Flipped microlecture classes: satisfied learners and higher performance?

ABSTRACT

Multimedia in blended learning provides the ability to enhance students’ performance and satisfaction. Microlectures are one type of multimedia that comprise of short modular videos that introduce basic theoretical concepts. Literature provides limited insight into how this specific format supports engineering classrooms. This study, therefore, explored what the benefit of flipped microlecture classes is for students in a course with mainly traditional face-to-face lectures. To this end, we allocated ten students to a flipped microlecture and twelve to a traditional lecture setting; conducted pre and post-tests; and, organised a focus group. Results show a positive effect of the flipped microlecture class on the test performance. Students’ qualitative evaluations also indicate their satisfaction with increased control over learning, reduced discomfort of traditional lecture settings, enhanced classroom interaction, and better knowledge retention. These findings resonate with blended learning and multimedia literature and support future decision making about microlectures in engineering education.

KEYWORDS:

• Microlecture
• BIM/engineering
• flipped classroom
• performance
• satisfaction

1. Introduction

Modern higher engineering education aims to enhance the quality of interaction that takes place between students, their peers, and instructors. This current approach replaces the teaching-centred paradigm that considers traditional teacher-based instruction as an endpoint in education (Barr and Tagg 1995) with active learning methods (Fear et al. 2003) that strive to give students ownership of their learning trajectory, and ultimately improve their outcomes.

One specific format in which instructors let students prepare for interactive lecture sessions is the flipped classroom. Such a setting expects that students study instructional material before they physically attend classroom sessions. During in-class sessions, students use the available time not to listen to a plenary face-to-face presentation but to do exercises and discuss concepts in detail. The literature review conducted by Lo and Hew (2019) suggests that flipped classroom approaches in engineering education provide benefits to learning achievements over traditional learning – such as, for example, self-paced learning and availability of classroom time for problem-solving activities.

Along with the introduction of flipped classrooms came the increase in possibilities for producing, sharing, and distributing videos via the internet (van der Meij 2017). Multimedia became learning resources for a flipped classroom setting in which instructors transfer knowledge outside classes, and where interactive learning takes place during the face-to-face time with students (Hamdan et al. 2013).

Existing literature assesses the effectiveness of existing video types, such as, for example, voice-over recordings of presentation slides (Fogarty 2017; McClelland 2013), lecture captures (Chen and Wu 2015), and demonstration videos (Grossman et al. 2013; van der Meij 2017). These types differ in terms of format, accessibility, length, purpose, and content. The length of a video influences how it contributes to a student’s control over his learning experience. Furthermore, the different purposes of videos (e.g. demonstration, task instruction, and persuasion) support distinctive types of knowledge and skills development.

Despite these differences, multimedia studies seem often not to explicitly define the type, length, and the number of videos that they used (Kay 2012). Since such differences can result in different learning outcomes, literature would benefit from clearly defined details about new types of videos for blended learning. This would, in turn, refine the understanding of the value that distinctive multimedia types in flipped mode classrooms add to traditional face-to-face teaching.

One recent video type that gained popularity in higher education is the microlecture. Microlectures are short instructional videos or audio fragments that explain a tightly defined topic (Sweet 2014). They have a duration of five to ten minutes and are segmented in modules. Each microlectures is furthermore segmented into a sequence of self-containing units, following segmentation principles in multimedia learning (Margulieux, Guzdial, and Catrambone 2012; Schittek Janda et al. 2005). Aimed at knowledge and comprehension (Bloom 1956), their purpose is to transfer basic conceptual knowledge or provide instructions about a concept. Although literature explored how this specific and contemporary video type can be implemented in education (Sweet 2014), it remains unknown how its implementation in a flipped classroom impacts students’ performance and satisfaction.

One typical course in engineering education that benefits from flipped teaching with microlectures is Building Information Modelling (BIM). The course covers concepts from the disciplines of Civil Engineering, Construction Management, and Information Technology. Short videos about those concepts could prepare students for in-class discussion and hands-on exploration with BIM software. Engineering education literature in this field previously assessed the impact of online learning platforms on distant software learning (Hewitt et al. 2016; Neill, DeFranco, and Sangwan 2017), the sentiments towards online learning (Suwal and Singh 2018), and the value of visual computer aids on comprehension (Kösa and Karakuş 2018). It did not, however, explore the value that the flipped microlecture classes could bring to students’ performance and satisfaction.

To bridge this gap, we explored the impact of short and modular microlecture videos within a flipped BIM-classroom setting. Specifically, we evaluate how the lecture design contributed to students’ performance and their satisfaction by comparing it with traditional face-to-face lectures.

The remainder of this paper is outlined as follows. First, it discusses the theoretical background of multimedia in BIM engineering classrooms. It then elaborates on how we compared students’ performance and satisfaction in flipped microlectures vis-à-vis traditional face-to-face lectures. Next, we present results showing positive effects of the experimental microlecture group on performance and satisfaction. This paper concludes by reflecting on these findings for BIM classrooms and future research.

2. Theoretical background: multimedia in flipped engineering classrooms

BIM is a topic within Civil Engineering that addresses the development of virtual models to support construction project management. It uses digital models, visualisations, and simulations to streamline collaboration between the many stakeholders that are involved in a construction project. BIM tools support advanced analysis of buildings, bridges, and road infrastructure. The topic gained momentum in the past decade and most large firms in the construction industry nowadays use it in their practices.

Scholars stress the relevance of placing BIM in college and university curricula (Jagiełło-Kowalczyk and Jamroży 2016; Nushi and Basha-Jakupi 2017; Sampaio 2015). BIM-students benefit from a high-quality interaction during their in-class time with lecturers, as they need to obtain conceptual knowledge about virtual design, and master technical modelling skills. The literature discusses, therefore, how these skills should be taught in BIM curricula worldwide (Lopez-Zaldivar et al. 2017; Zhang, Xie, and Li 2017), and what challenges exist for its further integration in engineering teaching (Abdirad and Dossick 2016).

The development of suitable BIM curricula is supported by trends in multimedia and learning approaches. Increased internet bandwidth and the rise of online social media like YouTube, for example, created a new landscape of media content suitable for educational purposes (Kay 2012). This, in turn, shaped new learning spaces comprising virtual workspaces and digital libraries (Zhang et al. 2006). Furthermore, the affordability of mobile devices and omnipresent internet access increased students’ demand for more flexible educational approaches (Ravishankar, Epps, and Ambikairajah 2018).

