Keywords

1 Introduction

In the last 20 years, the development of technology considerably changed the way people interact to each other. Before that time, mobile phones had very little use besides making calls, fewer people had Internet access and most of online communication was done by email. Today people communicate with each other by sending and receiving messages through various forms such as text, audio, images, video or symbols, synchronous or asynchronously.

During this period, education has always kept its doors open to the use of new technologies such as simulators, just to mention a single example. In spite of that, communication between teachers and students in class seem not to have taken full advantage of the possibilities the advances of technology have brought. Most of the classes are still based on oral lectures, using few visual aids, where teachers just speak and students ask their questions when they think something is not clear [1]. This type of class hardly draws students’ attention [2] and they, as a result, usually have a low rate of content retention [3].

Students commonly feel little involved in the communication process of more traditional teaching methods since, in general, they allow them have a more passive posture towards their teachers that, on the other hand, must be very active in their classes in order to keep students motivated [3]. Also, those methods do not usually offer the students many opportunities to apply that recently acquired knowledge to real life situations, leading to a superficial learning level that could be represented by the initial categories of the cognitive domain of Bloom’s taxonomy [4], for instance.

Receiving considerable attention over the past several years, active learning has found many advocates among faculty looking for alternatives to traditional teaching, even though there may still exist some that regard it as just an educational fad [5].

Human-Computer Interaction (HCI) is an important area in computer related degree programs, such as computer science [6], information systems [7] and software engineering [8]. Despite its importance, there seems to be a lack of literature on practice-level issues about its implementation in the classroom [9]. Likewise, technological tools specifically designed to aid HCI teaching are practically nonexistent [10]. Considering this scenario, the present paper conducts a systematic mapping study that aims at checking how HCI is being taught and if any tools to support the process are being used.

The remaining of this paper is organized as follows: Sect. 2 presents related work. Section 3 describes the systematic mapping with the presentation of its results. Section 4 discusses the implication of the results, answering the research questions and presenting the threats to the validity of the study. Conclusions and comments on future work are described in Sect. 5.

2 Background

The importance of HCI can be stated when looking at the Curriculum Guidelines for Undergraduate Degree Programs for Computer Science [6], for Information Systems [7] and for Software Engineering [8]. Computer science students are expected to have at least 6.4 lecture hours of HCI core topics and have a choice of some other elective ones [6]. When majoring in information systems, the students might take Introduction to HCI as an elective course, though it brings significant coverage to the application developer and user interface designer career tracks, whereas the program for software engineering [8] considers that it is essential that students have 10 hours of HCI.

By looking at those programs, and understanding the role HCI plays on interface designing, one might think that a considerable number of software developers have at least been initiated on the matter. Edwards et al. [11], however, argues that this might be true, but teachers are failing at HCI education. As an evidence of his position, he points out the abundance of poorly designed interfaces and great number of papers published in the HCI literature that do not offer much more than criticisms of interface designs. He concludes that there is a vast scope for improvement in HCI teaching.

Battistella and Wangenheim [12] carried out a systematic review of the literature to understand what kind of games are available for teaching computing in higher education. They encountered a considerable number of 107 games, indicating that there exists a trend to game-based learning also in computing education. However, only one ranked match in the HCI area was found. The game called “3DAR Lego Game” provides a tool to improve spatial ability for a wide range of ages [13].

In 2012, Sommariva [14] conducted a systematic mapping study to search for games or simulators specifically developed to support usability teaching, and also to understand what usability topics were being taught and how. His study found no games developed to support usability teaching. He later developed and proposed a game to help teach usability engineering life cycle [10]. When it comes to usability teaching, his study focused on the activity proposed by the teacher, without mentioning the underlying teaching approach. Later, in 2014, Ferreira et al. [15] presented a game to teach Jacob Nielsen’s heuristic evaluation. The game makes use of analogies to reinforce the heuristics understanding, building a story to motivate the students during their learning process.

Nevertheless, systematic reviews (or systematic mappings) that analyzed approaches and computational solutions specifically for teaching HCI were not identified. Therefore, this article proposes: “Let’s talk about tools and approaches for teaching HCI”.

3 Systematic Mapping

This work presents a systematic mapping study carried out to get an overview of HCI classes at the undergraduate level. A systematic mapping is usually used to investigate a wider research area than a systematic literature review when little or no evidence on a topic is known [16]. The interest here is to get an indication of the quantity of that evidence and classify it for further studies.

3.1 Planning and Conducting the Mapping

The phase of planning the mapping consists of developing a review protocol which defines the methods to undertake a specific systematic review. Having a well defined protocol reduces the possibility that this review can be driven by research expectations.

