Edited by:
Reviewed by:
*Correspondence:
This article was submitted to Human-Media Interaction, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Special education teachers for visually impaired students rely on tools such as raised-line maps (RLMs) to teach spatial knowledge. These tools do not fully and adequately meet the needs of the teachers because they are long to produce, expensive, and not versatile enough to provide rapid updating of the content. For instance, the same RLM can barely be used during different lessons. In addition, those maps do not provide any interactivity, which reduces students’ autonomy. With the emergence of 3D printing and low-cost microcontrollers, it is now easy to design affordable interactive small-scale models (SSMs) which are adapted to the needs of special education teachers. However, no study has previously been conducted to evaluate non-visual learning using interactive SSMs. In collaboration with a specialized teacher, we designed a SSM and a RLM representing the evolution of the geography and history of a fictitious kingdom. The two conditions were compared in a study with 24 visually impaired students regarding the memorization of the spatial layout and historical contents. The study showed that the interactive SSM improved both space and text memorization as compared to the RLM with braille legend. In conclusion, we argue that affordable home-made interactive small scale models can improve learning for visually impaired students. Interestingly, they are adaptable to any teaching situation including students with specific needs.
With more than 285 million visually impaired people in the world, including 19 million visually impaired children below the age of 15 (
The two raised-line maps (RLM) with their corresponding legends used in the control condition. The first map
Tangible interaction has primarily been studied with sighted users, and has been defined as the interaction with a computer through the use of physical objects (
Recent studies have demonstrated the benefit of using tangible interfaces in the context of visual impairment.
As mentioned above, it has previously been shown that SSMs are beneficial for the acquisition of spatial knowledge by visually impaired people (
We recruited participants from an education center for visually impaired students (INJA in Paris, France). All participants or their legal representatives gave written informed consent to participate. The protocol followed the Declaration of Helsinki and was approved by the CLERIT ethics committee.
Twenty-four blind students (10 females; age range: 14–19 years;
We divided participants into two groups: the control group explored the RLMs, and the test group test explored the interactive SSM. Within each group of participants, we included five participants with low to medium spatial skills (hereafter called “non-experts”), and seven participants with a good to very good level of spatial skills (called “experts”; see
Number of early and late blind subjects with high (experts) and low (non-experts) spatial skills.
Condition | Blindness | Spatial skills | Number |
---|---|---|---|
Early | Experts | 5 | |
Early | Non-Experts | 4 | |
Late | Experts | 2 | |
Late | Non-Experts | 1 | |
Early | Experts | 4 | |
Early | Non-Experts | 3 | |
Late | Experts | 3 | |
Late | Non-Experts | 2 | |
Total | 24 |
We prepared this study in collaboration with a teacher specialized in low-vision education. His objective, beyond the scope of the current study, was to design his own teaching material based on low cost prototyping tools. The spatial layout represented the city where the students are living (Paris, France), in order to be used for giving Geography and Orientation lessons after the study. For the current study, we turned the layout upside down (i.e., North facing 6 o’clock) and created a scenario based on the history of a fictitious kingdom. We checked afterward that none of the students was aware that the model represented the map of the city they lived in. In addition to the specialized teacher, a person in charge of producing tactile documents checked that all materials (RLM, 3D model, braille booklet, text) were appropriate for tactile exploration by the visually impaired students.
The information content of the two conditions (control vs. test) was identical. Two levels of display described the kingdom. The first level showed the main landmarks and roads of the kingdom in its current state, as well as the river going from west to east. The second level illustrated the enlargement of the kingdom with the fortifications that were built over time. The evolution of the kingdom was associated to a complex history
The control condition encompassed two regular RLMs with a legend (
The interactive SSM was designed to be affordable and simple to make (
The first level of the interactive small-scale model created with a laser cutter and a wooden plate. This level
Before the experiment, we carried out a set of pretests in order to check that all documents and the prototype were usable. Pretests were completed by three sighted users that were blindfolded, and three visually impaired users that did not participate in the study. The main objectives of the pretests were: (1) assess the duration needed for the complete exploration of the maps; (2) verify the readability of the tactile maps (check that the different symbols and textures were distinguishable and readable), as well as the usability of the 3D model (check that there was no problem in triggering the audio feedback); (3) check that the whole experimental protocol was adapted to visually impaired people, and did not take too long. As a result of the pretests, we improved the prototype (reliability of interactive zones), and set the maximal exploration time during the experiment to 35 min. This duration was sufficient for a person with regular braille reading skills to explore the whole display and read the legends and booklet in both conditions.
As previously mentioned, participants were included either in the control group (RLMs) or test group (interactive SSM), which corresponds to a between-subject experimental design. Experiments were carried out individually in a quiet and separate room. The experimenter gave instructions, and read the questionnaires after exploration.
