Abstract
The current study investigated whether collaborative quizzing impacted quiz performance, overall grades, exam grades, pass/fail rates, and attendance in upper-level psychology courses. Students enrolled in two sections of upper-level psychology courses completed collaborative quizzes, whereas students enrolled in two noncollaborative sections completed quizzes individually. Students enrolled in the collaborative classes maintained quiz performance from beginning to end of the semester, whereas students enrolled in the noncollaborative classes demonstrated a general decline in quiz grades. There was also a slight increase in overall course grades for students enrolled in collaborative compared to noncollaborative classes; however, there was no difference in exam grades, pass/fail rates, and attendance between the collaborative and noncollaborative classes. Collaborative quizzes may encourage a deeper understanding of course material and/or encourage students to come prepared for the group quiz activities.
Group work is required in many facets of life, such as work and academics. In fact, many employers rate this ability as critical for workplace success (Casner-Lotto & Barrington, 2006; Landrum & Harrold, 2003; Robles, 2012). During the past decade, pedagogical research on collaborative learning techniques in education has increased (Delucchi, 2007; Drouin, 2010; Gorvine & Smith, 2015; Koçak, 2008; Schroeder, Scott, Tolson, Huang, & Lee, 2007; Tsay & Brady, 2012; Woodzicka, Ford, Caudill, & Ohanmamooreni, 2015). Many of these studies demonstrate that collaboration on assignments that have no or minimal impact on students’ overall grade, such as group projects, group study, and lab reports, tends to have a positive impact on learning outcomes, such as exam grades or overall course grades. Fewer studies have examined the impact of collaboration on more substantial assignments (i.e., assignments that make up a large percentage of the student’s overall grade) on learning outcomes (Tsay & Brady, 2012).
Learning outcomes for collaborative studies can be separated into three groups: analytical processes (Gokhale, 1995; Lazonder, 2005; Woodzicka et al., 2015), attitudes (Drouin, 2010; Koçak, 2008; Stein, Cislo, & Ward, 1994; Terenzini, Cabrera, Colbeck, Parente, & Bjorklund, 2001; Tsay & Brady, 2012), and course material retention (Delucchi, 2007; Drouin, 2010; Gorvine & Smith, 2015; Perkins & Saris, 2001; Schroeder et al., 2007; Shibley & Zimmaro, 2002; Smith, Hinckley, & Volk, 1991; Tsay & Brady, 2012). Students are more likely to engage in analytical thinking with collaborative activities. For example, collaborations helped students understand the research process (Woodzicka et al., 2015), increased performance on critical-thinking application-based exam questions (Gokhale, 1995), and helped students use more sophisticated web-based search strategies (Lazonder, 2005). Collaboration also positively influences students’ perception of learning (Terenzini et al., 2001; Tsay & Brady, 2012), satisfaction with the collaborative activity (Drouin, 2010), and social skills (Stein et al., 1994). Koçak (2008) found that collaborative learning decreased loneliness and social anxiety and increased happiness.
Considering course material retention, positive outcomes include better performance when students learned new material using a jigsaw classroom compared to learning individually (Perkins & Saris, 2001; Smith et al., 1991). In addition, solving problems and completing lab experiments in groups increased overall course grades (Gorvine & Smith, 2015; Shibley & Zimmaro, 2002). Working on graded group projects (Delucchi, 2007; Tsay & Brady, 2012) or participating in a nongraded study group (Drouin, 2010) increased exam scores. In fact, collaborative group work (e.g., working in groups on lab exercises) has the highest impact on student test performance in science classes compared to direct instruction, posing questions at the start of the lecture, capturing attention by alerting students to the intent of the lesson, practice with physical objects, and relating learning material to students’ previous experiences (Schroeder et al., 2007). With the exception of Tsay and Brady, in which a majority of the final course grade was based on group work, most collaborative activities are not graded or minimally contribute to the final course grade. Thus, in general, the actual collaborative assignment does not impact or directly contribute to students’ grades. It is possible that if the actual collaboration is graded, students may benefit more from that collaboration. This research focuses on how collaboration on quizzes, which impact students’ grades, influences quiz performance, exam grades, final course grades, and pass/fail rates. Another goal of this research is to determine how graded collaborative quizzes influence class attendance. Class attendance is a strong predictor of academic performance, outweighing factors such as SAT scores and study skills (Credé, Roch, & Kieszczynka, 2010). Likewise, absenteeism is related to decreased exam scores (Lin & Chen 2006; Marburger, 2001) and poor overall course performance (Thatcher, Fridjhon & Cockcroft, 2007). Students report attendance as being important to their performance (Gump, 2006), but life events and motivational factors can impede class attendance (Fjortoft, 2005; Van Blerkom, 1992). It is possible that pedagogical interventions such as implementing collaborative quizzes can improve motivational reasons to attend class. Alternatively, collaborations may foster a sense of community among the students (Top, 2012) and/or increase the feeling of belonging, which is associated with student success in various contexts (Yeager & Walton, 2011).
