Keywords

1 Introduction

Museums as informal learning environment to include learners of various disciplines and levels, have gained significant attention from both the research and practices of education [1, 2]. How to manage the learning materials, how to cooperate appropriate technology in exhibition and education have become important concerns for museums. Combing tangible user interface (TUI) and augmented reality (AR) technology, recently tangible augmented reality has been used in displays in exhibitions [4, 5]. Tangible AR using sandbox allows users to control both real and virtual objects with their bare hands to obtain real time feedbacks. For the applications of geoscience, for instance, sandbox-style TAR enables learners to shape various landforms and understand their structures at the same time. Apart from traveling to real fields, students can also create a field indoors by sandbox-style TAR [3, 6,7,8].

With sandbox-style TAR in museums, it is possible for visitors to experience the outdoor setting in a simulated yet instructional environment, and it also extends the physical entities of museum artifacts and buildings to a boundless and resourceful learning environment. Stimulated by the aforementioned potentials which is regarded as an effective way to interest students in science learning. Its quality, hence, plays a critical role, and the effects which sandbox-style TAR possess would be important and worthy of exploration. Stimulated by the aforementioned background, this study intends to design and develop a tangible AR environment for students’ geoscience learning in museums.

2 Literature Review

2.1 Sandbox-Style Tangible AR

Since tangible AR is a kind of AR whose user interface is tangible [4], its development is aligned with tangible user interface [9]. Sandbox-style TAR comes into use in the form of a sandbox with a projector and a depth camera hanging above. The continuous, kinetic materials such as sand or clay in the sandbox give users more flexibility in manipulation [10]. Considering people’s instinct to explore the world by touching, tangible user interface allows users to control digital information directly rather than using media like mice and key-boards which are necessary in other modern user interfaces [11]. The materials in sandbox-style tangible AR need to imply positions information so that the information can be controlled while users are manipulating the materials. And then the projector will give feedback once changes are sensed by the depth camera, making it possible for users to manipulate virtual objects in AR [12].

Basically in a sandbox-style tangible AR environment, the manipulation is that users change positions of materials and see feedback directly projected in the sandbox. The team of Tangible Landscape, however, has enabled other more modes of interaction and types of feedback [8]. In their original design of Tangible® [8], there are six kinds of interaction and they can be further classified into three categories based on the similarity between modes. First, what users can do in category one is change the form of materials whether they use their bare hands or tools, which is the basic design of sandbox-style TAR. And the second category allows users to place several markers such as pins, patches or outlines on the materials; different markers have their own functions (e.g.: river, forest…etc.). The third category includes the interaction that users establish viewpoints in the sandbox, in which external monitors are necessary.

The current ways how sandbox-style presents feedback can be divided into two. Users can get feedback either from the surface of materials or from a monitor aside [8]. The former is the common way to present feedback, serving as the basic setup of sandbox-style TAR; however, the latter, for certain purposes, can simulate what virtual or analog objects will look like in reality on a monitor equipped. Users who intend to experience the third category of sandbox-style TAR’s modes of interaction will get corresponding feedback only if there is a monitor displaying changes of viewpoints.

Recognized by its positive effects, sandbox-style TAR is used to train a variety of abilities. For one, users acquire such basic skills as students use to understand topography. Users show their better abilities to read contour lines via sandbox-style tangible AR. On the other hand, ones are more likely to have difficulties when using planar materials [13, 14]. For another, users also undergo drills in engineering. And in the study of [8], the results suggest that work made by users trained with sandbox-style TAR tend to be of higher quality than that made with either analog or digital methods.

In view of the above, it implies that the purposes of sandbox-style TAR involve both theoretical and practical aspects. And its users are mainly aimed at college students or above. The user studies which have been conducted so far show that college subjects gain positive experience from sandbox-style TAR [8, 13, 14], while other researches discussing TAR’s application indicate that it is of the design of intuitive manipulation and can be easily accessed [4]. The state that few studies have dealt with other groups of users doesn’t seem to be consistent with such original intention, however.

2.2 Learning with Sandbox-Style Tangible AR

In the research of [15], the factors influencing students’ motivation in science learning are divided into six constructs: self efficacy, active learning strategies, science learning value, performance goal, achievement goal and learning environment stimulation. In this research, we focus on the last one given sandbox-style TAR’s role of a pedagogic medium, which can be a stimulus in a learning environment.

TAR generally features four attributes, namely tactile richness, manipulability, real-time computational analysis, and multiple access points [4, 8, 12,13,14]. Via tactile contact, users can feel material qualities. Manipulability allows users to carry out manipulation by their own volition and learn by doing. Real-time computational analysis refers to the immediate and relevant information tangible augmented reality generates after computational analysis. Multiple access points enable all users of tangible augmented reality to see what’s going on and get their hands on the central objects of interest. The combination of manipulability and real-time computational analysis, in particular, reflects the essence of constructivism, which considers that human learn from the accumulative experience of observing, interpreting and processing. And such learning theory has been recognized in geoscience education [5]. That is, tangible augmented reality can be an ideal medium of learning geoscience: users may have various interpretations following their observation on the information given by the system, and then they can process what has been done accordingly. Several empirical studies, in correspondence, have indicated its positive effects on geoscience learning [8, 13, 14].