After exploration, participants were asked to recall the names of all roads, landmarks and ramparts that were present on the map, without receiving any clue from the experimenter. Following free recall, the experimenter read aloud the entire list of the points of interest in order to provide all participants with the same amount of lexical information before the next series of questions. Then, she asked a list of geographical and historical questions about the kingdom. For each question, the subject had to pick the correct answer among four propositions, thus for each question, the probability to answer correctly by chance was 0.25. The presentation order of the geographical and historical questionnaires was counterbalanced across subjects. The order of the questions and of the four choices were also counterbalanced across participants. The last questionnaire aimed to evaluate the subjective satisfaction (SUS) of participants. The whole experiment lasted about 85 min including map exploration.
We used two questionnaires to assess the memorization of geographical and historical knowledges. The geographical questionnaire was inspired by previous studies (
The three types of route (R) questions concerned: (1) Route distance estimation: Subjects had to select the two landmarks separated by the shortest route among four propositions; (2) Route recognition: Four routes between two points were described. Participants had to pick the correct one; (3) Wayfinding: a starting point, a destination, and the beginning of the route between these two points were provided. Participants had to choose the road that completed the route among four options.
The three types of survey (S) questions were: (1) Direction estimation: a starting point and a goal were given. Participants had to indicate the overall direction to the goal using a clock reference system (e.g., three o’clock for east) among four possible options; (2) Location estimation: a landmark was provided, and subjects had to choose the rampart that covered this landmark among four possible options; (3) Survey distance estimation: Participants had to choose the pair of landmarks that were separated by the longest distance (“as the crow flies”) among four propositions.
The historical questionnaire to evaluate the memorization of historical knowledge about the kingdom was inspired by
We also assessed subjective satisfaction for each device using the System Usability Scale (SUS,
As already mentioned, we made the general hypothesis that the interactive SSM would provide better learning for both geographical and historical knowledge. The main independent variable was the map type (RLMs vs. interactive SSM).
The dependent variables were the scores in response to geographical and historical questions (with a maximum of 24), and the score to the SUS (with a maximum of 100). We compared these different scores by map in order to test the following specific hypotheses:
Hypothesis 1: Users get more correct answers to a questionnaire about the geography of the kingdom after exploring the interactive SSM than after exploring the RLMs;
Hypothesis 2: Users get more correct answers to a questionnaire about the history of the kingdom after exploring the interactive SSM than after exploring the RLMs.
Hypothesis 3: Users get a higher score to the SUS questionnaire after exploring the interactive SSM than after exploring the RLMs.
Shapiro–Wilk test confirmed that each set of data (historical and geographical scores, SUS) was normally distributed. Statistical tests (
Percentage of correct answers to the historical and geographical questions, as well as SUS score depending on the condition (∗∗
Mean scores (
Dependent variables | Raised-line map | Interactive small-scale model | |||
---|---|---|---|---|---|
58% (12.7) | 72.6% (12) | -2.9∗ | |||
Overall | 28.8% (10.4) | 54.9% (10.3) | -6.14∗∗ | ||
Route | 32.6% (16.1) | 57.6% (14.4) | -4.01∗∗ | ||
Survey | 25% (12.3) | 52.1% (17.5) | -4.39∗∗ | ||
76.7% (15.1) | 85.4% (12.3) | -1.6ns |
Among the geography questions, the correct answers to Route [
It is interesting to note that the enhancement of spatial memorization provided by the SSM affects both the early and late blind, but also experts and non-experts.
Scores to geographical questions according to onset of blindness and spatial expertise. Error bars represent standard deviation. (RLM, Raised-Line Maps; SSM, Small-Scale Model). Effects of the Tukey
We carried out additional analyses on the recall of landmarks (i.e., number of correctly remembered names). The mean scores were 9.2 (
Finally, there was no significant difference for the SUS scores between the two conditions [
We also checked that results did not depend on spatial skill of the subjects. There was no significant correlation between spatial skills and score for the historical questions [
Users explored the interactive map with both hands, i.e., 10 fingers, exactly the same way that they would explore a paper relief map. The interactive map contained no Braille text but audio output was triggered when touching the conductive markers. No further input or output interaction was provided to ensure functional equivalence with the paper map.
We noticed that the participants first explored the content of the whole map with both hands, whether it was the 2D RLMs or the interactive SSM. Then, they typically focused on details. For the 2D RLMs, they explored a landmark on the map, then they moved back and forth between textual and pictorial elements in order to establish the correspondence. On the opposite, when they found a landmark on the interactive SSM, they kept the finger steady on that landmark while listening to the description. In addition, they sometimes moved the second hand to explore the space around that landmark. None of the participants tried to place the first 2D RLM on top of the second 2D RLM in order to make a correspondence between landmarks and ramparts, whereas for the SSM they removed and replaced the 3D pieces and explored which landmarks were under the rampart.
The participants reported that the interactive SSM was much better than the 2D RLMs because of the modularity provided by the 3D pieces, as well as its interactivity. Interactivity was appreciated because then back and forth movements between the map and the braille legend were not necessary. They reported that exploration was easier and faster. They also enjoyed to listen to the verbal descriptions (with earphones when they needed it due to surrounding noise or to focus). However, they declared that the interactive zones were sometimes too sensitive and should be improved.
