Abstract
The opportunity to give oral presentations is important because it helps develop skills that are necessary for many aspects of professional careers. Competencies for oral presentations include appropriate content, design and organization of the presentation, and delivery skills. Self-assessment has been shown to help students develop meta-cognitive skills that can be used in all aspects of higher education, and it has been shown to improve student achievement in a range of different academic areas. In this study, students were recorded while giving two PowerPoint presentations on a topic of their choice. One section was required to complete a self-assessment rubric of their first presentation, while the other section was encouraged to watch themselves, but it was not required. All students also received instructor and peer evaluations. For their second presentations, the students who completed the self-assessment rubric received higher scores than the other group; significant differences were seen in content explanation, timing, and speech skills. Peer assessment scores correlated moderately well with the instructor scores, whereas the self-assessment scores did not. Despite the lack of congruence between self- and instructor scores, the combination of instructor, peer, and self-assessment resulted in consistently high-quality presentations. The students found the multiple types of feedback, use of video technology, and the opportunity to repeat their presentations useful.
The use of video to record presentations and other training activities
Many reports support the idea that students, and adults in general, are anxious about public speaking (Lucchetti et al., 2009). For example, Smith and Sodano (2011) report that around half of students say they are anxious during the preparation process and feel tense and tongue-tied during the presentation itself. Public speaking and giving presentations cover a range of related competencies (e.g. organizational skills, professional language use, eye contact, and body language) that are necessary for successful professionals in careers typically pursued by undergraduate or postgraduate students.
Video recording has been used for many decades to allow students to watch and assess their performances (Bourhis and Allen, 1998; Quigley and Nyquist, 1992). This technology has been used in speech/communication classes and also during career training, for teachers, medical health professionals, social workers, and lawyers (Baecher et al., 2013; Grant et al., 2010; Hawkins et al., 2012; Hecimovich et al., 2010; Mort and Hansen, 2010; Murphy and Barry, 2016; Pereira et al., 2014; Reitmeier and Vrchota, 2009; Seidel et al., 2011; Yoo et al., 2009). The use of video technology during a performance allows students to gain the observers’ perspective (Quigley and Nyquist, 1992). Having students engage in self-assessment is widely held to have the benefits of self-reflection and engagement in meta-cognitive activity that encourages the development of a life-long learner (Falchikov and Boud, 1989; Yoo et al., 2009).
There are two broad benefits to the use of video technology for self-review in presentation/performance classes: skills development and improved attitude of the students toward their presentation abilities. Self-efficacy in presentation skills has been shown to be important in achievement (De Grez et al., 2009a, 2010; Lucchetti et al., 2009). But obviously, tangible evidence of skills improvement is critical. Quigley and Nyquist (1992) reviewed literature on videotaping university students performing different activities, for example, in speech classes and in classes of students pursuing social work, dental, and law careers. They found that in some cases skills improved when students watched themselves, but in some examples they did not. Quigley and Nyquist (1992) concluded that self-assessment during videotaping can be most effective when specific skills are being addressed and when comparisons with others are made. Bourhis and Allen (1998) conducted a meta-analysis of 12 reports on videoing presentations during a range of university classes and concluded that it is an overall useful learning experience. In addition, many studies report a positive attitude of the students toward the course when video technology is used (e.g. Barry, 2012; Bourhis and Allen, 1998; Grant et al., 2010; Murphy and Barry, 2016; Seidel et al., 2011; Yoo et al., 2009). Although some increased student anxiety might be expected from the use of this technology, Bourhis and Allen (1998) reported that the effect is negligible.
There are more recent examples in the literature in which, as found by Quigley and Nyquist (1992), the effect of self-assessment in performance tasks is mixed. Hecimovich et al. (2010) conducted a study in which postgraduate chiropractic students performed a history-taking exercise and then either received instructor assessment or self-assessed. Self was sufficient for the students to perform as well as those who received instructor feedback on a final class assessment. But for students at an earlier stage in their higher education, this was not the case (De Grez et al., 2009a); those who received instructor feedback improved the most in oral presentations. De Grez et al. (2009a) found that there was no difference in development of presentation skills between students who self-reflected on their performance compared with those who did not, although they did observe an improvement in content skills when students focused on specific areas of presentation.
