“Your words matter!” Relevance of individual participation in xMOOCs

Abstract

Massive open online courses are increasingly popular. One characteristic of most massive open online courses is that they are offering educational contents to masses of different individual learners. At the same time, a particular challenge for the individual learner could be the massiveness of such courses, that is, that one is part of a large crowd of other learners. Being one of many other individuals can have negative consequences on learning activities because an individual might experience a sense of virtual crowding. The experiment described in this article investigated how course type (xMOOC vs virtual seminar) and relevance of individual participation (low vs high) impact online learners’ engagement. Results showed that participants in an xMOOC condition referred to themselves and their peers to a greater extent when their participation was introduced as being highly relevant. Implications for the design of xMOOCs are discussed.

Keywords

massive open online courses references to oneself references to peers relevance virtual crowding

Individual learning in massive open online courses

Recent technological developments have led to new ways of delivering instruction online, which thereby have the potential to change existing views of how information is delivered. For example, the immensely popular online courses called massive open online courses (MOOCs) allow educational content to be presented to very large numbers of learners. Yet, these one-to-many learning settings present new challenges for instructional communication (e.g. Kruiderink, 2013). Overall, two main models of MOOCs can be distinguished: connectivist MOOCs (cMOOCs) and extended MOOCs (xMOOCs). cMOOCs are based on the idea of connectivism (see Siemens, 2005), that is, learners work together in a network structure. A more popular and far more widespread model of MOOCs are xMOOCs (Brahimi and Sarirete, 2015). xMOOCs follow a behaviorist approach, that is, the learner primarily interacts with materials individually (Conole, 2014). Put differently, xMOOCs are driven by a self-guided design (McAuley et al., 2010).

In general, online courses have several advantages over traditional lectures and seminars, for example, being independent from time and space and allowing to regulate one’s own learning pace (e.g. Young and Norgard, 2006). Alongside the advantages, studies have also indicated possible disadvantages of online courses, for example, that students become bored of course materials over longer periods (e.g. Szpunar et al., 2013). In addition, online courses tend to experience rather high attrition rates (Moody, 2004; Shaw et al., 2016), which especially applies to very large online courses like xMOOCs: Since xMOOCs have become popular, their greatest criticism has been their high dropout rates, arising from students not completing a course (e.g. Jordan, 2014; Liyanagunawardena et al., 2013).

xMOOCs have been viewed under different pedagogical foci, whereby a majority of empirical studies focused on rather internal factors of individual learning (for a systematic review, see Zhu et al., 2018). At the same time, a better understanding of the social environment in which individual learning takes place is lacking importance. As the impact of massification can be crucial for the quality of individual learning (see, for example, Abualrub et al., 2013), a factor that is neglected so far in the research literature on xMOOCs is that individual participants often know that they are not alone on large online courses. Participants might even, as it is typical for xMOOCs, be directly presented with the exact number of participants as an indicator of a company’s success (see, for example, Coursera, FutureLearn). Therefore, the individual learner is aware of a very large number of others when participating in an xMOOC, which might be a challenge for individual engagement in course activities.

While xMOOCs give masses of individual learners the opportunity to expand their knowledge on a variety of topics (McAuley et al., 2010), many courses still show a fundamental lack of instructional quality (Margaryan et al., 2015). That is, the individual learner is not supported sufficiently in them (Spector, 2014). As such, xMOOCs deliver content for a potentially infinite number of learners who are responsible for organizing their own learning. Therefore, a better understanding of psychological factors is required that particularly support the individual learner in xMOOCs. In a cMOOC, for example, one factor that promotes self-directed learning is feeling activated (Kop and Fournier, 2011).

However, being part of an xMOOC—and, accordingly, potentially having the feeling of being only one among many others in a virtual crowd—might in itself impair self-directed learning. As such, an xMOOC participant might be indifferent toward engaging in course activities when it is not pointed out whether one’s participation is relevant at all. In order to address this problem, so-called SPOCs (small private online courses) are emerging in university teaching (e.g. Fox, 2013; Kaplan and Haenlein, 2016). This format is equivalent of what is here denoted as virtual seminars; these are small seminars that contain online components, which are managed via Learning Management Systems (LMS). Compared to xMOOCs, participants might perceive a higher personal relevance in a virtual seminar by default and should therefore be more likely to engage themselves in course activities.