The results show that correct answers to the historical and geographical questionnaires were significantly higher with the interactive SSM than with the RLMs. For the geographical questions, we observed significantly more correct responses for route and survey questions with the interactive SSM. Therefore, the interactive SSM provides an overall benefit for visually impaired students, and the effect size is large. Indeed, 63% of the variance was explained by the type of map for the geography score and 27% of the variance for the history score. These results are consistent with the results of
The SUS score was not significantly better for the interactive SSM, which is notable for two reasons. First, it confirms that visually impaired students were satisfied with the regular RLMs that were made for them, according to their specific needs, relying on the expertise of a person in charge of tactile document making. Indeed, they judged that RLMs were usable with a score of 76.7. Second, it showed that the students have not been sensitive to a “novelty” effect, which may be observed with new devices (
We did not observe any significant difference between the two conditions regarding the free recall of landmarks. Both devices allowed a good memorization of landmarks, independently of their location. In addition, the recall rate was correlated with history scores for the interactive SSM [
At least two main factors make interactive SSMs an appropriate teaching material for visually impaired students: (1) haptic exploration of spatial content, and (2) multimodal interactivity. First, several authors showed that the haptic sensory system is more efficient when exploring 3D objects than 2D drawings, and hence more efficient to explore SSMs than RLMs in our case. Indeed, in a study by
The teaching material (SSM) that we used in this study was designed in collaboration with a special education teacher. This type of device is relatively easy to design, adapt, and make. In fact, it was mentioned as an initial design recommendation because the teacher wants to be able to produce similar teaching material on his own in the future. First, using the same prototype, it is easy to modify the interactive description associated to each conductive zone. Indeed, it is enough to change the sound files on the micro SD card plugged into the Touchboard. Then, the teacher is free to modify the descriptions according to the perceptual and cognitive abilities of the students, as well as the pedagogical objectives of each lesson. More generally, any professional can design similar interactive teaching materials that fit to her/his own needs (see
Comparing costs for producing these different tools is another important aspect to consider. Teachers probably need 10 to 20 sheets of swell paper for preparing each lesson, depending on the number of failures during the printing process, necessary updates, as well as the number of students. Each pack of 100 sheets costs between 175€ and 330€ depending on the format (A4, A3). In addition to the paper, an inkjet or laser printer as well as a fuser (around 1300€ for ZY-Fuse or PIAF fusers) are needed. The initial cost for producing raised-line graphics is thus important. Obviously, for special education centers that already own a printer and a fuser, the cost is less important. However, although the production of RLMs might seem easy, it requires know-how regarding the design of the map (choice of element sizes, distances, textures, etc.) as well as the handling of the fuser. Then, making a tactile RLM is not an easy task, and is, in general fulfilled by a specialized person called “tactile document maker.” Obviously, 3D printing also relies on dedicated material (3D printer) and know-how. Currently, 3D printers are available at low costs in many cities around the world (thanks to the FabLab movement). But, more importantly, many centers which are specialized in visual impairment have already bought 3D printers and know how to design and print models (this was the case in the school we collaborated with). It is also important to compare the cost and usability of the prototypes designed in this study to the cost and usability of education tools in the market. Very few adapted tools exist and, hence, other tools (e.g., GeoSafari Talking Globe) are presented as educational tools for the blind. In fact this globe has been designed for sighted children and requires the presence of a sighted person to be used. Consequently, it is not adaptable to different usages, children and contents. Its current cost is around 130€.
We showed that a home-made affordable interactive SSM improves both space and text memorization in visually impaired students. Because such teaching materials are easy to make and affordable, they may cover many different use cases for visually impaired students. Through these technologies, professionals are able to design and build their own teaching material according to their needs (teaching of orientation and mobility, history, geography, mathematics, science, etc.). We are convinced that such accessible, adaptable and low-cost technologies will be accepted and used by special education centers in the future, which will improve education access for visually impaired people.
All authors made substantial contributions to the conception of the work and interpretation of data. They all participated in the drafting of the work and approved the version to be published. SG and AB carried out data acquisition, and SG carried out data analysis.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We thank INJA, Paris, and more specifically Julien Berthier, special education teacher, as well as the 24 visually impaired students who volunteered to participate in the study. We thank Andrea Boisadan for assistance during the experimentation. Finally, we thank CESDV-IJA and “Cherchons Pour Voir”, and more specifically Nathalie Bedouin for precious assistance in producing tactile documents.
In short, the kingdom first appeared under the reign of king Hector, who died from a poisoned arrow received in the shoulder. After the death of king Hector, the kingdom was enlarged by prince Angelot. Later on, a large rampart was added by the Leader of Sprites. The kingdom was then transformed by the Great Wizards who built the fortifications. The last ramparts were built by the Army of Gnomes around the whole kingdom, which were destroyed during the reunification of all the people. Finally, the kingdom was populated by elves, fairies, gnomes, sprites, humans, wizards, and witches. In total, one river, six roads, five ramparts, and thirteen landmarks (e.g., the forest museum, the gnome tower, etc.) were represented.