New approaches are visible in classroom designs that blend digital learning concepts with elements of traditional teaching. Thai, De Wever, and Valcke (2017) categorise these along a continuum scale of four classroom designs. The first type is traditional classroom teaching which combines face-to-face lecturing with guidance questions and immediate feedback during one session. During this time, the lecturer mostly sends information through presentations and the blackboard, while students can ask a limited number of questions. The second type is e-learning where online digital tools form a platform for instruction, interactions, and feedback activities. The third type is blended learning. This combines elements of the previous two: it uses digital tools to execute either instruction or interaction activities outside the classroom. The fourth type is the flipped classroom. This is one version of blended learning that uses virtual lectures and in-class interactions and feedback (Baytiyeh and Naja 2017). It replaces passive traditional face-to-face in-class information transfer with flexible and active construction of knowledge (King 1993).

The flipped classroom design has been introduced in chemistry (c.f. Olakanmi 2017), medicine (c.f. Lewis, Chen, and Relan 2018; Sharma et al. 2015; Tune, Sturek, and Basile 2013), and architecture and engineering (c.f. Baytiyeh and Naja 2017; Lo and Hew 2019; Lucke, Dunn, and Christie 2017; Newman et al. 2016; Ravishankar, Epps, and Ambikairajah 2018). In this setting, various video types could be implemented to support teaching. Literature that distinguishes four of those video types: first, podcasts contain presentation slides with voiceover audio explanation (Chen and Wu 2015; Fogarty 2017), recordings of lectures (Chen and Wu 2015; Newman et al. 2016; Tune, Sturek, and Basile 2013), or other supplementary illustration to demonstrate, summarise and introduce a problem (Freguia 2017; Kay 2012). Second, videos may refer also to screen recordings of the instructor’s computer (Chen and Wu 2015; Mason, Shuman, and Cook 2013) on which a task, simulation, or software is demonstrated. Third, instruction videos can demonstrate specific procedural tasks and behaviour (Rosen et al. 2010), such as management skills (Grossman et al. 2013) or software instruction (van der Meij 2017; van der Meij and van der Meij 2016a). Fourth, the term video may more broadly refer to the less standardised public ‘YouTube’ videos (Sharma et al. 2015), and TED Talks (Yelamarthi and Drake 2014) that often have a persuasive character and are not tailored to educational purposes.

Altogether, videos can influence learning in various ways. Sung and Hwang (2013) state that such impacts can appear in both the cognitive and affective domains. Cognitive changes take place in the form of enhanced performance after instruction sessions. Performance is the temporary fluctuation in retained knowledge that can be measured immediately after the acquisition process (Soderstrom and Bjork 2015). Review studies of multimedia (c.f.Kay 2012), and flipped classrooms (c.f. Lo and Hew 2019) generally show positive impacts on performance. Experimental results also show that webcasts (Freguia 2017; Traphagan, Kucsera, and Kishi 2010), and video material (Zhang et al. 2006) improve students’ performance compared to traditional lectures.

Next, changes in the affective domain relate to students’ altered motivation and satisfaction with learning resources (Sung and Hwang 2013). Satisfaction is the direct experience resulting in expressions of interests or liking (Blunsdon et al. 2003). Literature associates videos and the flipped classroom with satisfaction. Empirical studies of students that used webcasts, for example, reported that learners increased their satisfaction with the course (Ravishankar, Epps, and Ambikairajah 2018; Traphagan, Kucsera, and Kishi 2010), and had enjoyable experiences (Copley 2007; Dupagne, Millette, and Grinfeder 2009; Winterbottom 2007).

Other studies reported satisfying experiences together with qualitative evaluative statements about the instructional format. These were, for example: being motivated by the ability to sustain attention (Kay 2012), having no distraction (Winterbottom 2007), benefitting from taking notes at own time and pace (Copley 2007; Dupagne, Millette, and Grinfeder 2009; Lucke, Dunn, and Christie 2017; Ravishankar, Epps, and Ambikairajah 2018; Zhang et al. 2006), enjoying the ability that the resource provided to learn independently and self-directedly (Green et al. 2003; Piccoli, Ahmad, and Ives 2001), having the flexibility to pause and rewind as many times one likes (Winterbottom 2007; Zhang et al. 2006), having the ability to cover material multiple times, and being forced to learn a topic to be prepared for class (Ravishankar, Epps, and Ambikairajah 2018).

Although media in flipped classroom settings show effects on performance and satisfaction, these results may be affected by the unique characteristics of multimedia video in terms of its length, format, and interactivity (Kay 2012). Microlectures are one type of video that has not been extensively addressed in flipped classroom studies. Their short duration (five to ten minutes) make them an attractive resource in addition to in-class activities, to prepare for lab sessions, create curiosity about the future discussion topic, or recap information (Sweet 2014). Brevity makes sure that only essential information is transferred and the attrition of viewers’ attention or dropout while watching video instructions, are prevented (Guo, Kim, and Rubin 2014).

In comparison with podcasts, screen recordings, instruction videos, and public videos, microlectures have a unique set of features. Specifically, (1) they are much shorter; (2) cover only basic concepts rather than detailed knowledge; and (3) focus on informing and explaining concepts rather than pursuing audience, or demonstrating tasks. These differences could give students distinctive learning experiences, and hence hinder any prior reasoning about the impact of flipped microlectures classes on students’ performance and satisfaction.

The goal of this study is therefore to evaluate whether the microlecture format supports students’ performance and satisfaction enhancement in the flipped BIM-classrooms, relative to an existing state-of-practice in which the traditional lecture is dominant.

3. Research method

We implemented a flipped micro lecture classroom setting for one lecture topic in our existing BIM course. Before this study, this BIM-course primarily used traditional face-to-face teaching. These traditional lectures served as a point of departure for a systematic comparison with the new flipped microlecture format.