Relevance of participation and individual engagement

Students may ask themselves: “Is it worth it for me to fully engage in this MOOC?” Clark and Mayer (2016) distinguish between behavioral and psychological engagement in online learning: Behavioral engagement refers to conduct actions, for example, typing on a keyboard to produce a text, whereas psychological engagement refers to cognitive processes, for example, the mental effort involved in producing a text. In addition to social-psychological conceptualizations of individual relevance, further engagement in a particular subject can be achieved when an individual experiences an issue as being relevant (Petty and Cacioppo, 1986; for further proposals to support learners’ engagement, see, for example, Zepke and Leach, 2010). In addition, Park and Choi (2009) showed that this is also applicable to online courses: The experienced relevance of an online course was particularly predictive for dropout rates of adult learners, and, as already stated, dropout rates are a large problem in xMOOCs, too. To deal with this problem and to support learners’ engagement in online courses, participants should be personally addressed (e.g. McCombs and Vakili, 2005). This is also especially important because Phan et al. (2016) found that when learners are more motivated to engage in an xMOOC, they are more likely to achieve learning success.

Therefore, it is necessary to explore whether psychological engagement could also be supported in xMOOCs. When being part of a rather impersonal xMOOC (vs a virtual seminar), the perceived relevance of an individual learner’s participation could be an important factor for individual engagement in course activities.

Dealing with tasks in a virtual crowd

Besides the influence of crowding on the extent of individual engagement in an xMOOC, a further important factor is its impact on how learners deal with tasks. For traditional lectures and seminars, the number of participating students can strongly influence many important cognitive variables in learners. Studies revealed negative consequences of large class sizes on different outcomes for students in higher education, for example, their reported learning success (e.g. Monks and Schmidt, 2010), their grades (e.g. Johnson, 2010; Kokkelenberg et al., 2008), or their evaluation of courses and lecturers (e.g. Bedard and Kuhn, 2008; Johnson et al., 2013; Westerlund, 2008). Students perceive “crowding” issues, such as insufficient space for individual learning, as highly problematic (Campbell and Shlechter, 1979). This is indicated, for example, by students’ tendencies to choose seats in a library that are most distant from others (Eastman and Harper, 1971). As being in a crowd often means that people do not have enough personal space (Hall, 1966), this can lead to the experience of social crowding (Stokols, 1972), which can also have negative consequences for the cognitive elaboration of presented material, particularly while dealing with complex tasks (e.g. Evans, 1979).

Varying information about the virtual presence of others (crowded vs non-crowded) can affect how often learners refer to themselves and others when answering an open question (Hellmann et al., 2016). Therefore, it can be assumed that an individual may not only be affected by crowds in a real social environment but also when having many others in mind, such as by feeling virtually crowded in large online courses like xMOOCs.

When students are aware of the size of an online course, they arguably have the virtual presence of other learners in mind. Furthermore, participating learners may perceive the relevance of their individual participation differently when experiencing virtual crowdedness, which might influence references to oneself and others in response to an open question (see Hellmann et al., 2016). The perception of individual relevance can play an important role for learners’ engagement (see Petty and Cacioppo, 1986), which is also supported by findings of Park and Choi (2009), who showed that experiencing relevance in online courses is an important factor to maintain participation.

In sum, the number of other learners in online courses, particularly in xMOOCs, might be challenging for the engagement of individual learners. For this reason, the individual learner might benefit from receiving information regarding the relevance of his or her participation. Put differently, there is a need to explore whether pointing out relevance of individual participation supports learners’ engagement in xMOOCs. Therefore, this study investigates the extent to which the number of virtual others (xMOOC vs virtual seminar) and the perceived relevance of participation (low vs high) can affect an individual’s references to oneself and other students in an online course when answering an open question regarding one’s opinion on learning with online videos. The amount of those references is conceptualized as an indicator for learners’ engagement.

In line with this, it is predicted that in an xMOOC condition (vs a virtual seminar), an individual should refer to oneself less often when the individual’s participation is seen as less relevant (low) than when it is seen as more relevant (high), as participants in a large virtual crowd should be less engaged in written contributions when relevance of individual participation is not pointed out. In particular, the amount of references to oneself, measured via first-person singular pronouns, should be lowest when participants are in an xMOOC condition (vs virtual seminar) and the relevance of the individual’s participation is low (vs high)—Hypothesis 1.

Furthermore, references to other students and references to university lecturers are differentiated. Participants who feel activated in a cMOOC are more likely to show a desire to interact with other students in the course (Kop and Fournier, 2011). Therefore, references to other students (but not to university lecturers) should also differ on the dimension of perceived relevance in an xMOOC condition (vs virtual seminar). In particular, an individual participant in an xMOOC should refer to other students less often when the relevance of the individual’s participation is low (vs high)—Hypothesis 2.

In addition to this, interest, motivation, and cognitive preconditions were taken into account. Whether participants’ level of need for cognition is high or low might affect their learning behavior. Furthermore, participants’ impressions of the instructional information can indicate whether any unintended negative effects of the wording are distributed unequally across conditions. Participants’ previous experiences with online learning offers and their previous knowledge of the McGurk effect are additional factors to rule out that they have an effect to the assumptions of this study.