Self versus instructor assessment
Self-assessment of presentations and other evaluated activities is often compared with instructor or peer assessments. Frequently, but not always, there is only low congruence between self and instructor scores (Falchikov and Boud, 1989; Lew et al., 2010; Reitmeier and Vrchota, 2009), although this does not necessarily mean it is not still a valuable exercise (Tan, 2012). This low congruence has been observed in many research studies (reviewed by Falchikov and Boud, 1989; Lew et al., 2010; Topping, 1998), including performance-specific activities in which students saw themselves on video (De Grez et al., 2009a; Langan et al., 2008; Mort and Hansen, 2010; Pereira et al., 2014; Quigley and Nyquist, 1992). Indeed, this lack of congruence extends beyond traditional academic skills (Heidmeier and Moser, 2009; Kruger and Dunning, 1999; Lew et al., 2010).
The low correlation between self and instructor scores can in part be linked to level of achievement. Topping (1998) and Mort and Hansen (2010) observed that higher achieving students tended to under-rate themselves or be unreliable. In contrast, underachieving students tend to over-rate themselves (De Grez et al., 2012; Langan et al., 2008; Mort and Hansen, 2010). Lew et al. (2010) found that correlations between self and instructor assessment were better for higher achieving than lower achieving students, although it was not mentioned whether either group tended to undergrade or overgrade themselves. This poor ability to accurately self-assess could be due to a poor understanding of the quality required or an over-inflated opinion of self. The lack of ability to self-assess objectively is thought to stem from a lack of meta-cognitive skills (Kruger and Dunning, 1999; Langan et al., 2008; Lew et al., 2010).
Formative self-assessment using video recordings
The emphasis on correlation between self and instructor scoring tends to focus the use of self-assessment into a summative role: can it be used to replace instructor scoring and provide students with more empowerment within the class structure? However, to foster the reflective aspects of self-assessment it is more useful to consider it as a type of formative assessment. In this context, it is important when conducting research to determine whether it has an impact on student performance in subsequent related activities. There are studies in which students gave multiple presentations in the same class with the same general themes (De Grez et al., 2009a, 2009b), and improvement is seen. Repeated performance with self-assessment, but no other feedback, gives mixed results in terms of improvement of skills (reviewed by Quigley and Nyquist, 1992). Mulac (1974) found that students who did not receive any feedback after presentations still improved through the semester, although not as much who were required to self-assess their presentations using video. Thus, even un-assessed practice has some value.
For oral presentations, the use of video technology allows focus on particular aspects of presentations; content requirements can be separated from delivery skills. In a study on public speaking skills in undergraduates, Dunbar et al. (2006) found that students tended to be at the level expected by national standards (United States) for topic choice, communicating purpose, and organization, but were below the standard for delivery skills. This suggests that delivery skills should be emphasized just as much as content skills even in classes for which communication is not the sole focus (Smith and Sodano, 2011). Interestingly, De Grez et al. (2009a) found that through three rounds of short presentations, content was improved in the second presentation and delivery skills were improved more during the third presentation. Similarly, De Grez et al. (2009b) observed more improvement in content than delivery. These studies suggest that content areas in a presentation are somewhat easier for students to address appropriately than the intangible delivery requirements, which require more maturity and experience. However, it is worth noting that the student participants in all three studies were at the lower level.
In order to address whether self-assessment of presentations does in fact help students, it would be useful to provide them with an opportunity to focus on improving the same presentation, rather than starting again with new content. The studies of Dunbar et al. (2006) and De Grez et al. (2009a, 2009b) suggest that because presentation content is an area in which students tend to be more competent, when they are assigned a new content area they will predominantly focus on that. By keeping the content related, this would ensure more attention on other aspects of the presentation, that is, the organization and delivery skills. A survey of the literature indicates that this method of examining the role of self-assessment using video-recording presentations has not been attempted.
