Weiboard, 2003) and who accessed more number of online course materials presented in the course (Crampton, Ragusa, & Cavanagh, 2012) performed better than those who only accessed the online course materials infrequently and reviewed only limited content.

On the other hand, student inactivity on the online course site had a high predictive value on student grade, as inactive students had a higher risk of failing the course than those who accessed the course site frequently (Fritz & Whitmer, 2017). Analyzing student online activity data, Zacharis (2015) found that student engagement with learning activities, such as reading and posting on discussion board, email and chatting, as well as taking optional online quizzes, was positively correlated to students’ success in hybrid learning courses. Another interesting finding from the study was that total time logged in the online course site and total clicks in the course site, analyzed as basic quantitative measures were only weakly correlated to students’ performance, but participation to optional online quizzes had significant effect, similar to the effects of participation to the required quizzes (Macfadyen & Dawson, 2010). Additionally, Fritz and Whitmer (2017) reported that students’ access to gradebook and checking on their progress during the course helped students improve their grades. As students learn from the online course materials, use of active learning strategies such as highlighting and annotating the e-textbook content had significant correlation with final course grade as a predictive factor (Junco & Clem, 2015). In this context, analyzing the student activity data can offer useful hints on understanding student learning progress (Martin & Whitmer, 2016) and where to focus efforts to provide adequate support to students would most benefit for their learning, and especially those who might be prone to drop out or failing the course (Siemens, 2013).

Additionally, analyses of student participation in online learning activities indicate that although the total amount of time spent online may not show a significant effect on students’ learning (Kupczynski, Gibson, Ice, Richardson, & Challoo, 2011), the amount of time that students spent on specific online videos (Smith-Chant, 2010), specific tasks, and the amount of delays in submitting assignments shows significant correlation with students’ performance (Cerezo, Sanchez-Santillan, & Paule-Ruiz, 2016). Also, the point in time during the course term when students access the course material is relevant, as students who only access the course site at the beginning of the course would most often end up dropping the course (Hershkovitz & Nachmias, 2011). For example, undergraduate students in compulsory courses would eventually decrease their access to the course site, leading to increased procrastination (Geri, Gafni, and Winer, 2014). Levy and Ramin (2012) found that delayed access and procrastination would increase drop out from the course.

Students make decisions to enroll in the hybrid course, gauge their probable success, assess their expected progress and actual achievement, and reinforce their work to continue in the course. At the initial stage, students will make the decision to take an online or hybrid course based on their motivations and needs. Maslow (1943) suggested a hierarchy of human needs that determines different areas of motivation. Students come with various needs, from most basic physiological needs to self-actualization, and strive to meet their needs accordingly. If a student has mismatching needs, he or she will not start, or continue working in the class that they enroll. When considering studying or beginning to study in a course, students would develop their own idea of how well they would perform in the course, or self-efficacy (Bandura, 1997), based on various factors that they can relate to. Students who have previous experience with online or hybrid learning course would establish self-efficacy, especially if they were successful or unsuccessful previously, which in turn may shed a prediction whether a student will continue to be successful or not.

Additionally, students’ performance will be influenced by their psychological needs in autonomy, competence, and relatedness, in conjunction with the way they find motivation, by intrinsic or extrinsic motivation, or no motivation (being motivated). Based on the self-determination theory, Deci and Ryan (1985) explained the relationship between different types of motivation and individual’s decision on their behavior, in meeting their psychological needs. Ryan and Deci (2000) also explained that a person’s decision is influenced in the continuum of different levels of regulation—external regulation, introjection, identified regulation, and integration of intrinsic motivation. When individuals find autonomy, competence, and relatedness in their work or tasks, they may find increased levels of intrinsic motivation, which in turn, can increase the productivity and level of achievement. The self-determination theory could provide a useful framework to examine student motivation in online or hybrid learning courses, where students’ motivation can be influenced by the delivery modes of the course that are mediated through technology.

Based on the review of current literature, below listed research questions guided this study:

1. Which of the students’ demographic and academic background variables have significant predictive value toward final course grades of students who are enrolled in a hybrid learning course in introductory psychology?

2. To what extent is student participation and persistence in online activities related to final course grades of students?