Method

Participants and design

Participants were students from a cross-section of disciplines at the University of Münster (Germany). Data collection took part—from March to April 2015—by using the online platform Questback Enterprise Feedback Suite (EFS) Survey . The data from 31 participants were excluded from the final data set, because these participants were not enrolled at a university at the time of their participation (13), responded incorrectly to the manipulation check (see below) (12), indicated that they had already taken part in this or a similar study (4), reported technical problems during participation (1), or explicitly indicated that their data should not be used for data analysis (1). Furthermore, a suspicion check was carried out to determine whether the used manipulations were too salient. In particular, participants were asked for their views on the real purpose of this study. An analysis of replies revealed that none of the participants fully guessed the experimental purpose.

The final data set included N = 169 students (n = 109 females) with a mean age of 22.79 years (SD = 2.96) and a mean study time of 4.67 semesters (SD = 3.66). They took part in a 2 (course type: xMOOC vs virtual seminar) × 2 (relevance of individual participation: low vs high) between-participants design. The experiment was approved by the Ethics Committee of the University of Münster. Participants were recruited via email lists and social media with an implemented hyperlink leading to the study named “learning with videos.” Participants were compensated with a 5€ voucher from a large online retailer.

Procedure and materials

All participants provided informed consent on the first page in the online experiment. After that, they were randomly assigned to one of the four experimental conditions and were given further instructions.

Experimental manipulation

Participants, in their respective condition, received information about different online course scenarios and the relevance of individual participation in this study. Two experimental manipulations were implemented in the instructions: Course type was manipulated via information on the respective course (xMOOC vs virtual seminar) and via ostensible quotes as accounts of two learners’ experiences within this course type. Participants read statements of two ostensibly real students, Laura and Frank. In the xMOOC condition, for example, Laura stated that she participated in a MOOC with about 12,000 people; in the virtual seminar conditions, she reported having participated in a virtual seminar with 18 people. This manipulation was necessary to enable the participants of this study to build up a mental image of a specific online course type, in particular, identifying with the individual learning situation of being one among many or few other online learners. Relevance of individual participation was also manipulated in the instruction. Participants were informed about the subsequent use of their answers to the open question in one of two ways: In the low-relevance condition, the text stated that “Individual answers in this experiment will be anonymized and analyzed for scientific purposes only!” In the high-relevance condition, the text mentioned that “Individual answers in this experiment have a high degree of relevance for us, will be analyzed carefully, and will be taken into account for planning future teaching events!” (see Figures 1 and 2 for the exact wording in each condition).

Figure 1.

Materials. Course type: xMOOC versus virtual seminar.

Figure 2.

Materials. Relevance of individual participation: low versus high.

Next, all participants were informed that they would watch a short educational video from their respective online course, depending on the experimental condition. On the following page, participants watched a learning video with a duration of 2 minutes and 49 seconds, which started automatically. The short learning video contained an explanation and illustration of the McGurk effect (the interplay of auditory and visual senses within speech perception; McGurk and MacDonald, 1976) and was produced by the authors of this study. The McGurk effect is a rather unknown phenomenon and was simply used as one possible material here for a short learning video in an online course. Because the experiment was conducted at a German university, the featured language in the video was German. The same video was presented across all conditions.

Thereafter, participants answered a short knowledge test regarding the effect they had just learned about. This is a typical learning task in xMOOCs (e.g. Toven-Lindsey et al., 2015) and was here used as a filler task. Participants then replied to an open question about their individual attitude toward online learning videos (see below). In the last part of the experiment, demographic data and a set of control variables were assessed. Finally, participants were told how they could claim their voucher and were fully debriefed and thanked. An email address to contact the experimenter was provided in case participants had any further questions.

Dependent measures

The main goal of this study was to analyze participants’ written contributions to an open question, namely, “To what extent do you think online learning videos allow knowledge to be transmitted in a useful manner?” The answers were analyzed along different dimensions, which are outlined in the following sections.

References to oneself

Using the text analysis software Linguistic Inquiry and Word Count (LIWC; Pennebaker et al., 2007; German dictionary by Wolf et al., 2008), participants’ answers to the open question were analyzed. Among other features, LIWC provides detailed information about the frequency of certain word categories (e.g. pronouns) in relation to the total length of a text. References to oneself were analyzed with the LIWC category “I,” which contains first-person singular pronouns (e.g. I, me, myself).