Review Objective and Research Questions. The main objective of this mapping is to identify how HCI is being taught to undergraduate students. In order to guide this analysis towards its achievement, it is important to know what are the teaching approaches that are mostly used in class and what tools, if any, are used to support it. With this in mind, three research questions are proposed:

  1. RQ1

    What are the main approaches used to teach HCI to undergraduate students?

  2. RQ2

    What are the available technological tools specifically developed to support HCI teaching?

  3. RQ3

    How have the approaches or technologies used to teach HCI been evaluated?

Search Strategy. An automated search strategy was used in three different scientific databases: ACM Digital LibraryFootnote 1, IEEE Xplore Digital LibraryFootnote 2, Google ScholarFootnote 3. These results were complemented with manual search on selected conferences:

  • International Conference on Applied Human Factors and Ergonomics and the Affiliated Conferences, AHFE 2015 [17] and AHFE 2017 [18];Footnote 4

  • HCI International 2013–2017 [19,20,21,22,23,24,25];Footnote 5

The search string used is divided into three parts:

  1. Part1

    This part of the string used expressions relating HCI and teaching. By using only the expression “HCI”, without relating it to teaching, most of the studies returned by the search discussed on the applications in the area, but outside the education context. If on one hand the search was restricted, on the other hand, it was extended by adding term “usability”. The choice for this term is supported in [26], that states “Usability and HCI are becoming core aspects of the system development process to improve and enhance system facilities and to satisfy users’ needs and necessities.”

  2. Part2

    Here the string was limited to search for papers that dealt with teaching undergraduates.

  3. Part3

    The last part had the intention to find among those papers the ones that explicitly mentioned how their teaching was conducted, using some kind of tool, approach or both.

In this way, the string used for the search was:

(“teaching hci” OR “hci teaching” OR “hci education” OR “teaching usability” OR “usability teaching” OR “usability education”) AND undergraduate AND (software OR game OR simulation OR tool OR environment OR methodology).

To complement the automated procedure, a snowballing search, backward and forward, was conducted following the guidelines in [27]. All the works selected from the automated search were used as seeds.

Inclusion and Exclusion Criteria. Some inclusion and exclusion criteria were defined for selecting papers for the final review.

The mapping should only include:

  • Peer-reviewed publications appearing in journals, conferences, and workshops;

  • Publications written in English;

  • Papers published from 2012 to 2017;

  • Papers that discussed about or presented a methodology or an approach for teaching HCI on undergraduate level;

  • Papers describing computational solutions to support HCI teaching.

The following works should be excluded from the mapping:

  • Presentation documents, such as PowerPoint slides, and short/extended abstract papers;

  • Papers that did not relate to undergraduate HCI courses;

  • Papers that, although related to HCI teaching undergraduates, discussed about other aspects of the course or program, such as curriculum proposal, for instance;

  • Papers that described computational solutions applying concepts related to HCI area, but that did not focus on teaching those concepts.

Data Extraction and Data Synthesis Strategies. The mapping was conducted from December 2017 to February 2018 with the execution of the protocol, and resulted in the selection of 21 papers for data extraction. However, during the data extraction process it was observed that in three selected studies there was no consistent information for extraction, that is, the information was not sufficient to respond to the RQs. So these 3 papers were excluded. The number of retrieved, examined and selected papers from each resource is summarized in Table 1. The selection process is illustrated in Fig. 1.

Data were extracted from each selected paper following 15 previously defined information items (Table 2).

Table 1. Search result
Table 2. Data extracted from each primary study selected

Results. Table 3 summarizes the data extracted for the 17 articles selected considering some information from groups 1 and 2 of Table 2:

  • The Context column indicates where the research described in the paper was conducted, i.e., which course, with its major in parenthesis;

  • The In-Text Topics column presents those HCI topics the authors mentioned they had worked with the students during their research;

  • The HCI Topics column classifies the topics surveyed by considering an HCI course in a curriculum guideline for undergraduate degree programs [7];

  • The methodologies/approaches used to teach HCI are in the column Approach;

  • The last column, Tools, indicates the tools used during the teaching process that were mentioned in the text.

Table 3. Selected papers summary

During the period researched, the publication of selected studies seemed to have been fairly regular, with 3 to 7 papers being published a year. The exception was in the 2013–2014 period, when no papers were published (Fig. 2).

Fig. 1.
figure 1

Process of selecting primary studies

From the 17 selected papers, 12 (70.5%) conducted their research in a HCI course, whether mandatory or elective, whereas 5 of them (29.5%) had their researches in some other courses (technical communication [TC], system design [SD], interaction design [ID] or winter school), where some topic of HCI was taught.