3.1. Experimental design

We use an experimental study design to compare the flipped microlecture classroom with traditional lecturing. This research design is similar to that of Baytiyeh and Naja (2017) that used a traditional face-to-face classroom design as a control group and compared this with learning experiences in a flipped classroom with voice-over slideshows. Another example is the work of van der Meij and van der Meij (2016a), who studied demonstration video learning based on two experimental groups. We selected this approach because it divides the student population in an experiment and control group, and helps infer causality between the intervention and the dependent variables (Bishop-Clark and Dietz-Uhler 2012).

The control and experimental setting were similar except for the instructional style (i.e. traditional presentation or microlectures with an in-class session). In the experimental condition, students watched five microlectures as a preparation for classroom exercises and discussions. The control condition consisted of a traditional face-to-face lecture in which students listened, did exercises, and discussed a statement. The study material, exercises, and discussion topics in both conditions were identical.

3.2. Course intervention

The intervention took place during a twelve-week-long (5ECTS) BIM course at the graduate level in 2018/2019. The lecture entitled ‘standardization and interoperability’ took place in week two of this term. This lecture was selected for the flipped microlecture intervention because its basic concepts could be explained without interaction between the instructor and student, while it requires exercise and interactive discussion to deepen understanding of how standards are applied. The first author taught the standardisation and interoperability lecture twice: once in a traditional face-to-face format (control condition), and subsequently by using the flipped classroom microlecture format (experiment condition). The lectures took place sequentially on the same day.

The reason for selecting only one lecture was pragmatic. This study was part of a teaching innovation project of the first two authors. There was limited time available to innovate and study the way they delivered their BIM-course. Within the period, they spent over twenty hours recording microlectures, following microlecture training, revising slides, and preparing scripts. At the start of the experiment, most students had little to no experience with flipped classrooms or microlectures. This novelty may have influenced their attitude.

3.3. Participants

Students participated voluntarily. After the participants received instructions and signed informed consent, we divided the group of 27 students into two. We randomly allocated thirteen participants to the experimental, and fourteen to the control group. Seven participants from the experiment group also took part in the focus group discussion that took place one week after the in-class session.

3.4. Instructional material

Experimental group: we designed microlectures that complied with principles of multimedia design. The first author followed a two-day training on micro lecturing. Based on this, he was able to use Mayer and Fiorella’s (2014) principles coherence, signalling, redundancy, spatial contiguity, and temporal contiguity to design his microlecture slides and autocue script. The principles helped to avoid the cognitive overload that redundancies and misalignment in messages and media may cause. Furthermore, principles like segmenting and modality were applied to avoid cognitive overload caused by dense and fast presentations (Mayer and Pilegard 2014). Finally, the use of conversational language and the human voice and embodiment principles (Mayer 2014) helped design social cues in the microlectures blueprints. The second (content expert) and third author (multimedia expert) reviewed the lectures. These efforts resulted in five microlectures with a duration of respectively 3:55, 6:36, 3:24, 2:42, and 6:45 min (see Appendix 4).

We provided the experiment group access to the videos via our Canvas learning management system. One week before the session, we instructed the students to watch the microlectures before coming to the in-class session. During the session, students asked questions about the microlecture content, made exercises, and participated in discussions. The in-class session had a duration of around 75 min, excluding the fifteen-minute break.

Control group: the first author reused the identical presentation slides from the microlectures in his face-to-face lecture. He also integrated the exercises and discussion topics of the in-class session of the experimental group into the face-to-face-presentation for the control group. The face-to-face lecture was shorter than the total learning time that the experiment group had. It had a duration of 90 min, excluding one ten-minute break.

3.5. Measurement instruments

Figure 1 schematically shows the research setup and data collection moments. To measure performance data, we used performance scores and self-reported data (Kay 2012). As for performance scores, literature uses course marks (Ravishankar, Epps, and Ambikairajah 2018), group report marks (Freguia 2017), exam scores (Boevé et al. 2017; Traphagan, Kucsera, and Kishi 2010), tests (Chen and Wu 2015; Tune, Sturek, and Basile 2013), and pre-posttests (Olakanmi 2017; Thai, De Wever, and Valcke 2017; van der Meij and van der Meij 2016b; Zhang et al. 2006). We used a simple pretest to test the homogeneity of the control and experiment group, and a more difficult post-tests to assess significant performance differences between the groups after they took part in the experiment.

Figure 1. Schematic illustration of the experimental setup and data collection moments.

In specific, the first author developed two digital diagnostic tests as pre and post-test (see appendices 1 and 2). He implemented these in the online testing system Remindo (Paragin E-learning 2019). The second author validated the test questions. All students took the pretest on their laptops directly after the first lecture of the term and finished this in around fifteen minutes. The post-test had a similar duration. The control group took this test after the face-to-face lecture, while the experimental group took the test at the end of the classroom session (see Figure 1). The performance was calculated as the relative number of correct answers in an assessment.

We also assessed performance based on students’ perceptions (Boevé et al. 2017; Kay 2012; Newman et al. 2016). Since we dealt with a new phenomenon and smaller class size (n < 30), we used a focus group to interact with students and deliberate with them about the classroom setting (Bishop-Clark and Dietz-Uhler 2012, 52–3).

We modified and included questions from existing studies as a guide for our focus group protocol (see Appendix 3). Besides addressing performance, the protocol adopted questions about, for example, satisfaction levels (Zhang et al. 2006), learning pace, interaction, (Ravishankar, Epps, and Ambikairajah 2018), engagement (Lucke, Dunn, and Christie 2017), intrinsic motivation, and flexibility (Thai, De Wever, and Valcke 2017). The session had a duration of sixty minutes. An experienced moderator guided the discussions, and the researcher was present to provide clarification to panel members. We collected data during the session through audio recording and transcribed this data afterward.

3.6. Analysis

After a first random allocation of 27 students to the experimental and control group, two were switched between groups because of their agenda limitations. We excluded the students from the study when they did not fill out the complete pretest or post-test. This resulted in, respectively, 12 (pretest) and 10 students (post-test) for the traditional lecture group, and 13 (pretest) and 12 (post-test) for the flipped microlecture group. The analysis was done with the understanding that while this small sample size may not result in conclusive evidence from a strictly statistical standpoint, it can very well serve as a solid indication of potentials of flipped microlecture classes.