References to others

Participants’ answers to the open question were analyzed by means of verbal analysis (Chi, 1997). It was differentiated between references to other students, that is, peers on the one hand and references to university lecturers on the other hand. Two independent coders, who were blind to the respective condition, analyzed each written contribution. To use an equivalent measure to LIWC, the absolute frequency of respective references to other students and to university lecturers was coded. Afterward, the relative frequency of references for each contribution was calculated. Therefore, the absolute frequency of references to other students and to university lecturers was divided by the respective word count of each text. The correspondence of initial ratings between the coders was satisfactory, with Cohen’s κ = 0.759 for references to other students and Cohen’s κ = 0.766 for references to university lecturers (see Cohen, 1960).

For example, the following statement was coded as containing a reference to other students: “A good online course conveys knowledge in a simple way and makes it possible […] to get into contact with one’s online fellow students.” References to university lecturers were wordings such as “The interaction with lecturers, for example in a seminar, cannot be replaced by online courses, because they don’t allow for immediate questions.”

Manipulation check

Two questions with three response options assessed whether participants had carefully read the instructions and identified the information given regarding the two experimental factors: “How many other people did Laura take the online course with?”: (1) 18 people, (2) with about 12,000 people, (3) I don’t know, and “What will happen with your anonymous data?”: (1) they will only be used for this study, (2) they will be analyzed carefully and will be taken into account in the planning of further learning events, (3) I don’t know.

Interest, motivation, and cognitive preconditions

A short knowledge test was introduced after the video (five multiple-choice questions) to measure what participants had learned about the McGurk effect. Each question contained four response options (see Figure 3).

Figure 3.

Short knowledge test. All participants received the same five multiple-choice questions with four choices regarding the short online video (translated from German).

Students’ answers to each question were coded as 0 (incorrect) or 1 (correct). Therefore, participants could get a total score between 0 and 5 points. Furthermore, participants provided ratings on their response behavior (one item: “How important was it to you to fulfill the tasks in this study properly?” ranging from 1 = not at all important to 7 = very important), motivation (one item: “How motivated were you working on the task in this study?” ranging from 1 = not at all motivated to 7 = very motivated), perceived impact of their own contribution (one item: “What is your assessment of impact of your contribution while doing this study?” ranging from 1 = not at all relevant to 7 = highly relevant), and mood (one item: “How is your current mood?” ranging from 1 = very bad to 7 = very good).

A short scale measuring participants’ need for cognition was also included (Beißert et al., 2014). The scale consists of four items (ranging from 1 = does not apply at all to 7 = applies completely), for example, “Simply knowing the answer rather than understanding the reasons for the answer to a problem is fine with me.” Thereafter, participants were asked whether the direct quotes of Laura and Frank created a pleasant or positive impression for them on two scales, ranging from 1 = completely disagree to 5 = completely agree.

Furthermore, participants were asked to indicate whether or not they had previously used online learning offers and, if so, to rate their experience regarding valence on a scale, ranging from 1 = very negative to 5 = very positive. Furthermore, participants were asked whether they had any previous knowledge of the McGurk effect.

Results

Answers to the open question

All answers had a mean length of 87.02 words (SD = 57.45), with no differences between conditions, Fs < 1.485, ps > 0.225.

References to oneself

A 2 (course type: xMOOC vs virtual seminar) × 2 (relevance: low vs high) analysis of variance (ANOVA) revealed that there were no main effects of course type and relevance of individual participation relating to the linguistic dimension of the category “I,” Fs < 2.040, ps > 0.155. However, there was a significant interaction effect, F(1, 165) = 4.433, p = 0.037, = 0.026 (see Table 1 for relative means and SDs). Post hoc analyses using Least Significant Difference (LSD) comparisons indicated that the use of first-person pronouns differed between conditions of relevance of participation (low vs high) when participants were in the xMOOC condition, while there was no such difference in the virtual seminar condition. As part of the xMOOC condition, participants used fewer words from the first-person pronoun category when their individual participation was seen as less relevant (low) compared to when it was viewed as more relevant (high), F(1, 165) = 6.309, p = 0.013, = 0.037. Under the condition of a virtual seminar, use of first-person pronouns did not differ between the levels of low versus high relevance, F(1, 165) = 0.227, p = 0.634 (see Figure 4 for an illustration of significant group differences).

Table 1.

Relative Means (and standard deviations) of the lexical analysis of answers in each condition regarding references to oneself and others.

	xMOOC		Virtual seminar
	Low relevance	High relevance	Low relevance	High relevance
References “I” (LIWC)	1.69 (1.65)	2.90 (2.12)	2.40 (2.56)	2.17 (2.43)
References “Others” (verbal analysis)
References to students	0.11 (0.36)	0.51 (0.84)	0.37 (0.71)	0.34 (0.70)
References to lecturers	0.31 (0.65)	0.31 (0.58)	0.33 (0.62)	0.34 (0.79)

xMOOC: extended massive open online course; LIWC: Linguistic Inquiry and Word Count.

Figure 4.

References to oneself and to other students depending on course type and relevance of individual participation.