Several characteristics of sandbox-style TAR promote students’ motivation to participate in the class based on the construct of learning environment stimulation in [15] ’s questionnaire. Exciting and changeable content and opportunities to join discussions can be facilitated via sandbox-style TAR’s tactile richness and multiple access points respectively. Manipulation out of students’ volition can lead to less pressure from teachers, and system’s real-time feedback draws students’ attention on what they have done. Sandbox-style TAR promotes students’ learning motivation. In addition, the level where students engage in the class is considered to increase when they get interested or motivated [15]. That is, the more sandbox-style TAR motivates its users, the more the users get engaged in learning by manipulating the device, which can be another beneficial effect of sandbox-style tangible AR.

Regarding the learning effectiveness, previous studies have indicated that high levels of both motivation and engagement can lead to students’ greater achievements [15]. In most countries around the world, geoscience basically shares the same syllabus with science [3]; students’ achievements in geoscience, thus, are evaluated in a way similar to those in science, which are judged from three aspects: attitude, process and knowledge [16,17,18]. Among the three aspects, different levels of knowledge of a certain scientific concept, including knowing, understanding or applying [18], are regarded as a common indicator implying students’ achievements in many studies as well as in studies about sandbox-style tangible AR’s learning effects [8, 13, 14]. With a unique environment, sandbox-style tangible AR’s learning effects, however, should involve wider aspects, which has rarely been dealt with in previous studies yet.

Therefore, this study attempts to discover other possibilities and would like to put emphasis on the aspect of process because learning by doing/manipulating, one of sandbox-style tangible AR’s characteristics, is believed to improve this kind of achievement. The aspect of process tackles students’ scientific process skills, referring to the abilities to facilitate a scientific study [16]. With reference to the curriculum guidelines for primary and secondary education of science and technology in Taiwan, five scientific process skill are emphasized and included: observing, comparing & classifying, organizing & relating, generalizing & inferring and interpreting. Stimulated by the aforementioned issues, the purpose of this study is to explore sandbox-style tangible AR’s effects on learners who visit a museum. In the literature review above, studies have confirmed that sandbox-style TAR does have positive learning effects, yet several questions still remain unknown.

3 Method

Based on the literature review, this study intends to design and develop a tangible AR environment for students’ geoscience learning in museums. With reference to the basic functionality of a tangible AR, and the possibilities of AR to support meaning making for science learning, this study firstly developed a sandbox-style tangible AR environment to enable basic interactivity. And the school curriculum of geoscience in general, and topographic maps in specific, in both local and global communities [19, 20] was then referred in order to develop the instructional plans with the sandbox TAR. Finally, there were four lessons in total completed with expert teachers’ review. These four lessons consisted of different infusion of tangible AR elements that allowed learners to search, connect, collect and even generate. A preliminary user testing with 5 college students were conducted to confirm the system instruction and functionality.

Instruments:

The system is constructed on the platform and framework of ARSandbox, developed by Oliver Kreylos in University of California, Davis, U.S.A. The sandbox in use of this study is 75 cm(L) × 60 cm(W) × 160 cm(H) with 20 kg kinetic sand loaded. Microsoft Kinect Xbox 360 ® was set up to sense the surface change and transmit the data back to the Linux server. An ultra short throw projector with high luminous flux of 3000 lm was set on top of the sandbox to project visual information. The final system outlook is shown as Fig. 1 and Fig. 2. Physical objects as markers, including thumbtacks, shovels, flags, rulers and blocks were also distributed for users to conduct manipulations like indicate, measure and shape.

Fig. 1.
figure 1

The experiment setting of this study.

Full size image
Fig. 2.
figure 2

The ARSandbox used in this study

Full size image

4 Results

With references to the school curriculum of topographic maps [19, 20], there were four instructional plans of typographic maps developed to coordinate with the ARSandbox (see Table 1). The instructional objectives of all lesson plans were to acquaint learners with topographic maps, and these four lesson plans distinguished from each other by their required manipulation tasks of searching, connecting, collecting and generating, respectively.

Table 1. Learning goals of the experimental instructional plans
Full size table

In the searching tasks, learners were asked to seek possible water flows and watersheds from a landform with a topographic map (see Fig. 3). In the connecting tasks, learners needed to shape a terrain according to a given topographic map (see Fig. 4). For collecting tasks, learners were asked to judge the visibility of an assigned spot from three alternative reservoirs by collecting the information of heights and profiles of the landform (see Fig. 5). And in the generating tasks, learners needed to determine where to build a dam with evaluating the available information such as slope, catchment area, and cost of construction, all provided by the system (see Fig. 6).

Fig. 3.
figure 3

Searching

Full size image
Fig. 4.
figure 4

Connecting

Full size image
Fig. 5.
figure 5

Collecting

Full size image
Fig. 6.
figure 6

Generating

Full size image

These four lesson plans were reviewed by 2 subject experts of geoscience education to ensure the validity and independency of each lesson plan. Dynamic and immediate feedbacks were projected to the surface of sand when users activated the raining by putting their hands over the box. Static and real time instruction and feedback were shown on the display in front of the participants. In order to ensure the user flow and system functionality, 5 college students were invited to use the sandbox with their behaviors observed and recorded. The results of the user testing informed the technological affordance of the physical markers, and the modification of the physical markers and their functions were made accordingly as shown in Table 2.

Table 2. Physical markers used in each lesson plan
Full size table

5 Conclusion

This study reported our design and development of a series of geoscience learning materials using ARSandbox. The four AR learning packages of topographic maps with careful instructional design were developed and assessed by subject experts to ensure the validity of the content as well as the interactivity of the system. From the interviews with the users in pilot study, it was also proved that the learning tasks of manipulation engaged the participants and motivated their problem solving skills. With the instructional plans ensured, in our future works, the user studies will be conducted with a larger population of high school students who are learning geoscience.