We used the Shapiro Wilk W test to assess whether the performance test data had a normal distribution (Shapiro, Wilk, and Chen 1968). The W-test had a p-level lower than 0,05 for the pretest data of the microlecture group (Table 1, W(12) = 0,811, *p = 0.013), so we concluded that the data was not normally distributed. Consequently, we used the non-parametric Mann Whitney-U means test (Fay and Proschan 2010) to assess whether the pre and post-test performance data sets had an equal distribution and whether there was a significant difference between the post-test data (at significance levels of 0,05).

Table 1. Shapiro-Wilk normality test.

Condition	Pretest		Post-test
	Statistic	Significance	Statistic	Significance
Traditional (control, n = 12, 10)^a	0,915	0,351	0,954	0,733
Micro (experiment, n = 13, 12)^a	0,811	0,013*	0,929	0,371

^aNumber of participants in the pretest and post-test.

For the qualitative analysis of the focus group, we used the software ATLAS.ti to perform open and axial coding (Strauss and Corbin 1994). During open coding, we quoted and coded participants’ claims from the focus group transcripts on a sentence-by-sentence basis and marked excerpts that related to participants’ lecture experiences. This resulted in the first list of forty codes (e.g. benefits of microlecture, increased sense of control, modularity, length of lecture). During axial coding, these were mapped with the performance and satisfaction benefits from literature and culminated in the results presented below.

4. Results

We discuss the outcomes of the statistical test and focus group below.

4.1. Pre and post-test performance scores

Table 2 shows the performance test scores of both groups before and after they attended the lectures. Results show that the mean performance scores increased. The Mann Whitney-U Means Test demonstrates that the difference between the pretest scores of the experiment and the control group are not significant (asymp. Sig (2-tailed)) given that the p = 0,121 (U = 50,0; α = 0,05). This means that we consider the groups in the baseline situation as randomly distributed.

Table 2. Descriptive statistics of the pre-post test scores.

Condition	Pretest		Post-test
	Mean	Var	Mean	Var
Control, (n = 12, 10)^a	4,1108	1,600	5,3770	2,792
Experiment, (n = 13, 12)^a	4,8200	1,568	6,8250	0,814

^aNumber of participants in the pretest and post-test.

A repeated Mann Whitney-U test also demonstrates that the difference between the post-intervention scores of the experiment and control group is significant (asymp. Sig (2-tailed)) given that the p = 0,021 (U = 25,5; α = 0,05). In other words, the experiment group (microlecture) scored significantly higher than the control group on their post-test.

4.2. Perceived performance and satisfaction

This section uses selected verbatim quotes from the transcript of the focus group session to illustrate how the participants perceived the impact of the flipped microlecture classroom on their performance and satisfaction. After addressing the benefits, we end with a note from the focus group about the complementarity between the flipped classroom setup and microlectures.

4.2.1. Control over the learning process

The focus group explained how the ability to play and pause the microlectures at any time, provided them control over their learning process. One participant argued about this as follows:

in a microlecture, you will see before you start how long it will take. You then start watching and can make notes in a separate document […] What I mean is, yes, thus, before I start watching the ten minutes, I know that I will push myself to carefully pay attention for ten minutes. I will make notes and then I know that, after these ten minutes, I learned about a part of the learning material or at least the basis of it. So, […] the way I divide my attention is much shorter and focused compared to a two-hour [face-to-face] lecture. [participant 6]

This suggests that the ability to play short videos and take notes in parallel increases the students’ ability and motivation to stay focussed on the learning material that he watches.

Participant two added to this a comparison with the traditional lecture, by stating: ‘and you can pause. You can think and go back. And that is difficult in [a traditional] lecture’. Together, the two quotes above suggest that the fixed and short duration of the microlectures fit the students’ attention span better than a longer traditional lecture. Moreover, the ability to pause-and-play allowed students to flexibly plan when they watched the videos, allowing them to watch these at times when they could concentrate well.

Three other quotes explain how the length of the microlecture made it more attractive to repeat videos. One participant stated:

Look, what [participant 7] just stated. If you watch such a microlecture and it is just ten or five minutes and if it goes a little too fast for you to understand [at once], you may then think, [hey] how can that be, I should be able to understand it. You will then replay and seek back. I notice that I will then try to understand and playback for twenty seconds. [participant 6]

Similarly, another participant argued:

Yes, it is also about the length of the video. It is for ten minutes and not two hours. I found it pleasant that the videos were just around five minutes and not fifteen or so. Because if it is just five or ten minutes, it is very easy to watch. It is done quickly, and you will watch it back more easily. [participant 1]

Yet another participant added:

‘suppose the video would have taken ten minutes, then I would not have done that’ [participant 2].

These three statements suggest that students were satisfied with the ability to pause sections and that the short duration of microlectures lowered the threshold for students to repeat and pause sections multiple times.

Respondent 7 finally noted that he probably would not have repeated videos of longer recorded traditional lectures:

In a recorded traditional lecture, which has quite a high pace, and when you do not understand it. I will not watch this back [to mark] what I did not understand and where I need to google or read more about.

Again, this implies that the length and modularity of the microlectures stimulate the students to repeat and replay video material more often.

4.2.2. Classroom discomfort

Further, participants were satisfied that the microlectures reduced the discomfort that they experienced in traditional classroom settings. Participant 6 explained as follows how he perceived discomfort during a traditional lecture:

So then I notice, that I want to understand it and rewind the video for twenty seconds, and the threshold for doing this is quite low. I can imagine that, when you are in a [traditional] lecture, that you do not want to raise your hand all the time [to ask for the lecturer to repeat what he said]. And, you will definitely not do this when your classmates are silent after you have already asked a couple of questions before. That forms a social psychological barrier.

This quote shows that the format of microlectures could make it easier to engage with learning material, without feeling constrained by social pressures from (less active) peers.

One participant also commented on his satisfaction with the absence of other classroom discomforts related to the moods of the lecturer and fellow students:

The way I divide my attention is shorter and focussed than when I need to listen to a two-hour lecture of an instructor, who may have an off-day, and other students too. I feel I have more control over this. [participant 6]

This implies that the microlectures suffer less from the local and temporary situation in the classrooms. These situations that may reduce the quality of the instruction when noizes from outside or from peers disturb students from keeping attention during long lectures.