4 Discussion

With the data extracted from the selected papers it is possible now to answer the research questions.

RQ1 :

What are the main approaches used to teach HCI to undergraduate students?

From the data extracted here, a wide variety of teaching approaches or techniques can be identified (Fig. 2).

Fig. 2.
figure 2

Approaches used in selected studies

Most articles mentioned the use of a specific approach (blue dots in the graph of Fig. 2). However, some articles did not cite a specific approach, and in this case, they were ranked from the analysis of the text (red dots in the graph of Fig. 2):

  • Experiential activities: experiential learning is a guided process of questioning, investigating, reflecting, and conceptualizing based on direct experiences. In this learning process, the learner is actively engaged, has freedom to choose, and directly experiences the consequences of their actions [45].

  • Hands-on experiences: development of projects with real users, that is, real-world cases.

  • Project development: development of projects without considering real users.

  • Serious game: use of serious game (educational game) to support teaching.

Among the articles that did not mention a specific approach, the teaching of HCI through the development of group projects using as case study real needs, that is, real-world cases [33, 37, 42] can be highlight as the most recurrent.

Nine different approaches were cited in the selected studies. For Zaharias et al. [32] problem-based learning develops “experiential and social learning and calls for an active rather than passive approach to learning that leads to the development of critical thinking skills”. Jeon [38], on the other hand, defends that although a problem-solving approach may not work for novices (i.e., undergraduates), “they are still encouraged to be involved in the overall design process, but they start with a focus on part of the problem with well-defined guidance based on the scaffolding approach.” Culén [44] proposes the use of design thinking that “stands firmly on three main pillars: empathy with users and human centeredness, rapid prototyping to generate large number of alternatives in order to solve the right problem rather than a problem right, and last, but not least, their synthesis leading to best viable and feasible solutions that incorporate desired values” [44].

Chong [9] mentions the use of user-centered approach, whereas, Taylor et al. [43] prefer student-centered learning which are research-led, problem-based and flexible, where students can focus on the topics and content delivery methods which are most interesting and useful to them. Alnuain et al. [36] apply blended learning model “a learning approach that contains different types of education techniques and technologies. This learning model aims to provide more effective education experiences by combining features and functions of well-known learning and teaching techniques” [46].

Studio-based learning was the most cited approach [31, 40, 44]. Adapted from architecture and art education, as well as from collaborative problem-solving pedagogies, studio-based learning has shown great promise for computing education [47]. Key elements of studio based learning include exploring multiple solutions to a problem, justifying the choice of one solution, and being subject to, as well as providing, peer reviews [47].

Yang et al. [41] propose a new approach, called interest-based learning, in this approach “students are guided to organize teams by same personal hobbies or specialties, e.g. music or sports, and conduct research on HCI in these familiar and interested topic.”

It can be observed from this analysis that the teaching of HCI to undergraduates has used some kind of active learning approach, considered here as “any instructional method that engages students in the learning process”[5]. The popularity of active learning among HCI teachers seems to be clear when assuming that this approach tries to involve the students in a way that is closer to the professional environment. Most of the importance of HCI for the IT industry is related to practical activities such as interface prototyping and development or usability testing, for instance. The approaches identified on the selected papers try to actively involve the students at doing something rather than passively receiving information. For Hundhausen et al. [31], giving the students a more central participation “is akin to the situation of an expert designer presenting a user interface design to a design team in a real-world company.”

RQ2 :

What are the available technological tools specifically developed to support HCI teaching?

Although HCI is closely related to computers and some other technological devices, the use of tools specifically developed for HCI teaching does not seem to be very widespread. From the selected papers, only 6 mentioned some kind of tool during the process and, among them, only 2 described tools that were specifically designed for the purpose.

One of those tools is a serious game, called UsabilityGame [29], that offers the students the opportunity to practice the usability lifecycle by addressing requirement analysis, prototyping and heuristic evaluation. When playing the game, the students have to select the correct set of requirements for each scenario presented, develop a prototype to be evaluated by the teacher and evaluate real-world interfaces by choosing one of Nielsen’s heuristics [48] that would correctly improve the detected usability issue. The main role of the teacher, during the game, is to set up the game according to the student’s learning objectives, to conduct the evaluation of the prototypes, and monitor the performance of the class.

The other tool is WOZ Pro (Wizard of Oz Prototyper) [31], a low-fidelity prototyping environment that was developed to support prototyping creation and wizard of oz testing. Although it was conceived to be used by college students who are first learning about interaction design, this tool could also be used outside the educational context [49].