Peer assessment
In addition to self-assessment, peer assessment is used widely in education. When students engage in peer assessment, they are spending more time focused on the class content/skills, and it gives students more “ownership” of the class and responsibility toward their peers. From the perspective of the assessed, advantages include receiving more feedback, reducing any perceived instructor bias, and again spending more time reading and considering the multiple sources of feedback. Importantly, there is empirical evidence that this type of activity does help improve student learning (Topping, 2009), and students report learning as much from assessing the writing of their peers as from writing themselves (Mulder et al., 2014). In the specific area of oral presentations, peer assessment has been shown to improve student confidence during presentations (Topping, 1998), and students regard the use of peer assessment favorably (De Grez et al., 2010).
Peer assessment in a wide range of student activities (writing, professional performance, and group and individual presentations) has been compared with instructor assessment, and generally, the correlations are good, better than self-instructor. Falchikov and Goldfinch (2000) conducted a meta-analysis of instructor versus peer assessment studies in a range of student activities including oral presentations. They found that the highest correlations between peer and instructor scoring occurred when global rather than multidimensional criteria scoring was used.
Although peer assessment is used widely (Topping, 1998) and is regarded positively from the perspective of the reviewer and the person being reviewed (De Grez et al., 2010; Langan et al., 2008; Topping, 1998), various issues have been observed in the literature. Some studies have identified that the range of peer grades is narrower than that of the instructor, that is, peers are less likely to give as high or low scores as the instructor (Langan et al., 2005; Pereira et al., 2014). In some studies, it has been observed that there is a tendency for the students to be more generous than the instructors (De Grez et al., 2012; Langan et al., 2005; Lew et al., 2010; Reitmeier and Vrchota, 2009). Langan et al. (2005) also noticed that a small group of “popular” students were awarded even higher grades by their peers. Finally, while the use of peer-grading is perceived positively by students, inclusion of instructor grades is still appreciated (Mulder et al., 2014), suggesting that the “expert” opinion is still valued.
The theoretical framework for this research comes from a socio-cognitive perspective. Learning comes from observing others doing something, trying it out, and receiving feedback. Students in higher education watch professors give lectures on a daily basis, and in a class using oral presentations, they have the opportunity to watch and evaluate each other (whether formally as in this class or informally when no peer assessment is required). As a task is performed and feedback is received, the theory suggests that a process of self-calibration occurs. From the feedback received (whether from the instructor, peer, or self), adjustments are made so that the next attempt will be closer to the desired level of competence. Thus, this theory emphasizes the cycle of observation, practice, and feedback. In this study, because students are required to conduct peer and self-assessment, they have ample opportunity for observing and reflecting on the attributes of a good presentation. Furthermore, because students have the chance to perform twice, they have the opportunity to put into practice the self-reflection that has occurred as a result of the feedback.
From the literature, it is clear that students value both instructor and peer feedback; therefore, rather than comparing self-assessment with other forms in an either/or situation (e.g. De Grez et al., 2009a, Hecimovich et al., 2010), it is also pertinent to determine whether it would be useful in addition to these other forms of feedback. Hence, this study asked the question, “If students are required to self-assess their presentations (in addition to receiving instructor and peer evaluations), would the skills demonstrated in their second presentations be improved more in comparison with those for whom formal self-assessment is not required?”
Methods
Participants and class structure
Participants were 39 students enrolled in two sections of a class called Professional Communications in Biology at a US public university. The first section (19 students) was the control group and the second section the intervention group (20 students). All students were in the biology department, and the study was conducted during the Spring 2014 term. The male:female ratio was 9:10 in the control group and 6:14 in the intervention group. Other background information is provided in Table 2.
The assignments given to the students were all part of the regular course work. The students were required to write a literature review on a topic in biology of their choice; they ranged from ecology to medical treatments. They then gave two individual, 10-minute presentations based on their literature review. An anonymous questionnaire was used to establish the students’ background in giving PowerPoint presentations and their opinions on the assessment process.