4.2.3. Classroom interaction

Also, the specific combination of microlecture videos and the subsequent in-class session created benefits for the students. The combination requires students to watch the microlectures as preparation for the in-class session that directly engaged with learning material – without any presentation of the instructor. Participant 2 reasoned about this as follows, ‘When you have a traditional lecture, you can sit down and stay silent. In this context, you are expected to at least contribute’. This quote shows that the participant considered the availability of the videos as an incentive for him to prepare for the in-class session.

Participant 4 added to this:

[I still need to go to the in-class session still] but then I have prepared and already know what I do and do not understand. […] with the video lectures you prepare for what you can expect in a discussion and you see what you already understand and what not.

This explains that the videos helped the students to become aware of what they already understood, and what not. In turn, this provided them more specific questions and knowledge for the in-class session. The focus group moderator summarised this as follows: ‘so, you say you learn more, because, if you do not understand something, you will be confronted with that directly [in-class]’.

Further, the flipped microlecture classroom stimulated interactivity during the in-class session. Participant 6 explained that he perceived more lively interaction as the main benefit:

I think that the true benefit of such a microlecture [setting] as a whole is that you can use the second part [i.e. the in-class session] to go through, but not repeat completely, the themes from the microlectures. You can make it livelier, have discussions, and help to retain that knowledge. People will then be taken in-depth and can connect concepts with anecdotes, and applications.

This expression explains that students appreciated that they already mastered knowledge before the in-class session, which allowed them to more flexibly use and apply it in discussions and exercises.

4.2.4. Knowledge retention

The interactive in-class session further seems to have allowed participants to retain gained knowledge better. Participant two illustrated this by stating, in general:

I think this [approach] will stick better than in a traditional lecture. If I think about [another previous traditional] lecture, and try to say something about it right now, then I probably know more about this [microlecture] than [that traditional face-to-face] lecture.

This effect could be attributed to the fact that video motivated students to prepare for the classroom sessions. In turn, this seems to have stimulated the active engagement with the learning material in the class.

Further, participant 5 more specifically added that the repeated use of video and reflection on the material during in-class sessions could further help to retain knowledge: ‘I think this way of repeating is just as you do while studying for an exam: you repeat once or twice. In that case, you know it much better than when you just listen at it only once’.

4.2.5. Complementarity between flipped classroom and microlectures

The overall impression of the focus group was that the flipped microlecture classroom was a satisfying experience. They attribute this to the combination of microlectures and the in-class session. One participant stated this:

There is potential in this [instructional approach]. You just need to think carefully about, in the context of BIM, […] that you could pick [a concept] and use an example in class to explain how it works in [modelling software like] Revit. How does it look like when I apply this concept? […]. So, instead of demonstrating this for [only] three minutes in a [traditional] lecture, you could really demonstrate and clarify this [in class]. [Respondent 6]

This quote explains the synergy that can be achieved when the theory about Building Information Modelling – which was presented in the microlectures – is combined in the future with a range of practical exercises that apply this theory. This application of theoretical concepts could be done in BIM-software such as Revit. The students appreciated the idea that classroom time could be devoted to these applications.

Further, the group argued that complementarity leads to satisfying outcomes also because of the stimulating social contact that the in-class setting still provides. Participant three stated about this:

This [video] is not really to be considered instead of a lecture but in addition to it. I would not only replace it. An additional discussion session is needed. [in cases where there is no follow-up session,] I would be missing the contact with other students too.

This again shows that the ability to ask questions during a lecture was perceived as essential to creating increased satisfaction and performance benefits. This was not only because of the knowledge that will be exchanged during the sessions but also because of the presence of peers in the classroom that motivate students to learn.

5. Discussion

The comparison between the two groups shows that the flipped microlecture (experiment) intervention positively influences student performance. In addition, students perceive various performance and satisfaction benefits, when comparing the format with the face-to-face traditional classrooms.

Previous studies show that the video type – lengths, format, and interactivity (Kay 2012) – has a significant impact on the outcomes of multimedia learning (Chen and Wu 2015; Ravishankar, Epps, and Ambikairajah 2018). Microlectures have a specific non-interactive format, restrictive lengths (5–10 min), self-containing modules, and purpose of explaining basic theoretical concepts. This format has, however, not been extensively addressed in the existing literature. One main contribution of this study is, therefore, the identification of the performance and satisfaction benefits that this format creates for students in flipped classrooms. These benefits seem to be in line with existing literature on educational multimedia and blended learning. We discuss these findings in detail below.

First, this study provides evidence that the flipped microlecture classroom gives the students increased control over their learning. This resonates with existing studies of multimedia and flipped classrooms, where students reported that benefits were that they could take notes in their own time, and at their own pace (Copley 2007; Dupagne, Millette, and Grinfeder 2009; Lucke, Dunn, and Christie 2017; Ravishankar, Epps, and Ambikairajah 2018; Zhang et al. 2006), that they could pause and rewind videos as many times they liked (Winterbottom 2007; Zhang et al. 2006), and that they enjoyed the ability provided by the virtual learning environment to learn independently (Green et al. 2003; Piccoli, Ahmad, and Ives 2001). Besides this, we further show that the brevity of the microlectures provides more flexibility and incentives to watch learning material frequently, and at any desired moment.

Second, we found that flipped microlectures are less subject to the discomfort that may arise in traditional classroom settings. The focus group gave the example that peer pressures may reduce students’ willingness to freely ask lecturers to repeat material in a traditional (less interactive) classroom, while the flipped microlecture setting did not have this obstruction since videos could be repeated multiple times. Pre-recorded, concise microlectures seem to control better for uncertainties like disruptions from fellow students, or lecturers with an off-day. These findings confirm existing evidence that suggests that video streams enable students to sustain attention (Kay 2012), and suffer less from distractions (Winterbottom 2007).

Third, we found that the experimental setup increases students’ interaction during in-class sessions. It seems that the ease of preparing the classroom session, and the necessity to be prepared for interaction, stimulates the students to engage with learning material upfront. This increases the knowledge of students before they enter the classroom, but also provides the insight into what they do not yet understand. This feeling of being forced to prepare for interactive classroom sessions is also reported in the existing literature about ICT in flipped classrooms (Baytiyeh and Naja 2017; Låg and Sæle 2019; Ravishankar, Epps, and Ambikairajah 2018).