References to other students

A 2 (course type: xMOOC vs virtual seminar) × 2 (relevance: low vs high) ANOVA revealed that there were no main effects relating to references of other students, Fs < 3.253, ps > 0.073. However, there was a significant interaction effect, F(1, 165) = 4.110, p = 0.044, η_p² = 0.024 (see Table 1). Post hoc analyses using LSD comparisons indicated that the use of references to other students differed between the conditions of relevance of participation (low vs high) when participants were in the xMOOC condition, while there was no such difference in the virtual seminar condition. As part of the xMOOC condition, participants used fewer references to other students when their individual participation was seen as less relevant compared to when it was viewed as more relevant, F(1, 165) = 7.414, p = 0.007, η_p² = 0.043. Under the condition of a virtual seminar, use of references to other students did not differ between the levels of low versus high relevance, F(1, 165) = 0.025, p = 0.875 (see Figure 4 for an illustration of significant group differences).

References to university lecturers

There were no main effects and interaction effects above all conditions regarding references to university lecturers, Fs < 0.069, ps > 0.794 (see Table 1).

Interest, motivation, and cognitive preconditions

A 2 (course type: xMOOC vs virtual seminar) × 2 (relevance of individual participation: low vs high) between-participants ANOVA revealed that there were no main and interaction effects on the total score of correct responses in the short knowledge test, Fs < 2.194, ps > 0.140. With a possible total score of 5 points, participants generally scored very high across all conditions (M = 4.19, SD = 0.66).

A 2 × 2 multivariate analysis of variance (MANOVA) revealed that there were no main and interaction effects of conditions on the ratings of individual response behavior, motivation, perceived impact of individual contribution, mood, and need for cognition, Fs < 0.873, ps > 0.516. For the need for cognition scale, Cronbach’s alpha was 0.213 (this low reliability score can be explained with the loosely connected concepts of the few items of this short scale; see Beißert et al., 2014). Furthermore, a 2 × 2 MANOVA revealed that there were no effects of conditions on the ratings of pleasantness or positive impression regarding the direct quotes of Laura and Frank, Fs < 0.709, ps > 0.494 (see Table 2 for means and SDs of additional analyses).

Table 2.

Means (and standard deviations) of additional analyses in each condition.

	xMOOC		Virtual seminar
	Low relevance	High relevance	Low relevance	High relevance
Short knowledge test	4.19 (0.80)	4.40 (0.66)	4.15 (0.89)	4.02 (1.03)
Individual response behavior	5.93 (0.96)	5.90 (0.98)	5.83 (1.00)	6.18 (0.95)
Individual motivation	5.91 (0.78)	5.90 (0.82)	5.83 (0.85)	6.08 (0.97)
Perceived impact of individual contribution	4.91 (0.97)	4.83 (0.88)	4.74 (1.37)	5.34 (0.91)
Individual mood	5.60 (0.93)	5.60 (0.88)	5.43 (1.02)	5.66 (0.85)
Need for cognition	4.98 (1.04)	4.91 (0.91)	5.15 (0.78)	5.08 (0.99)
Rating of direct quotes (instruction)
Pleasantness	3.33 (0.81)	3.26 (0.80)	3.13 (0.65)	3.32 (0.90)
Positivity	3.33 (0.71)	3.14 (0.78)	3.09 (0.55)	3.13 (0.81)

xMOOC: extended massive open online course.

Moreover, there were no descriptive differences between conditions regarding the following frequencies: About half of the participants (50.3%) previously used online learning offers and judged their experiences rather positively on a scale, ranging from 1 = very negative to 5 = very positive, (M = 3.71, SD = 0.87). Most of the participants (90.5%) did not have any previous knowledge of the McGurk effect.

Discussion and conclusion

This study investigated how the number of other participants (xMOOC vs virtual seminar) and relevance (low vs high) of an individual’s participation can affect learners’ engagement. The results showed that learners in the xMOOC condition refer less to themselves and other students in their contributions when individual participation was not pointed out as relevant. Thus, the findings suggest that telling students that their individual participation is relevant in an xMOOC might be an effective way to promote their engagement in online course activities. This is in line with findings of Petty and Cacioppo (1986). The higher amount of references to oneself and other students, when students knew their individual participation in the xMOOC condition was relevant, can be interpreted as an indicator for learner’s engagement. Put differently, the learner is willing to bring their own opinion to the front, when being aware of the relevance of participation. Thus, especially in an xMOOC, an individual learner may benefit from being personally addressed in a virtual crowd. Moreover, pointing out the relevance of participation in an xMOOC might also help to reduce the enormous dropout rates among participants (see, for example, Jordan, 2014). And this, in turn, might also help to maintain individual participation.