PowerPoint presentations
The presentations used Microsoft PowerPoint and were recorded using Tegrity lecture capture software (McGraw Hill Education). The recordings (slides and view of presenter) were accessible to the instructor and enrolled students through the university learning management system. The instructor and students could access all presentations at any time during the semester. After the first set of presentations, the students in both groups received graded rubrics (Table 1), including written comments, from the instructor and (anonymously) 6–12 of their peers. In the control group, the students were also told to watch themselves afterward and “see how they did” and compare their observations with the assessments they received. The intervention group was required to complete the presentation rubric for their own performance. The peer and self rubrics were turned in for copying and then given to the presenters so that they could use all the feedback in preparation for their second presentation.
Rubric used to grade presentations.
The rubric used was based on content requirements for the class plus organizational and delivery criteria used in studies of similar focus and for which reliability was demonstrated (De Grez et al., 2009a; Langan et al., 2005; Smith and Sodano, 2011). The same rubric was used for instructor, self, and peer evaluations (Table 1). It was divided into 15 points: 5 were specific for the class assignment and 10 more generally applicable to all presentations. Each criterion was graded using a 1–5 point Likert scale, 1 representing “poor/absent” and 5 “no change needed.” The grade received from the first presentation was an average of the instructor grade and the average peer grade.
The second presentation grade was based on improvement in the presentation including content, structure of the presentation, and delivery skills. Points were assigned based on how well each student addressed feedback received from the first presentation. Thus, even if the student performed quite well and received a high grade for their first presentation, if they did not address the comments received, they would receive a low grade for the second round. At the side of the numerical score for each criterion, specific details were provided describing areas for improvement. This was then used as a checklist for scoring the second presentation. This method of grading was also used in two previous written assignments.
Three other instructors who had previously taught the class were invited to grade 5 of the presentations. The Tegrity system was used to watch the presentations (scores were not given to the students). Pearson’s r values for the primary instructor scores versus the other instructors were 0.89, 0.51, and 0.77. The lower correlation between the primary instructor and the second guest evaluator was due to a very different score for one particular student. When the data for that student were removed from the analysis, the r value was 0.82. These data show reliability for the primary instructor consistent with previous research (Dunbar et al., 2006).
Data analysis
Data were analyzed using Microsoft Excel. Personal identifiers were removed once grades were transferred from the university gradebook software. The study received review exempt status from the Office of Research Assurances.
Results
These show that for their second presentations, the students who completed the self-assessment rubric received, on average, higher scores than the control group. Peer assessment scores correlated moderately well with the instructor scores, whereas the self-assessment scores did not. Despite the lack of correlation between self- and instructor scores, the combination of instructor, peer, and self-assessment resulted in consistently high-quality presentations.
Student backgrounds
The mean age and proportion of juniors to seniors were very similar for both groups (see Table 2). Questions were also asked to determine previous experience the students had with PowerPoint presentations and being recorded. The majority of students reported first doing this kind of presentation at middle or high school (77% control and 75% treatment), with a small proportion not having done this before their adult education experience (Table 2).
Student background information.
All 39 students responded.
The majority of students estimated they had done six or more PowerPoint presentations before this class (Table 2: 89% control and 75% treatment). Most students had been recorded in an informal setting (e.g. home movies) and also in circumstances such as theater or musical events. Many less reported experiencing a situation similar to this class, in which they were recorded in a formal setting. There was little difference between the previous experiences of the two groups.
Effect of intervention on second presentations
The learning management software allowed the instructor to determine when and for how long students visited the website. Although it was not possible to determine whether the control students actually watched their first presentations, as they were instructed to do, time spent logged in gave an indication. All of the intervention students accessed the website for an appropriate length of time soon after they had given their presentations (1–5 days), indicating that they all likely used the recordings to complete their self-assessment. In contrast, 47% of the control students were logged in for enough time to have watched their presentations.