Fourth, we found evidence that students performed better. This was visible in the increased quantitative results on the post-test scores as well as in the qualitative evaluative expressions from focus group participants. In the latter, students perceived that they retained knowledge better. These results confirm previous findings that short and modular videos would reduce viewer dropouts and increase task performance (Guo, Kim, and Rubin 2014; van der Meij 2017) and that an overall significant effect on performance exists for flipped classrooms over traditional lecturing in engineering education (Lo and Hew 2019).

Like any other study, this research also has limitations. For one, limitations in the data make the results from the conducted qualitative analysis statistically less conclusive. The sample size (n  =  27) were small, and our experiment and control groups were both just below the minimum target size of 15 (see Table 1) to draw statistically significant conclusions. The unequal number of students that were partaking in the pretest and post-test may have influenced the results since students could have decided not to make the post-test in case they had scored low on the pretest.

The test data that shows positive results on the impact of the flipped microlectures on performance (quantitatively) should thus be interpreted with caution and considered as a suggestive result rather than conclusive evidence. For future research, we suggest that data from larger sample sizes are collected to draw a more convincing statistical conclusion about the causal relationship between the flipped microlecture format and performance. This would be possible by repeating this study over various cohorts or in larger class sizes.

Another limitation is that we organised just one focus group with seven, out of fourteen, experiment group members. The participants of the focus group found consensus about the most discussed topics, indicating that our findings are shared amongst students. We did not, however, organise a follow-up session to evaluate whether we the data collection reached an analytic point of saturation. To determine whether the benefits that we identified are complete, clear, and generally applicable, we therefore suggest follow-up studies that include data collection from successive focus groups, and with more engineering graduate students.

Further, this study does not lend itself for demonstrating any superiority of the microlectures over other educational media, nor did it show how this improves the flipped classroom setup. The focus group indicated that students appreciate the complementarity of the modular and brief videos for theory explanation, and in-class sessions for its application. They also added that the social element of meeting students in the classroom is valuable to them. It is this boundedness between the microlectures on one hand, and the flipped classroom on the other, that makes it difficult in the current experimental setup to infer any causality to the satisfaction and performance benefits and microlectures alone. Consecutive research should systematically compare multiple multimedia types in flipped classrooms contexts to assess this.

Another potential limitation concerns the bias that students who are subject to a study, may perform better just because they were observed. This is referred to in the literature as the Hawthorne effect. The students in the flipped microlecture classroom group may have spent additional time and effort in studying the microlectures just because there were research subjects. The fact that microlectures were new to them may also have influenced that they perceived the experiment as more satisfying. We suggest that consecutive research, therefore, repeats this study for multiple successive lectures to reduce this bias.

A point of attention is also that the experiment group had 24 min of microlectures and 75 min of in-class discussion. The experiment group thus exceeded the 90 min of learning time that the control group had. Furthermore, the experiment group had the ability to watch microlectures multiple times. The positive findings of the flipped microlecture context may therefore partly be due to this additional exposure to learning material. This suggests that additional research into learning material engagement may be needed to explore how this influenced outcomes. Such triangulation of data from student engagement with recorded lectures, and qualitative data collection will help to understand whether there is a significant difference in the exposure to learning material between the two groups, and, whether this could influence their performance and satisfaction (Gorissen, Van Bruggen, and Jochems 2013).

The findings of this study finally have practical implications for teachers. The seemingly positive impact of microlectures in flipped classroom settings signals that teachers in (civil) engineering classes could consider this as a fruitful alternative for their classical lecturing approaches. The study could lead to more evidence-based decision making in education design, and in the end, may spur the uptake of innovative learning approaches in engineering curricula.

6. Conclusion

Flipped microlecture classrooms in civil engineering show a positive effect on performance scores and students’ perceived learning performance and satisfaction. We found this by experimenting with twenty-seven M.Sc. students of Building Information Modelling. During the experiment, we allocated the student to two groups. One watched five microlectures and attended an in-class session (experimental group); the other attended a traditional face-to-face lecture (control group).

Outcomes of a pre- and post-intervention test show that participants in the flipped microlecture group performed better than students in the traditional face-to-face lecture group. Qualitative analysis of the focus group shows that the short, modular microlectures helped students to increase control over their learning, minimise discomforts from classroom settings, increase classroom interaction, and retain knowledge better. The complementarity of the microlecture and in-class session seems to make learning more satisfying than traditional face-to-face lectures. All in all, these findings confirm that the flipped microlecture classrooms create benefits similar to those reported in the literature.

These findings contribute to the body of knowledge by demonstrating that one specific video format, i.e. microlectures, influences students’ performance and satisfaction in a flipped classroom. It aligns with previous findings that short and modular multimedia material enables deeper learning in class, and provides more control over learning processes. We argue that these are valuable contributions to the teachers of and other similar engineering courses since our case shows how microlectures can be valuable investments to enhance interaction and performance, and facilitate student-led learning.

Acknowledgments

This study was conducted as partial fulfilment for the first author’s Senior University Teaching Qualification. We greatly acknowledge the students partaking in the study, and in particular the students that voluntarily attended the focus group session. We also thank Monique Duyvestijn for co-organizing and moderating the focus group session; Arthur Veugelers and Martin Bosker for their help with the development of the microlectures; and Kim Schildkamp for supporting the reflection and discussion sessions around this educational research topic.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Léon olde Scholtenhuis

Léon olde Scholtenhuis is an assistant professor in construction management and engineering. He teaches Building Information Modelling, 5D and Planning, Subsurface Utility Engineering, Smart Cities, and Digital Technologies in Construction. Léon is interested in innovations that make teaching more engaging for instructors and students.

Farid Vahdatikhaki

Faridaddin Vahdatikhaki is an assistant professor in construction management and engineering. He teaches Building Information Modelling, 4D and Planning, Simulation and Optimization, and Digital Technologies in Construction. Farid won the University of Twente Teacher of the Year Award in 2017.

Chris Rouwenhorst

Chris Rouwenhorst is an educationalist at the Centre of Expertise in Learning and Teaching (CELT). He is programme coordinator of the 4TU Centre for Engineering education, and faculty educational consultant at the faculty of Science and Technology. Chris supervises teaching qualification programmes and is member of the Technology Enhanced Learning & Teaching team.