A few other tools such as an environment for editing and analysis of task models (CTTE – ConcurTaskTrees Environment [30]), an e-learning environment (SCELE – Student Centered e-Learning Environment [33]), a virtual world platform (OpenSimulator [32]) and even simple art supply [49] were mentioned to be used during the teaching process and they can all be as valuable as the computer-based tools described above. However, HCI teachers seem not to have many choices when picking out some supporting tool that could meet their educational purposes.

RQ3 :

How have the approaches or technologies used to teach HCI been evaluated?

The selected studies were also classified following the research type classification proposed by [28]. According to this study, most of the primary studies (76%) were classified as experience papers, presenting the authors’ experiences in class. The importance of these papers rely in the relevance of the lessons learned by the author from that experience [50]. Three other studies (18%) were classified as evaluation research, with the proposal of a novel piece of knowledge in HCI education. At last, 1 study (6%) was classified as an opinion paper (see Table 4).

Despite the importance of experience papers, that contribute with experiences of in-practice approaches, revealing their results when applied in real contexts, the number of studies evaluating new developments in HCI education does not seem to be very large. This high percentage of experience studies on HCI education seems to indicate that researchers are mostly concentrated on applying their efforts to understanding the area, producing data to pave the road of further developments.

When evaluating the result of the applied approaches: (i) a simple survey was used in 7 studies [9, 34,35,36,37, 41, 44] without statistical tests; (ii) 4 studies applied statistical analysis to evaluate their results (Wilcoxon test [29], Pearson Product-Moment Correlation Coefficient [40], Kruskal-Wallis, Mann-Whitney [42], chi-squared test and ANOVA [31]); (iii) interview [9, 40] and structured interviews were used by [32]; (iv) observation was the technique used by [32, 36, 40]; and (v) content analysis (video) used by [31].

The size of the sample used by the researchers ranged from 8 [44] to approximately 1,150 [35] students participating in the process. When analyzing the mean and the median numbers (159 and 40, respectively), they are considerably higher than those of the studies published at CHI2014 [51].

Table 4. Research type classification [28]

4.1 Threats to Validity of the Mapping

One of the major problems with systematic reviews is finding all the relevant primary studies (evidence). In this case, three search strategies were used to ensure that the largest number of related studies would be found: (i) Automatic search in 3 different databases; (ii) Manual search in specific proceedings of two conferences of HCI area not indexed on the searched databases; and (iii) backward and forward snowballing.

In addition, our search strings were designed to find the maximum number of relevant papers. Nonetheless, it is possible that they have missed studies that used different terminology to describe any part of the string.

The search resulted in some papers that, in spite of having their titles and abstracts in English, were actually written in other languages. Those papers were excluded from the selection, but, due to language limitations, it was not possible to assess whether they had relevant information or not.

Readers must consider that a systematic review is by definition limited by the search date, the electronic sources and key terms used in the search. Therefore, it is possible that other papers may be included in a future replication of the study. The results are limited by the previous features, and by the evolution of the HCI education area itself.

5 Conclusions

Although HCI teaching is present in many computer-related programs, it seems that HCI teaching in undergraduate level has not been drawing enough attention from researchers, what is evidenced by the small number of selected papers. This conclusion gets patent when looking at the dates of the publications and realizing that for two years in a row there were no selected publications on the theme. The lack of selected papers from the proceedings of prestigious conferences shows that the road is still open for research in HCI education.

Most of the selected papers presented researches on either development or evaluation methods teaching. Nevertheless, the teaching of some topics, such as Relevance of HCI and Devices, was not mentioned. The lack of research on those topics does not mean they are less important than others, it simply means that there is a field ahead to be explored.

Active teaching approaches as a whole seem to be the main option when teaching HCI to undergraduate students. The present study did not try to categorize the multitude of approaches or evaluate its results when applied to this area of knowledge, because the prevalence of one approach over the others could not be stated in small number of primary studies found. This categorization might eventually become an object of future work.

HCI teachers have very few options when choosing a supporting tool exclusively developed for their classes. The same does not happen in other areas of knowledge on computer field. On a paper of 2009, for example, Wangenheim and Shull [52] found 16 games used in software engineering education, most of them computer-based, but also card or board games. Although the paper concludes that more research in the area is necessary, it is clear that there were a higher number of educational games on software engineering 10 years ago than there are of educational tools for HCI today. Future work on these tools would bring great benefit for HCI teachers and students.

Finally, it is noted that few empirical studies have been conducted to verify the results obtained by the different teaching approaches and tools used. Thus, this paper suggests that more empirical studies be performed with sufficient rigor to improve the body of evidence in the HCI education field.