3. To what extent are the student participation and persistence in in-class activities related to final course grades of students in a hybrid learning course?

The purpose of this study was to identify and examine factors that have predictive values toward students’ success in hybrid learning courses such as: (a) students’ demographic and academic background variables, (b) students’ participation and persistence in online activities, and (c) participation and persistence in inclass learning activities in a large hybrid learning course.

The current study used quantitative analyses to identify the potential factors that have predictive values on student learning. In order to examine the potential factors and their relationships with student achievement, students’ standardized final grades were compared among different subgroups of students within the same class, based on their demographic and academic backgrounds. Students’ records of participation and persistence in online learning activities and in-class activities were analyzed to examine the predictive values toward students’ final grades.

For this study, the researchers analyzed student records from an undergraduate hybrid learning class offered at a public university located in southeastern United States. The course, Introduction to Psychology (PSYC 1101), enrolls about 300 students per section and is offered in hybrid format, incorporating online instructional videos and online quizzes following the videos to help students check their understanding. The course is structured with online sessions and face-to-face sessions alternating during the week, where students learn the topical contents online, and follow up with in-class activities and assessments. The Center for Teaching and Learning at the university led large class redesign projects, and the Psychology department participated in the project to offer an enhanced learning experience to the large number of students that the course serves. The course serves as part of the general education requirements for students who are not majoring in psychology and as a required course for those who major in psychology. The initial redesign of the course indicated positive outcomes in student learning, so the course has continued to be offered in hybrid format. At the pilot stage in spring 2011, the hybrid learning course indicated a significant improvement in student retention and final grade, with reduced percentage of students (30%) who dropped or withdrew from the course or failed, compared to the traditional face-to-face section of the course (38%) (van Wallendael et al., 2011). On the other hand, other courses in different departments or disciplines that were redesigned as hybrid courses in the initial project, found varying results. The findings from the pilot project invited questions regarding which factors predict students’ success in hybrid course format.

With the approval from the institutional review board of the university and the psychology department, a set of deidentified individual student data of 260 students enrolled in a typical section of PSYC 1101 in the hybrid learning format during spring 2016 was analyzed for this study. The psychology department has extracted and shared deidentified student data including students’ detailed course grades and the matching students’ background information at the point of admission to the university. Student data included students’ demographic data, being in the first year at the University, class standing (freshmen–senior), and major (if declared or undecided). The academic background data included standard test scores for entrance to the college, precourse GPA, weighted high school GPA, online video quiz scores, in-class quiz scores, and final grades out of a total of 765 possible points. However, background information was only available from the first year students (N = 169) who have been recently admitted. For this reason, analyses regarding students’ background variables have been conducted using the data from first-year students.

In response to the proposed research questions, quantitative data analysis methods were used:

1. To examine the relationship between students’ demographic and academic background variables and their final course grade, ANOVA procedures, and multiple regression analyses were conducted to identify the potential factors that have significant predictive value.

2. To examine the relationship between students’ online video quiz scores and their final course grades, linear regression analyses were conducted.

3. To examine the relationship between students’ in-class quiz score and their final grades, linear regression analyses were conducted.

4. Standardized multiple regression analyses were conducted to establish a predictive model toward the final grades.

5. Standardized multiple regression analyses using multiple imputations for missing values were conducted to establish a predictive model toward the final grades.

Background variables of the sample first year students were examined for significance in relationship with final grades. The data indicated that there was statistically significant difference by gender. The ANOVA results was F(1, 167) = 8.9, p < .05. Specifically, female students outperformed male students at a statistically significant level. The average final grade of female students was 3.7% higher (28.8 points out of the total 765 points) than male students’ average, with smaller standard deviation.

Students’ academic backgrounds, such as SAT writing test scores and weighted high school GPA were statistically significant in correlation with final grades and in linear regression analysis in predicting their final grades (see Table 1).