The current study investigated the impact of collaborative quizzing on quiz performance, overall course performance, exam grades, pass/fail rates, and student attendance. Given the interactive nature of the quiz design, we expected that the students enrolled in the collaborative sections would have higher quiz grades, higher overall course grades, higher pass rates, and lower fail rates than students enrolled in noncollaborative sections. Additionally, we hypothesized that the students enrolled in collaborative sections would attend more often than the students enrolled in the noncollaborative sections.
Method
Participants
One hundred and seventy-three students enrolled in upper-level psychology courses (i.e., two sections of Statistics and two sections of Drugs and the Nervous System) at Indiana University Northwest participated in this study. The data from one student were excluded due to incomplete attendance data, and 12 students withdrew before the courses were completed. One section of each course either participated in collaborative quizzes or served as a control. Collaborative and noncollaborative sections were determined by the instructor. Thus, there were 78 students who enrolled in and completed the collaborative classes and 82 students who enrolled in and completed the noncollaborative classes.
Materials and Procedures
All classes were taught by one instructor. Both sections of Statistics were taught in the summer; both sections of Drugs in the Nervous System were taught in the spring semester. In all classes, students completed up to five quizzes evenly spaced throughout the semester. Students were given 10 minutes to complete 10 multiple-choice questions individually. Quizzes were scheduled, noncumulative, and covered recently discussed material. Across all sections, quizzes were worth approximately 15% of the students’ overall course grade, exams were worth approximately 55% of the students’ overall course grade, and the remaining 30% of the total course grade consisted of papers or homework assignments, depending on course.
In the collaborative quiz classes, after individual quiz sheets were submitted to the instructor, the students self-organized into groups of three to six participants. These groups were maintained for the entire semester. Each group was given one fresh quiz sheet and an additional 10 minutes to discuss the questions and complete the quiz collaboratively. The instructor first graded the individual quizzes, then, as an additional incentive, gave each passing student (individual quiz grade > 50%) bonus points based on group performance. An individual quiz mean based on all passing scores was calculated and compared to the group quiz grade. If the group quiz score was greater than the individual quiz mean, the difference was added to each individual passing score as a bonus.
Data Organization
Quiz performance
Quiz performance was figured by computed by comparing the individual quiz percentages at the beginning of the semester to the end of the semester. Change in quiz performance was used as a measure due to differences in initial baseline differences in quiz scores between control and experimental groups. As each course varied in the number of quizzes offered per semester, the beginning of the semester was defined as the first quiz for all sections. The end of the semester was defined as either the final quiz (for sections with only three quizzes offered) or the average of the last two quizzes of the semester (for sections with five quizzes offered). Thus, each change score was figured as the beginning of semester quiz percentage subtracted from the end of semester quiz percentage. Means of the individual change in quiz performance scores were computed for collaborative and noncollaborative classes. The additional group incentive bonus points were not included in these calculations. On average, for students enrolled in the collaborative sections, the point incentive for the entire semester was equivalent to 8 points of 500 (1.6% of the total course grade).
Course performance
Overall course grade was based out of 500 total points for the semester. For the collaborative classes, all group quiz incentive bonus points were removed from the overall total grade calculations. Means of the overall course grades were calculated for each the collaborative and noncollaborative classes.
Exam grades
Average exam grade was calculated as the average percentage of three noncumulative exams for the semester. Means of the average exam percentages were calculated for each of the collaborative and noncollaborative classes.
Pass/fail rates
The number of students passing and failing in each condition was counted. A passing grade was considered to be a 70% or greater in the overall course grade, and a failing or DFW (D/F/Withdrawal) grade was considered to be either a withdrawal from the course or if the student earned less than a 70% in the overall course grade.