Furthermore, when pointing out relevance of individual participation in the xMOOC condition, results showed differences in references to other students (peers). That is, students refer more to their peers when they perceived their own participation as being relevant. Therefore, the findings are also an extension to the findings of Kop and Fournier (2011), who could show that participants, who felt activated in a cMOOC, also were more likely to interact with other students. In turn, cooperative work in general can again support students’ engagement (see, for example, Herrmann, 2013). Drawing on this, it is also important to enable learners to interact with others in xMOOCs, because only some of them offer discussion forums (e.g. Kaplan and Haenlein, 2016). In this vein, participants might also especially benefit from being salient to others in terms of social presence (see, for example, Kehrwald, 2008). However, there were no differences for references to university lecturers over all conditions. This might be explained by the fact that lecturers are not directly available in xMOOCs (e.g. Kay et al., 2013). To counteract this problem, the use of peer assessment is popular in xMOOCs (e.g. Bali, 2014).

Moreover, references to oneself and other students did not differ for the virtual seminar condition when participants were told how relevant their participation would be. This might be explained by the fact that participants in a virtual seminar perceived themselves as more visible and therefore had a higher perceived relevance of participation than when they were part of a virtual crowd in an xMOOC. Like SPOCs, virtual seminars offer the advantage of helping participants to not feel like they are lost in a one-to-many learning setting. In order to make participants in an xMOOC feel that they are personally relevant, it is necessary to identify further factors that promote this feeling in order to improve their individual engagement in course activities.

Overall, there was no difference between conditions regarding the total score of correct responses in the short knowledge test, possibly due to a ceiling effect, that is, the test might have been too easy. The test served as a filler task and only contained five questions with four response options, which could be insufficient to reveal effects between conditions regarding recall of previously learned information. Furthermore, there were no differences between conditions regarding the ratings of individual response behavior, motivation, perceived impact of individual contribution, mood, need for cognition, and ratings of pleasantness or positive impression to the given instructional information.

The limitations are as follows. All assumptions were tested in an experimental setting by using a scenario design. That is, participants were not part of a real online course environment, which affects ecological validity. Furthermore, especially the participation in a real xMOOC would differ in various points from the mere confrontation with vignettes employed in this study. Participants normally choose their own subject which they would like to study. In the work described here, learning materials were prescribed. In addition, the amount of individual experiences with xMOOCs might be important for one’s learning behavior. Both, in turn, might affect learners’ motivation and performance (see, for example, De Barba et al., 2016). With regard to motivation, there was no difference between groups in this study. Moreover, a real xMOOC includes more as well as other materials and a more complex course structure. That is, a single learning topic is embedded into a larger context. This study focused on one single phenomenon (here, McGurk effect), which could not be reflected in a larger context. Furthermore, repetition of content was not provided here, whereas real xMOOCs might, for example, offer previews and summaries. In addition, and as another limitation of this study, not all xMOOCs contain writing tasks, which allow for bringing one’s own opinion to the front (like it was measured via references to oneself here). Furthermore, the used manipulation of relevance was rather artificial. That is, relevance of participation was not pointed out for participation in an xMOOC or virtual seminar but, in the study itself, which was about learning with videos. Due to the extent of real xMOOCs, the duration of participation is probably longer (despite the issue of high dropout rates; see Jordan, 2014). Thus, longitudinal effects could not be examined here, although they might play an important role for learners when being part of a large virtual crowd. Furthermore, and besides the aspect that xMOOCs are, namely, “open” to everyone, such courses are often used by people with an university education (Littlejohn and Hood, 2018). Participants of the sample in this study were students from different disciplines at one university in Germany. In particular, it is another limitation that only a small cross-sample of students from one country was used here. Therefore, results do not allow for giving advice to a specific academic discipline or to students from any country other than Germany. In line with that, participants of xMOOCs can differ in educational and cultural backgrounds, learning experiences and abilities, study habits, aims, and intentions.

Taken together, a real xMOOC experience is determined by various factors that have to be controlled or considered when studying “real learning.” As a starting point, a scenario was used in this study, which is much easier to control and manipulate than a real xMOOC environment. Future research should place more emphasis on the relation of individual learning and social learning environment in real xMOOCs, that is, testing whether negative impacts of virtual crowds could also be found for individual learning in such courses—particularly with different types of learners all around the globe and various topics and tasks. Based on the findings of this study, a better understanding of learners’ psychological needs could also be beneficial for the instructional design of real xMOOCs such that the individual’s learning activity could be improved.

In sum, findings particularly showed that supporting individual engagement in an xMOOC scenario can strengthen the communication of one’s own opinion, which, in turn, might prevent feelings of being lost in a virtual crowd. In addition, findings also indicate that individual learning in xMOOCs is related to social exchange with other learners. Therefore, participants should also be provided with sufficient opportunities to communicate with their peers to further enrich their individual learning experience in xMOOCs.