The second presentations were graded based on the suggested areas for improvements. Figure 1(a) shows that there was a modest but statistically significant difference between the two groups. The control group received an average improvement grade of 21.3, while the intervention group’s average score was 23.6 (p = 0.03). The control group scores varied much more than the treatment group (standard deviation (SD) control = 4.17 and treatment = 1.47); about half the students in the control group did not address many of the comments appropriately although the rest did (Figure 1(b)), while in the intervention group all the students made good attempts to address the comments.
Effect of intervention on second PowerPoint grades: (a) Second scores for PowerPoint presentations and two written assignments in the same class. Student t-test indicates no significant difference for assignments 1 and 2 (p=0.46 and 0.4, respectively), and asterisk indicates statistical difference for PowerPoint grades (p=0.03). (b) Second scores for PowerPoint presentation shown as number of students in each grade range for each group.
In addition to comparing the overall scores for the second presentations, the data were also broken down by specific criteria. There were four areas in which the intervention group scores were statistically significantly higher than the control (p < 0.05). These were explaining graphs/figures (No. 4 in Table 1), presentation timing (No. 6), correct grammar/spelling (No. 11), and clarity of speaking style (No. 14). There were several other categories worth mentioning in which the intervention group scored higher than the control, although the differences were not statistically significant (p > 0.05). These categories were introduction and body (Nos. 2 and 3), well-practiced (No. 12), and intonation and speed of speech (No. 15). It is also interesting to note that for both groups eye contact was a common area that needed improvement, but the intervention did not help in this category. Thus, overall, the differences between the two groups were in both areas of content and more general presentation/speaking skills.
The trend in the literature is for content to be improved rather than speech skills when multiple presentations were performed. However, in this study, both broad categories were improved; this may be because of the more challenging content level in this class (which for some still had room for improvement) and the maturity of the students—many already had good delivery skills.
In addition to the PowerPoint presentations, there were two written assignments in the class for which the students were graded twice. As for the PowerPoint presentations, the second grade was based solely on attempting to address the comments, rather than overall quality. Figure 1(a) shows that for these other assignments there was no difference in the mean scores between the two groups. Thus, the difference between the two groups’ behavior could not be explained by the students in the intervention group simply being more conscientious. The data were examined to determine whether the same group of students who were scoring lower on their second PowerPoint was more generally weak in terms of addressing comments from evaluations. Table 3 shows correlative data within each group. There was no correlation between the scores for the PowerPoint presentations and Assignment 1; this assignment was unrelated to the presentation topic. In contrast, there was a modest correlation between the scores for the PowerPoint presentations and Assignment 2. This assignment was the review paper on which the presentations were based. The correlation was higher for the intervention group (intervention r = 0.54 and control r = 0.3). This result suggests that the intervention pushed that group’s behavior toward more consistency in improving both assignments focused on the literature review.
Correlations between second PowerPoint scores and two written assignments.
Pearson’s correlation coefficient (r) was determined for each pair of data. Assignment 1: writing assignment unrelated to presentation; Assignment 2: literature review on which the presentations were based.
Another factor considered was overall achievement of the student in the class. Figure 2 shows the relationship between second PowerPoint grade and final class grade. In the control group, there is a low but significant correlation between the second PowerPoint grade and overall grade (r = 0.36). Figure 2 shows that in this group several students did well in the class overall, but did not score well on their second PowerPoint. For the intervention group, there is a stronger correlation between the PowerPoint score and final class grade (r = 0.59). This suggests that the students in the intervention group performed according to their overall abilities. There will always be some variation in student achievement, but the intervention raised some to a more appropriate level.
Relationship between PowerPoint grade and final class grade. Lines indicate best fit for both data sets. Correlation coefficient for control group: r=0.36; intervention group: r=0.59.
Comparisons of assessment type
Table 4 shows that there were moderate correlations between instructor and peer evaluations for both groups. Both groups combined had an r value of 0.55. This value is very similar to that found by Falchikov and Goldfinch (2000) in their meta-analysis of 48 peer assessment studies. They found that when multidimensional (vs. global) scoring was used, the average correlation between peer and tutor grades was 0.53.