Using the student background variables that have statistically significant predictive values toward the final grade such as gender, weighted high school GPA, SAT Verbal, SAT Math, and SAT Writing, a multiple regression analysis was conducted to examine the contribution of the variables toward standardized final grades. The analysis of data indicated that the predicted contribution of the sample students’ background variables toward the standardized final grades was statistically significant (F(5, 99) = 6.40, p < .001). The predicted contribution of the student background variables toward the standardized final grades can be calculated as follows:

The result of the multiple regression analysis (Table 2) indicated that among the background variables, Weighted High School GPA, and SAT Writing scores had statistically significant (p < .05) predictive values toward the standardized final grades. The coefficient of determination (R² = .24) indicated that the variance in the background variables could be accountable for up to 24% of variance in the final grade.

Next, in order to further test the predictive model with background variables, multiple imputations (Rubin, 1987) were conducted 40 times, to impute missing values in weighted high school GPA (missing 41cases) and SAT writing scores (missing 64 cases respectively). Based on the result, the predicted contribution of the variables for the first year students was statistically significant (F(5, 6859) = 414.46, p < .000).

The predicted contribution could be calculated as follows:

The result of the multiple regression analysis using multiple imputations (Table 3) indicated that gender, weighted high school GPA, SAT Verbal, SAT Math, and SAT Writing scores had statistically significant predictive values toward the final grades among the sample students. For all variables, collinearity tests were conducted and the highest level of the VIF did not exceed 1.50. The coefficient of determination (R² = .232) indicated that the variance in the combined background variables could be accountable for up to 23.2% of variance in the final grades.

In the sample course, students watched 36 short online videos and took online video quizzes for five points following each video. The highest 33 scores were counted toward the final grades. In order to examine the relationship between the sample students’ online video quiz scores toward final grades, a correlation analysis was conducted between the online video quiz scores and final grade without the online video quiz scores. The results indicated that there was a medium positive correlation between students’ online video quiz scores and their final grades (r =.454, p < .001, N = 260). Based on the correlation, a linear regression analysis was conducted to examine the predictive value of online video quiz scores toward the final grade without the online video quiz scores included. The result of the linear regression analysis was statistically significant F(1, 258) = 67.07, p < .001. The coefficient of determination (R² = .21) indicated that the variance in online video quiz scores could be accountable for 21% of variance in the final grade without online video quiz scores.

However, regarding the online video quiz scores, there was a statistically significant difference among students in different quartiles divided by their final grades (ANOVA F(3, 256) = 41.75, p < .001). In order to locate the difference between specific groups, a Post Hoc test was conducted using Tukey HS procedure (Huck, 2008). The results from multiple comparisons of the student groups by quartiles of their final grades indicated that the first quartile was significantly different from all other quartiles, and the second was significantly different from the 4th and the 1st quartiles, while the fourth and third were not significantly different from each other. For example, the correlation between the online video quiz scores and the final grade without including the video quiz was statistically significant for the second quartile, with medium negative correlation (r = –.35, p < .01) and the result of the linear regression analysis was statistically significant (F(1, 63) = 8.97, p < .01). The coefficient of determination (R² = .13) indicated that the variance in online video quiz scores was accountable for 13% of variance in the final grade without online video quiz scores, among students in the second quartile. Considering the class as a whole group, the general direction of the correlation was that students in lower quartiles in their final grades (first through third quartiles) had a negative correlation between the online video quiz scores and the final grade without online video quiz scores when compared to the highest performing group of students (fourth quartile) with positive correlation.

Persistence in Online Learning Activities

The data revealed that the students in the sample course showed a unique pattern of participation, especially among the students who failed from the course. Among the sample students (N = 260), 12 students (4.3%) received an F grade but 5 students among them completed 36 online video quizzes out of 36 online quizzes, and 6 students completed 34 out of 36 online quizzes. In total, 11 out of 12 unsuccessful students (91.6%) completed more than the required number of quizzes (33 quizzes) throughout the semester. Among the sample students, 21 students received a D grade, of whom 13 students (62%) completed 33 or more online video quizzes, and all of the 21 students who received a D grade completed at least 23 quizzes (70%) of the required 33 quizzes. Additionally, there were 4 students who dropped out of the course (marked W as their final grade). Among them, 2 students participated in the online quiz until the 19th quiz out of 36 and another student until the 13th. These students stayed later than the early drop period, but eventually withdrew from the course, before passing the formal withdrawal deadline, to receive a W grade, rather than a possible F grade. Although the reasons and motivations of the unsuccessful students to persistently complete the online quizzes more than 70% of the required quizzes over the entire term are beyond the scope of the current study, it would be helpful to understand what components of hybrid courses and what kind of support would be most effective to facilitate students to succeed in hybrid learning courses.