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.

Author biographies

Jens Riehemann (PhD in Psychology, University of Münster) is a Postdoctoral Researcher and Lecturer at the University of Münster, Germany. His research addresses digitization in higher education with a special focus on supporting individual learners in massive open online courses. Address: Department of Psychology and Sport Studies, University of Münster, Fliednerstraße 21, 48149 Münster, Germany. [email: jens.riehemann@uni-muenster.de]

Jens H Hellmann (PhD in Psychology, Jacobs University Bremen) is a Postdoctoral Researcher and Lecturer at the University of Münster, Germany. He has published in journals of various areas of psychology, including social and educational psychology. His main focus is on coordinating a project on psychological aspects of refugee integration. Address: Department of Psychology and Sport Studies, University of Münster, Fliednerstraße 21, 48149 Münster, Germany. [email: jens.hellmann@uni-muenster.de]

Regina Jucks is a Full Professor at the University of Münster, Germany. She received her PhD in Psychology in 2001 and her Habilitation in 2005. Since 2011, she has headed the local Center of Teaching in Higher Education. Her research addresses various fields of learning through communication with a special focus on digitized interactions. Address: Department of Psychology and Sport Studies, University of Münster, Fliednerstraße 21, 48149 Münster, Germany. [email: jucks@uni-muenster.de]

References

Abualrub

Karseth

Stensaker

(2013) The various understandings of learning environment in higher education and its quality implications. Quality in Higher Education 19(1): 90–110.

Bali

(2014) MOOC pedagogy: Gleaning good practice from existing MOOCs. Journal of Online Learning and Teaching 10(1): 44–56.

Bedard

Kuhn

(2008) Where class size really matters: Class size and student ratings of instructor effectiveness. Economics of Education Review 27(3): 253–265.

Beißert

Köhler

Rempel

, et al. (2014) Eine deutschsprachige Kurzskala zur Messung des Konstrukts Need for Cognition. Die Need for Cognition Kurzskala (NFC-K) [A German Short Scale for the Measurement of the Construct of Need for Cognition. The Need for Cognition Short Scale (NFC-K)]. Köln: GESIS.

Brahimi

Sarirete

(2015) Learning outside the classroom through MOOCs. Computers in Human Behavior 51(Part B): 604–609.

Campbell

Shlechter

(1979) Library design influences on user behavior and satisfaction. The Library Quarterly 49(1): 26–41.

Chi

(1997) Quantifying qualitative analyses of verbal data: A practical guide. The Journal of the Learning Sciences 6(3): 271–315.

Clark

Mayer

(2016) E-Learning and the Science of Instruction: Proven Guidelines for Consumers and Designers of Multimedia Learning. San Francisco, CA: John Wiley & Sons.

Cohen

(1960) A coefficient of agreement for nominal scales. Educational and Psychological Measurement 20(1): 37–46.

10.

Conole

(2014) A new classification schema for MOOCs. The International Journal for Innovation and Quality in Learning 2(3): 65–77.

11.

De Barba

Kennedy

Ainley

(2016) The role of students’ motivation and participation in predicting performance in a MOOC. Journal of Computer Assisted Learning 32(3): 218–231.

12.

Eastman

Harper

(1971) A study of proxemic behavior: Toward a predictive model. Environment and Behavior 3(4): 418–437.

13.

Evans

(1979) Behavioral and physiological consequences of crowding in humans. Journal of Applied Social Psychology 9(1): 27–46.

14.

Fox

(2013) From MOOCs to SPOCs. Communications of the ACM 56(12): 38–40.

15.

Hall

(1966) The Hidden Dimension. Garden City, NY: Doubleday Publishing.

16.

Hellmann

Adelt

Jucks

(2016) No space for others? On the increase of students’ self-focus when prodded to think about many others. Journal of Language and Social Psychology 35(6): 698–707.

17.

Herrmann

(2013) The impact of cooperative learning on student engagement: Results from an intervention. Active Learning in Higher Education 14(3): 175–187.

18.

Johnson

(2010) Class size and student performance at a public research university: A cross-classified model. Research in Higher Education 51(8): 701–723.

19.

Johnson

Narayanan

Sawaya

(2013) Effects of course and instructor characteristics on student evaluation of teaching across a college of engineering. Journal of Engineering Education 102(2): 289–318.

20.

Jordan

(2014) Initial trends in enrolment and completion of massive open online courses. The International Review of Research in Open and Distributed Learning 15(1): 133–160.

21.

Kaplan

Haenlein

(2016) Higher education and the digital revolution: About MOOCs, SPOCs, social media, and the cookie monster. Business Horizons 59(4): 441–450.