Pairwise comparisons of scores after the first PowerPoint presentations.
Table 4 shows that in contrast to peer–instructor assessment, there was no correlation between self and peer or self and instructor scoring, indicating that overall the intervention group was not at all consistent in their self-assessment ability. When self and peer assessments were compared to instructor scores in a range of student activities including oral presentations, self-instructor assessments were generally more weakly correlated than peer–instructor (Falchikov and Boud, 1989; Falchikov and Goldfinch, 2000; Langan et al., 2005, 2008).
It was observed that some students tended to extreme peer-scoring, giving everyone much higher or lower grades than average, but there were no correlations between extreme scoring and either presentation grades or overall class achievement. Thus, there were no caveats found to the use of peer assessment.
This self-assessment data were analyzed further to determine any trends in the unreliability. Figure 3 shows a significant negative correlation (r = −0.49) between how much the self-score deviated from the instructor score and the final class grade. Most of the students who gave themselves higher scores for the presentations were lower ranking students. While several of the students who gave themselves significantly lower scores than the instructor were students who earned high final grades. But there were also several of these high percentile students whose self-assessment scores were close to or higher than instructor.
Negative correlation between self-assessment (deviation from instructor grade) and class grade for students in the intervention group. Pearson’s coefficient r=−0.49.
Over the whole intervention group, there was no correlation between deviation from the instructor score and score in the second presentation. However, when the students were divided into over-estimators and under-estimators, there was a significant negative correlation in the over-estimator group between their deviation from instructor score and their second presentation score (r = −0.64). In the under-estimator group there was a small positive correlation between deviation from the instructor score and second presentation grade (r = 0.38). This indicates that the weakest over-estimating students were those who improved their presentations the least.
This observation that lower achieving students tend to over-rate themselves when giving presentations has been noted before, whether they were (De Grez et al., 2012; Mort and Hansen, 2010) or were not been recorded (Langan et al., 2008).
Student reflections
During the final class, the students were asked to provide their thoughts on the process of being recorded and the evaluation types. When asked whether knowing they were going to be recorded made them more nervous, the average response was in between the extremes of “It didn’t make any difference” and “It made me very nervous” (see Figure 4). The SDs of the data from both groups were large, and there were several students in each group who chose the extremes of the scale. There was no statistical difference between the two groups. There was a small but significant difference between the groups when asked whether they practiced more knowing they were going to be recorded (control = 2.28 ± 1.07, intervention = 3.05 ± 0.97, and p = 0.01).
Student responses to the questions below. A 5-point Likert scale was used for each, 1 being “It didn’t make any difference” or “not at all helpful” and 5 being “It made me very nervous” or very helpful.”
The mean score was high when students were asked whether they thought it useful to repeat their presentation, and there was no significant difference between the groups (Figure 4) (p = 0.3). Similarly, both groups indicated that they found evaluating the presentations of others useful for improving their own skills.
When comparing usefulness ratings for feedback types, both groups ranked the instructor evaluation the highest and self the lowest (Figure 5). There is a significant difference between the control group rating of usefulness between self and peer, this group finding the peer evaluations more useful than the self (mean 3.11 vs 3.78, p = 0.02). For the intervention group, however, there is not a significant difference between self and peer usefulness (p = 0.28).
Student ranking of usefulness of different evaluations. Error bars=SD. Student t-tests indicate there are no significant differences between two groups for each category (p>0.05). Asterisk indicates significant differences between the control group’s self and peer mean scores (p=0.02).
Conclusion
The students who were required to formally self-assess their first presentation improved in their second presentations more than the control group. The results from the intervention suggests that students who had to take the time to complete the self-assessment on paper then continued on and took the time to improve their presentations using their feedback and other feedback provided. This demonstrates that the combination of all three types of evaluation results in most consistent improvement of presentations. The areas in which differences were observed are in all of the general categories assessed (content, organization, and delivery), specifically the content explanations were better, timing was more accurate, and the students in the intervention group showed more improved in their speaking style.