Appendices

Appendix 1: pretest lecture 3 inter-operability and standardisation/IFC

Question 1 (mc)

Statements about the interoperability problem in BIM:

1. The interoperability problem is about difficulties between digital and paper integration

2. The interoperability problem influences only to BIM systems, not practitioners

   1. Statement I is true and II is false

   2. Both statements are true

   3. Statement I is false and II is true

   4. Both statements are false

Question 2 (check correct answers)

Why does BIM collaboration need common information representation schemes? Check the valid boxes (more answers are possible).

1. To assure that all buildings and infrastructure designs can be manufactured offsite

2. To assure that a common language and format is used to represent designs

3. To be able to re-use designs in successive stages

4. To convince clients that designs are not only visual but also data-driven

Question 3 (MC)

What is the inter-operability problem in BIM?

1. It relates to the problem of internationalisation in construction and IT

2. It relates to hurdles of data exchange between systems that construction professionals use

3. It relates to the problem of updating and versioning design models

4. It relates to the problem of using different hardware to design BIM models

Question 4 (true false)

The ultimate goal of using standards in BIM is to create standard buildings and infrastructure assets

1. True

2. False

Question 5 (MC)

What does the term ‘proprietary format’ mean in the context of BIM? Proprietary format means:

1. that a BIM model has one owner who is legally responsible

2. that BIM model exchange takes place in the closed format of a software developer

3. that the BIM format that is used to develop a real estate property

4. that a BIM model has multiple owners who are legally responsible

Question 6 (MC)

Which of these examples describe a situation of ‘open BIM’ exchange?

1. An engineer exchanges data by using direct links within a structural design and infrastructure design suites of one single software supplier

2. An engineer exchanges data between his two suites for structural design and infrastructure design by using the standard proprietary format of one software supplier

3. An engineer exchanges data by using lock-in techniques between an open structural design suite and closed infrastructure design suite

4. An engineer exchanges data between his structural design suite and infrastructure design suite by using an open standard

Question 7 (MC)

What is the lock-in problem?

1. Stakeholders who spend money on a BIM system do not know how to use it. There is a lack of training so users are currently locked

2. Stakeholders who use a BIM system of one supplier are restricted to the features and formats of this system: they are locked

3. Stakeholder who use an open BIM system are not flexible and therefore locked in a situation where they do not benefit from other commercial BIM systems.

Question 8 (true false)

When using object models rather than old CAD models (like *.dwg files), we can make multiple visualisations based on the same single data model.

1. True

2. False

Question 9 (MC)

Language, process and data are important parts of establishing a digital communication flow between systems. IFC, the industry foundation class, is relevant to thee three aspects of digital communication flows. What is IFC:

1. An international object-model for describing data exchange processes

2. An international object-model for describing building terms and language

3. An international object-model for describing building data

Question 10 (MC)

What is the key feature of Model View Definitions?

1. They define top-bottom-perspective views for a 3D model

2. They define how the same data model can be loaded (e.g. visualised) in different ways

3. They define for a particular BIM model how it should be rendered by clients

4. They support user-defined views for 2D, 3D and 4D models

Question 11 (true/false)

The Model View Definition specifies how disciplines view data. They need different IFC data models for this.

1. True

2. False

Question 12 (MC)

The advantage of the existing IFC is that it is construction domain specific. Why is this also a disadvantage? Which statement is incorrect:

1. It results in the limitation to integrate data from other sources within the IFC schema

2. It results in the limitation that it cannot be extended easily

3. It results in the limitation that it is not suited for most recent BIM systems

4. It results in the limitation that it cannot be used in a linked data approach

Question 13 (check box)

Linked data is used to combine buildings data sets and other data sets. Which statement is true?

1. Linked data uses machine learning and therefore automates the process

2. Linked data defines structures that allow data sets with a different format to be combined

3. Linked data requires that all combined data sets are stored on one central location

4. Linked data uses virtual reality to visualise various data sets in one screen

Question 14 (MC)

1. A uniform resource identifier describes one set of subject, object and predicates

2. A uniform resource identifier defines the unique ID of a data piece

Choose the correct answer:

1. Statement I is true and II is false

2. Both statements are true

3. Statement I is false and II is true

4. Both statements are false

Question 15 (mc)

Linked data essentially relates data sets by defining the relationships between the different conceptual models behind these distinct data sets. What is the name of the process of retrieving, filtering these linked data?

1. Querying (through SPARQL)

2. Ontological modelling (through OWL)

3. Big data analysis (through machine learning)

4. Building Information Modelling (through IFC)

— End

Appendix 2: post-test lecture 3 inter-operability and standardization/IFC

Question 1 (MC)

BIM systems use schemes such as typologies, taxonomies, ontologies, object libraries, and data models. What is the commonality between such schemes?

1. Their purpose is to structure specific, project-related object-knowledge

2. They have no similarity

3. Their purpose is to structure conceptual, project-independent knowledge

4. They are the main cause of the inter-operability problem

Question 2 (MC)

Information standards in BIM:

1. Need to be redefined each and every project by all stakeholders

2. Should be defined as such that they work for a larger set of common projects in an industry

3. Should be defined by the BIM-software developer as he is the most knowledgeable about construction collaboration

Question 3 (MC)

What does the term ‘proprietary format’ mean in the context of BIM? Proprietary format means:

1. that a BIM model has one owner who is legally responsible

2. that BIM model exchange takes place in the closed format of a software developer

3. that the BIM format that is used to develop a real estate property

4. that a BIM model has multiple owners who are legally responsible

Question 4 (MC)

Which of these examples describe a situation of ‘single platform closed BIM’ exchange?

1. An engineer exchanges data by using direct links within a structural design and infrastructure design suite of one single software supplier

2. An engineer exchanges data between his two different suites for structural design and infrastructure design by using the standard proprietary format of one software supplier

3. An engineer exchanges data by using lock-in techniques between an open structural design suite and closed infrastructure design suite

4. An engineer exchanges data between his structural design suite and infrastructure design suite by using an open standard

Question 5 (true false)

Open formats are commonly flexible than proprietary formats as they are adjustable by all developers.