In addition to the online video quizzes, the students in the sample course also took six inclass quizzes with 10 points each (total 60 points over 765 points of course total), using a personal response system, also known as “clickers.” The quizzes were used to check students’ understanding of the readings assigned during the online portion of the course, and also as a measure of attendance verification during the face-to-face sessions.

The in-class clicker quiz scores were statistically significantly correlated with the final grades not including the points from the same item (r = .38, p < .01). Based on the correlation, a linear regression analysis was conducted to examine the predictive value of the in-class clicker quiz scores toward the final grades without the in-class quiz scores. The result of the linear regression analysis was statistically significant (F(1, 258) = 44.03, p < .001) and the coefficient of determination (R² = .15) indicated that the variance in online video quiz scores could be accountable for 15% of variance in the final grade without the in-class clicker quiz scores.

The data indicated that the pattern of participation in the in-class quiz among the 12 students who received an F grade was distinct. Seven out of 12 students (58.3%) started the course by missing the first in-class quiz, and one student never completed any of the inclass quizzes. Also, four students completed three to five quizzes, and two students completed the quiz only once. As a whole, seven students (58.3%) completed less than a half of the quizzes, and received an F grade at the end of the course. The data indicated that students’ participation patterns for in-class quizzes was positively correlated (r = .38) with their final grades and had a statistically significant predictive value, consistent with previous findings of persistence patterns of unsuccessful students (Dziuban, Moskal, Cassisi, & Fawcett, 2016; Levy & Ramin, 2012).

The results described the relationships between students’ demographic and academic background variables and their final course grades. The result of the multiple regression analyses using standard scores indicated that gender, SAT scores, and weighted high school GPA had statistically significant predictive values.

Regarding the persistence in taking the online video quizzes, the students in the sample course had extremely high participation, and even the majority of the students who failed in the course participated in the activity at near-complete level. Additionally, in-class quiz scores were also correlated with the final grade, and the direction of the correlation was consistent with the general trends of students’ final grades. The sample students’ behavior data for the in-class quiz was similar to those reported previously for low-performing students, and the in-class quiz had a significant predictive value.

The findings from this study implied that students who had stronger academic records as early as in high school would continue to do well in hybrid learning courses. For example, the sample students’ weighted high school GPA and SAT test scores had statistically significant predictive values toward the final grade. The results from the multiple regression analyses concurred with the findings of Keller et al. (2009) that students’ prior academic background variables are more relevant in predicting their success in hybrid learning courses, rather than just the course delivery formats.

As for the demographic backgrounds, a significant predictive value was associated with gender, but other variables did not have significant predictive value. Therefore, in addition to the delivery format of the course (Holley & Oliver, 2010), students’ academic background variables had significant predictive value toward students’ final grades (Deschacht & Goeman, 2015).

As students watched online videos and completed the quizzes during the online portion of the sample course, the online video quiz scores carried a significant predictive value toward students’ final grades. The results of a linear regression analysis indicated a correlation between the students’ online video quiz scores and their final grades and showed that the video quiz scores had a statistically significant predictive value toward their success in the sample course. Additionally, the data indicated that the majority of low-achieving students completed the online video quizzes more than required and many of them also attempted multiple times to attain high scores. It may be possible that the low-achieving students in the sample course were willing to invest their efforts to get the best scores they could achieve, as they considered that the quizzes are useful for their study and success in the course, and with low stakes for trying. The findings from this study showed a unique tendency, in which the lower performing students actively participated in the online learning activities (in this case, the online video quizzes), compared to the findings from the existing studies reporting that less successful online students show lack of participation starting as early as the second week of the course term (Dziuban et al., 2016).