It is interesting to note that while the requirement to self-assessment improved student performance in the second presentations, there was no correlation between self scores and either instructor or peer scores. The students understood the assignment requirements because the peer–instructor correlations were much better. Hence, the lack of congruence between instructor and self must stem from a lack of ability to objectively evaluate their own performance, at least for some students. The results suggest that the act of completing the self-assessment and taking the time to watch and reflect on the performance are valuable, but assigning scores similar to others is less so if the purpose of the self-assessment is formative.
Another somewhat surprising result from this research is that there was no difference between the groups when they were asked to rate usefulness of self-assessment. The control group ranked the peer feedback higher than the self, which is understandable since they were not required to analyze their recordings in a structured manner. But the intervention group was no more appreciative of self-assessment than the control students, despite the success of the intervention. Both groups ranked instructor evaluation highest. Despite the helpfulness of peer and self-assessment, perhaps it is not surprising that students value “expert” opinions most. In a study in which postgraduate students only received assessment from the self or instructor, self-assessment was sufficient for the students to perform as well as those who received instructor feedback (Hecimovich et al., 2010). But for students at an earlier stage in their higher education, this was not the case (De Grez et al., 2009a), and those who received instructor feedback improved the most in their presentations.
Instructor, peer, and self-assessment have been compared in other studies focused on presentations and other performance activities in higher education. But this study highlights the value of combining three types of formative assessment for students giving oral presentations. In the control group, formal instructor and peer evaluations and informally being told to watch their performance were sufficient for over half of the students to improve their presentations. However, this means that for the other students this was not sufficient. These are the students who benefited most from the additional requirement to formally self-assess. The less internally motivated, lower achieving students are the most likely beneficiaries of the increased structure.
It is perhaps an unusual way to assess students, that is, to ask students to improve the same presentations. However, this removes the necessity to address brand new content areas, which in this level of class is a great deal of extra work. This method is unique in the literature, but where class size and time constraints allow, this is a very valuable exercise. It allows a detailed analysis of improvement without added variables related to a change in information focus. Moreover, the students rank highly the usefulness of being able to rework their presentations. Since there was a wide range of topics covered over 2 months, from the audience perspective, there was no feeling of redundancy or ennui due to repetition of information.
The conclusions drawn from this research are based on a relatively small class size, although the effect was statistically significant. The students in the study were all junior or senior science majors, and the class is a majors’ requirement rather than a general education requirement. Both these factors mean that the majority of students were highly motivated. Hence, the conclusions drawn regarding the combined effect of the assessments could be limited to these academically successful and relatively mature undergraduates. On the other hand, considering the body of evidence that shows the use of peer assessment or self-assessment is generally positive in a range of student categories, the conclusion could be reasonably extrapolated to a broader range of students.
A future area of inquiry will focus on involvement of students in establishing assessment criteria; this is recommended as a mechanism by which students have more power within the class (Tan, 2012), and interestingly, the criteria provided by the students usually closely match those that come from the instructor (Falchikov and Boud, 1989). Aside from peer reviewing, the students in this study received little practice in assessing before they commenced their presentations. This type of activity has been shown to increase the correlations between instructor and student scores. In future research, the students will be given class time to practice and discuss presentations that have now been recorded and archived for future use. It will be interesting to determine whether performance is impacted because while this type of activity has been shown to improve student–instructor score correlations, this does not necessarily equate to improved performance (Baecher et al., 2013).
This study highlights the value in requiring students to use three types of assessment (instructor, peer, and self) to improve their presentation skills. The use of structured self-assessment resulted in improvement above that achieved with instructor and peer only. From the instructor’s perspective, the additional work required when the students self-assessed was negligible since the rubric used was the same. From the student perspective, the additional time they needed to take to watch themselves was clearly worthwhile in improving their presentation skills.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.