22.

Kay

Reimann

Diebold

, et al. (2013) MOOCs: So many learners, so much potential … IEEE Intelligent Systems 28(3): 70–77.

23.

Kehrwald

(2008) Understanding social presence in text-based online learning environments. Distance Education 29(1): 89–106.

24.

Kokkelenberg

Dillon

Christy

(2008) The effects of class size on student grades at a public university. Economics of Education Review 27(2): 221–233.

25.

Kop

Fournier

(2011) New dimensions to self-directed learning in an open networked learning environment. International Journal of Self-Directed Learning 7(2): 1–18.

26.

Kruiderink

(2013) Open buffet of higher education. Trend Report: Open Educational Resources. 54–58. Available at: https://www.surf.nl/binaries/content/assets/surf/en/knowledgebase/2013/Trend+Report+OER+2013_EN_DEF+07032013+%28HR%29.pdf

27.

Littlejohn

Hood

(2018) The many guises of MOOCs. In: Littlejohn

Hood

(eds.) Reconceptualising Learning in the Digital Age. Singapore: Springer, pp. 1–19.

28.

Liyanagunawardena

Adams

Williams

(2013) MOOCs: A systematic study of the published literature 2008–2012. The International Review of Research in Open and Distributed Learning 14(3): 202–227.

29.

McAuley

Stewart

Siemens

, et al. (2010) The MOOC Model for Digital Practice. Charlottetown, PE, Canada: University of Prince Edward Island.

30.

McCombs

Vakili

(2005) A learner-centered framework for e-learning. Teachers College Record 107(8): 1582–1600.

31.

McGurk

MacDonald

(1976) Hearing lips and seeing voices. Nature 264(5588): 746–748.

32.

Margaryan

Bianco

Littlejohn

(2015) Instructional quality of massive open online courses (MOOCs). Computers and Education 80(C): 77–83.

33.

Monks

Schmidt

(2010) The Impact of Class Size and Number of Students on Outcomes in Higher Education. Available at: http://digitalcommons.ilr.cornell.edu/workingpapers/114

34.

Moody

(2004) Distance education: Why are the attrition rates so high? Quarterly Review of Distance Education 5(3): 205–210.

35.

Park

Choi

(2009) Factors influencing adult learners’ decision to drop out or persist in online learning. Journal of Educational Technology and Society 12(4): 207–217.

36.

Pennebaker

Francis

Booth

(2007) Linguistic Inquiry and Word Count: LIWC (text analysis software program). Austin, TX: liwc.net.

37.

Petty

Cacioppo

(1986) The elaboration likelihood model of persuasion. In: Berkowitz

(ed.) Advances in Experimental Social Psychology, vol. 19. New York: Academic Press, pp. 123–205.

38.

Phan

McNeil

Robin

(2016) Students’ patterns of engagement and course performance in a massive open online course. Computers and Education 95(April): 36–44.

39.

Shaw

Burrus

Ferguson

(2016) Factors that influence student attrition in online courses. Online Journal of Distance Learning Administration 19(3): 1–8.

40.

Siemens

(2005) Connectivism: A learning theory for the digital age. International Journal of Instructional Technology and Distance Learning 2(1). Available at: http://202.116.45.236/mediawiki/resources/2/2005_siemens_Connectivism_A_LearningTheoryForTheDigitalAge.pdf

41.

Spector

(2014) Emerging educational technologies: Tensions and synergy. Journal of King Saud University-Computer and Information Sciences 26(1): 5–10.

42.

Stokols

(1972) On the distinction between density and crowding: Some implications for future research. Psychological Review 79(3): 275–277.

43.

Szpunar

Khan

Schacter

(2013) Interpolated memory tests reduce mind wandering and improve learning of online lectures. Proceedings of the National Academy of Sciences 110(16): 6313–6317.

44.

Toven-Lindsey

Rhoads

Lozano

(2015) Virtually unlimited classrooms: Pedagogical practices in massive open online courses. The Internet and Higher Education 24(January): 1–12.

45.

Westerlund

(2008) Class size and student evaluations in Sweden. Education Economics 16(1): 19–28.

46.

Wolf

Horn

Mehl

, et al. (2008) Computergestützte quantitative Textanalyse [Computer-assisted quantitative text analysis]. Diagnostica 54(2): 85–98.

47.

Young

Norgard

(2006) Assessing the quality of online courses from the students’ perspective. The Internet and Higher Education 9(2): 107–115.

48.

Zepke

Leach

(2010) Improving student engagement: Ten proposals for action. Active Learning in Higher Education 11(3): 167–177.

49.

Zhu

Sari

Lee

(2018) A systematic review of research methods and topics of the empirical MOOC literature (2014–2016). The Internet and Higher Education 37(April): 31–39.