1. True

2. False

Question 6 (multiple answers possible)

Why does a BIM system user want to use a commercial system despite the threat of facing lock-in problems? Mark the correct answer, there are multiple possible answers:

1. It offers a user specialised software packages for his engineering and design needs

2. It offers a user the best way to exchange data with other commercial systems

3. It provides a better support and training compared to most open systems

4. It is cheaper than the open systems that are available on the internet

Question 7 (MC)

What is the main different between the ‘old CAD based’ exchange of data and the ‘new BIM based’ exchange?

1. The old approach exchanges points and lines; while the new approach exchanges computer interpretable object data

2. The new approach exchanges points and lines; while the old approach exchanges computer interpretable object data

3. The old approach exchanges paper information while the new approach exchanges information digitally

4. The old approach exchanges complex data, the new approach simplifies the way in which data is stored

Question 8 (MC)

Language, process and data are important parts of establishing a digital communication flow between systems. IFC, the industry foundation class, is relevant to thee three aspects of digital communication flows. What is IFC:

1. An international object-model for describing data exchange processes

2. An international object-model for describing building terms and language

3. An international object-model for describing building data

Question 9 (MC)

1. IFC is specific to projects and needs to be refined for each different project

2. IFC structures building data from very global to specific

Choose the correct answer:

1. Statement I is true and II is false

2. Both statements are true

3. Statement I is false and II is true

4. Both statements are false

Question 10 (MC)

What is the difference between the specific object breakdown structure (OBS) for your design project, and the breakdown structure as used by IFC?

1. The OBS is project-specific, the IFC is a conceptual standard, which is applicable to multiple buildings

2. The OBS has a different scope and level of detail than the conceptual IFC standard.

3. The IFC is project-specific, the OBS is a conceptual standard, which is applicable to multiple buildings

4. There should be no differences, they should be identical

Question 11 (MC)

What is the key feature of Model View Definitions?

1. They define top-bottom-perspective views for a 3D model

2. They define how the same data model can be loaded (e.g. visualised) in different ways

3. They define for a particular BIM model how it should be rendered by clients

4. They support user-defined views for 2D, 3D and 4D models

Question 12 (MC)

Which data sources may be relevant to construction that are not included in IFC Building schemas yet? Mark correct answers:

1. Design data (e.g. geometry)

2. Supply chain data (e.g. manufacturing details)

3. Geographic data (e.g. project location data)

4. Construction data (e.g. costs, phasing)

5. Sensor data (e.g. real-time performance data)

Question 13 (link)

Linked Data uses the subject-predicate-object structure to describe data pieces. Link the subject, object, and predicate in the following statement: roof2 is on top of level3

Subject –

Object –

Predicate –

Question 14 (MC)

• III. A uniform resource identifier describes one set of subject, object and predicates

• IV. A uniform resource identifier defines the unique ID of a data piece

Choose the correct answer:

1. Statement I is true and II is false

2. Both statements are true

3. Statement I is false and II is true

4. Both statements are false

Question 15 (mc)

Linked data can support advanced analysis of linked data sets. How advanced is its application in the practical construction domain?

1. It is only hypothetical and theoretical

2. It is implemented in various pilot (study) projects

3. It is implemented internationally practice on a larger scale

Question 16 (true/false)

The traditional IFC standard and more recent linked data approaches try to overcome inter-operability issues. Both approaches require that a conceptual data model is defined to facilitate collaboration.

1. True

2. False

— End

Appendix 3: focus group questions

The description above leads to the following questions for the focus group

Part one (to get the discussion started): students’ philosophy on lectures

1. What in your view are the requirements for a good lecture?¹

2. How does the format in which the lecture content is presented matter in this regard?

On the microlecture / flipped classroom format

1. Do you think that working through the lectures before the in-class session makes it easier to, or harder to, understand the workshop material?²

2. Do you think that the new format has or has not helped you to better understand the course material so far?²

3. Did you encounter any difficulties while watching the microlecture? If so, which?

   1. Think about technical problems (e.g. web connectivity, slow performance, visibility of slides)²

4. Compare this microlecture approach with the traditional face-to-face lecture. What are your experiences?¹

   1. Consider that you can choose yourself when you watch the video-lecture³

   2. Consider the speed of lecture

   3. Consider technical needs

5. How enthusiast are you about the usage of the microlecture concept for a class like this?⁵

Generalisation to other lectures/students/courses

1. Would you like to take another lecture using the same approach? Please explain³

2. Would you recommend others to follow the microlecture class as well?

3. Do you think that this model is successful in engineering courses? Please explain⁴

Footnotes indicate that questions were adapted from the following previous work on impact of flipped classroom, multimedia learning and perceived performance and satisfaction.

1. Bishop-Clark, C., and Dietz-Uhler, B. (2012). Engaging in the scholarship of teaching and learning: A guide to the process, and how to develop a project from start to finish, Stylus Publishing, LLC.

2. Lucke, T., Dunn, P. K., and Christie, M. (2017). ‘Activating learning in engineering education using ICT and the concept of ‘Flipping the classroom’.’ European Journal of Engineering Education, 42(1), 45–57.

3. Ravishankar, J., Epps, J., and Ambikairajah, E. (2018). ‘A flipped mode teaching approach for large and advanced electrical engineering courses.’ European Journal of Engineering Education, 43(3), 413–426.

4. Newman, G., Kim, J.-H., Lee, R. J., Brown, B. A., and Huston, S. (2016). ‘The perceived effects of flipped teaching on knowledge acquisition.’ Journal of Effective Teaching, 16(1), 52–71.

5. Tune, J. D., Sturek, M., and Basile, D. P. (2013). ‘Flipped classroom model improves graduate student performance in cardiovascular, respiratory, and renal physiology.’ Advances in Physiology Education, 37(4), 316–320.

Appendix 4: links to microlecture content

Table

https://vimeo.com/298123819/76a354ab76	[inter-operability]
https://vimeo.com/298123781/2c92ba52dc	[open data structures]
https://vimeo.com/298123757/10e13a94e8	[open versus closed BIM]
https://vimeo.com/298123720/e58e865cbf	[viewing BIM data]
https://vimeo.com/298123678/1036e21e58	[linking BIM]