Attendance
Attendance was assessed by either student response systems (i.e., clickers) or through free web-based polling software (i.e., socrative.com) and was not factored into overall course grade calculations. The attendance for each student was computed as a percentage of the days actively participating out of all possible classes for that particular semester. Means of the individual attendance percentages were calculated for each the collaborative and noncollaborative classes.
Data Analysis
The data are presented as group means (n = 78 and 82 for the collaborative and noncollaborative sections, respectively). Most data were analyzed by t tests for independent samples using class type (collaborative and noncollaborative) as the independent measure. The pass/fail rate data were analyzed by a χ2 test for independence.
Results
Quiz Performance
An independent samples t test revealed a significant difference in the quiz change percentages (Figure 1) for students in the noncollaborative classes (M = −9.57%, SD = 30.63%) compared to students in the collaborative classes (M = 0.71%, SD = 28.92%); t(158) = −2.18, p = 0.031, d = −0.345, 95% CI [−19.59, −0.97].
Students enrolled in collaborative classes demonstrated a significant increase in quiz performance across the course of the semester compared to students enrolled in noncollaborative classes (group means ± standard error of the mean (SEM); independent samples t test).
Course Performance
An independent samples t test revealed a marginally significant difference in the overall grade point totals (Figure 2) for students in the noncollaborative classes (M = 374.8, SD = 85.83) and students in the collaborative classes (M = 397.15, SD = 69.94); t(158) = −1.80, p = 0.074, d = −0.285, 95% CI [−46.87, 2.27].
Students enrolled in collaborative classes demonstrated a marginally significant increase in overall course grades compared to students enrolled in noncollaborative classes (group means ± standard error of the mean (SEM); independent samples t test).
Exam Grades
An independent samples t test revealed no significant difference in the average exam grades (Figure 3) for students in the noncollaborative classes (M = 74.53%, SD = 18.38%) and students in the collaborative classes (M = 73.93%, SD = 15.90%); t(158) = 0.219, p = 0.827, d = 0.035, 95% CI [−4.78, 5.98].
Students enrolled in collaborative classes did not differ in average exam grade when compared to students enrolled in noncollaborative classes (group means ± standard error of the mean (SEM); independent samples t test).
Pass/Fail Rates
A 2 (course: collaborative, noncollaborative) × 2 (grade: pass, fail/DFW) χ2 test of independence was computed (Table 1). A marginally significant interaction was found, χ2(1) = 2.576, p = 0.109. Tests of simple effects revealed that there was no statistical difference in frequencies of passing rates between collaborative and noncollaborative courses, χ2(1, N = 128) = 0.781, p = 0.377. Likewise, there was no statistical difference in frequencies of failing/DFW rates between collaborative and noncollaborative courses, χ2(1, N = 45) = 1.80, p = 0.180.
Frequencies of Pass and Fail/DFW Rates for Collaborative and Noncollaborative Classes.
Attendance
An independent samples t test revealed no significant difference in the attendance (Figure 4) for students in the noncollaborative classes (M = 74.71%, SD = 23.49%) and students in the collaborative classes (M = 74.27%, SD = 18.66%); t(158) = 0.13, p = 0.897, d = 0.021, 95% CI [−6.21, 7.08].
Students enrolled in collaborative classes did not differ in attendance when compared to students enrolled in noncollaborative classes (group means ± standard error of the mean (SEM); independent samples t test).
Discussion
This study investigated the impact of collaborative quizzes on quiz performance, overall course performance, exam performance, pass/fail rates, and student attendance. Consistent with predictions, students enrolled in collaborative classes performed better on quizzes and had marginally higher overall course grades compared to students enrolled in noncollaborative classes. These effects are consistent with prior research that shows collaborative activities improve content retention (Delucchi, 2007; Drouin, 2010; Gorvine & Smith, 2015; Perkins & Saris, 2001; Schroeder et al., 2007; Shibley & Zimmaro, 2002; Smith et al., 1991; Tsay & Brady, 2012). Contrary to predictions, students enrolled in collaborative classes did not have higher exam grades, did not differ in their pass/fail rates, and did not attend more classes than students enrolled in noncollaborative classes.
The most striking difference between the noncollaborative and collaborative classes is the effect on quiz performance. Declining performance is common for the courses used in this study (i.e., Statistics and Drugs and the Nervous System), possibly because successful performance on end-of-semester material is dependent on a good understanding of beginning-of-semester material. The results from the current study suggest that the preparations made to engage in the collaborative nature of the quiz exercises may prevent this typical decline in performance.