Considering the predictive value of the online video quizzes, it is noteworthy that during the course redesign process, the psychology department had formed a curriculum development team to review the relevant video packages that were available for adoption for the sample course. The curriculum development team not only reviewed the topical contents of the videos, but also envisioned students’ use of the videos for their learning in the course and thus focused on students’ perspectives. The selection concurred with the recommendations from the existing literature suggesting that courses would be most effective when online videos were followed by reflective activities (Yilmiz & Keser, 2016) or online quizzes (Tune, Sturek, & Basile, 2013). In this case, the selection of the videos by the curriculum development team may have had a positive influence on students’ decisions on the usefulness of the videos and whether to use them for their learning. Also, as the department selected the particular series of the videos in favor of the online video quizzes that immediately followed the clips, the curriculum development team embraced the activities to the overall structure of the course. The existing literature supports that careful integration of learning activities, such as followed in this sample course would facilitate students’ learning (Zacharis, 2015).

In this course, the online video quizzes served as a ‘safe’ activity to facilitate students’ learning. The quizzes were low-stakes in terms of the weight of individual quiz and each question, with two attempts and no time limit. Therefore, the quizzes had relatively low cognitive and affective barrier, while allowing students to keep track of their learning and follow the pace of the course. This intentional design and student-focused planning of the course would distinguish the course from other courses that lack adequate instructional design (Willging & Johnson, 2009).

To further investigate the students’ motivation and decision to complete the online learning activities and the effects resulting from their work, it would be helpful to conduct longitudinal studies following individual students’ performances in different courses. Further investigation is also warranted to examine the factors that affect students’ success in different course formats and delivery modes.

The data from this study indicated that there were within-group differences among the students who were in different subgroups divided by their final grades in the course. Compared to Owston et al.’s study (2013) reporting that lower performing students performed worse in hybrid learning courses, the majority of sample students, and especially lower performing students in this study completed the online video quizzes persistently to contribute toward their final grade. It would be helpful to examine if the pattern of within-group difference is also present across the different delivery formats or if it is a unique pattern in hybrid courses. Also, it would be helpful to examine if there is a difference in student participation due to the nature and type of online activities (Olitsky & Cosgrove, 2014), and the gap in the amount of time and efforts that students actually invest. For example, some students tend to complete the tasks that they perceive to be easier than those that would require more substantive work, especially among the lower performing students. From this perspective, future research can be suggested to examine student perceptions of and participation in online activities of different types and nature.

Due to the limited scope of this study, it would be necessary to review the limitations that were applied. First, the student data used in the study was collected from a single section of a hybrid learning course. Since a convenient sample was used in this study, the result of the study cannot be generalized to other disciplines or institutions. Another limitation of the study was that the data used in this study were limited to quantitative information only.

Finally, the data and results of analyses in this study cannot be used to imply the causal effects of the variables. The results of correlation and regression analysis are limited to describe the potential relationship only and cannot be assumed for the causal effects.

The findings imply that students who had stronger academic records as early as in high school would continue to do well in hybrid learning courses. However, in the sample course, some students who were not as successful in their earlier study in high school or in college persistently completed online learning activities, carefully designed for the sample course. The data and results of analyses indicated that persistence in completing online learning activities had a significant predictive value in regard to students’ success in a hybrid learning course. To further investigate the students’ motivation and decision to complete the online learning activities and the effects resulting from their work, it would be helpful to conduct longitudinal studies following individual students’ performances in different courses to examine the factors that affect students’ success in different course formats and delivery modes.

While above questions initially focus on the course delivery and course components, it may be also helpful to examine what motivates students to put their efforts in the online activities within hybrid learning courses, or what strategies they use to succeed in hybrid course. A qualitative approach can be useful to learn about students’ decisions and rationale behind their decisions.

Using the data from a sample hybrid learning course, this study investigated to see if there were variables that can predict students’ achievement in the hybrid course, in terms of their demographic and academic background variables such as gender, high school GPA, and SAT scores. During the course, the students also indicated different patterns of participation and persistence in online activities, creating within-group differences. The within-group differences may reduce the size of effect as a whole class, however it has been brought to attention that many students in lower achieving group invested their effort to complete the online activity (online video quizzes), throughout the course.

Persistent participation among the students in lower achieving group was evident, even among the students who failed the course or dropped from the course. Persistent participation in online activities may have helped the lower achieving students improve their performance, which in turn contributed toward facilitating successful completion of hybrid learning course. Further studies are suggested to investigate the variables that may facilitate student learning in hybrid learning courses, and what motivates students to be persistent in the course and improve their learning.