This absence of a decline in quiz performance in collaborative classes may be due to the cumulative effect of students understanding the material at a deeper level, more motivation, and/or dispersion of cognitive load during quizzes. Students instructed to learn new material to teach others retained more material than students instructed to learn the material for a test (Bargh & Schul, 1980), presumably because students who taught had a more organized understanding of the content. Thus, it is possible that students preparing to collaborate created a more organized knowledge structure of the course content compared to students not preparing to collaborate. Students in collaborative sections are potentially more motivated to study, with the knowledge that they will be expected to contribute to the group discussion. In fact, collaborations help promote favorable attitudes toward learning (Springer, Stanne, & Donovan, 1999). Last, Kirschner, Paas, and Kirschner (2009) found that working in groups for a complex problem-solving task helped disperse the cognitive load, and subsequently, members of the groups formed more coherent memories of the problems. Therefore, it is possible that group quizzes helped create better memory of the material by dispersing the cognitive load across participants.
Students who collaborated had course grades approximately 5% higher compared to the noncollaborating students, although this effect was marginal. The small effect on overall grades may be due individual student characteristics. Some students, due to personality, knowledge, or experience, may be better equipped to benefit from the opportunities collaborative quizzing may offer. Indeed, implementing collaborations alone is useful, but the contribution of the individual student in the collaborative process is also important. The more students contribute to collaborations, the higher their performance in the course (Kirschner, Paas, & Kirschner, 2009; Tsay & Brady, 2012). Future studies should assess the quantity and/or quality of each individual student’s contributions to the collaboration. Alternatively, this small increase in course grades may be accounted for by the gains in quiz scores attained by the students in the collaborative course. Nonetheless, it appears that students in the collaborative classes benefited overall, even if this gain was small. If we consider another marker of student success, the pass/fail rate, we also see a small, albeit marginally significant effect. In the noncollaborative classes, the failure or DFW rate was 31.39%; however, in the collaborative classes, this rate drops to 20.69%. Thus, although it appears that collaborative quizzes do not have a large effect overall grades at first glance, collaborations may help keep students in the course and/or receive a passing grade.
Collaborative quizzes did not appear to influence exam grades. Although it appears that collaborative assignments appear to strongly influence the actual graded assignment, these effects do not transfer to other impactful assignments. Alternatively, as the quizzes were multiple choice, there may have been a specific transfer to only multiple-choice items on the exams. However, if there is a skill or knowledge transfer to these types of items on the exam, this effect may have been masked by other question types on the exams (e.g., fill-in-the blank, calculations, or essay questions).
Even though consistent class attendance influences various markers related to course success (Credé et al., 2010; Lin & Chen 2006; Marburger, 2001; Thatcher et al., 2007), it appears that collaborative quizzes do not encourage students to attend class more. The collaborative quizzes were well received by the students. Approximately 31% of respondents of the end-of-semester evaluations in the collaborative classes spontaneously made a positive statement about the quizzes in the open-ended section. The students also were more likely to rate their group interactions (4.432 out of 5 points in the collaborative classes vs. 4.240 in the noncollaborative classes) and self-perceived learning (4.208 in collaborative classes vs. 4.135 in the noncollaborative classes) more favorably in the numeric rating section of the course evaluations. However, these perceptions do not appear to outweigh other factors that influence class attendance. Fjortoft (2005) and Van Blerkom (1992) found that students’ self-reported factors such as illness, work for other courses, emergencies, and employment were major reasons for not attending. It appears that collaboration does not overcome these factors.
It should be noted that small sample sizes and self-selection of the specific courses may have influenced the results. Thus, the probability of sampling outliers is much higher. However, cumulative GPA for collaborative and control classes were relatively similar (3.027 vs. 2.915, respectively) before starting the courses under investigation. Nonetheless, one must interpret this study with caution. Future research should attempt to replicate these findings.
Collaborative quizzing is a quick, easy way to incorporate interactions in the classroom, and evaluations suggest students enjoy this technique. Because it may encourage deeper understanding of the course material and is easy to implement, the benefits of collaborative quizzing greatly outweigh the limitations. However, these benefits do not appear to transfer to other impactful assignments, like exams.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.