The Effects of 3D Printing on Social Interactions in Inclusive Classrooms

ABSTRACT

This research attempts to examine the role of 3D printer as a learning tool in inclusive education. It examines the quality of verbal and nonverbal interactions between students with disabilities (SWD) and students without disabilities (SWOD) in an inclusive classroom at primary schools, where 3D printers were used as a learning tool. The results presented in this paper are based on qualitative and quantitative analysis of interactions between 7 SWD and 31 SWOD. Even though both occurred, there was no statistical difference in generation of positive and negative verbal and nonverbal interactions between SWD and SWOD. Both, SWD and SWOD are of the opinion that cooperative learning with implementation of 3D printers contributes to their communication with peers, achieving better and durable knowledge in the subject where this learning approach was applied and they would like to learn in this way in the future.

KEYWORDS:

• cooperative learning
• 3D printers
• inclusive education
• interactions

1. Introduction

Application of different digital technologies in education is widely considered to be the most important step in modernisation and improvement of educational process. Digital educational technologies contribute to the quality of education for SWOD as well as for SWD. Yet, the process of gradual modernisation of classrooms has often left them insufficiently equipped. In general, most of the schools have more students than computers which consequently results in students’ group work while using one computer (Nikolić, Petković, Denić, Milovančević, & Gavrilović, 2018). This might not be an impediment in education seeing that computer supported group learning, or dyadic cooperative learning, may contribute to better students’ results in school (Blaye, Light, Joiner, & Sheldon, 1991). When cooperative learning is considered in an inclusive environment, there are different and opposing findings in research. On the one hand, research and literature show evidence that the SWD are vulnerable to social and academic exclusion and rejection in schools (Flem, Moen, & Gudmundsdottir, 2004; Nowicki, Brown, & Dare, 2017; United Nations, 2006). On the other hand, research by Ouherrou, Elhammoumi, Benmarrakchi, and El Kafi (2019), Creighton & Szymkowiak (2014) and Xin (1999) indicate that the environment in inclusive education is especially suitable for peer computer-supported cooperative learning (CSCL) between SWD and SWOD. CSCL today is reinforced with new emerging technologies such as 3D printing. The process of 3D printing or additive manufacturing is a process for transferring digital files (computer-created and designed models) into three-dimensional solid objects (physical three-dimensional models). The process of producing a 3D printed model could be divided into two main parts: 3D modelling and 3D printing. 3D modelling is the design and development of 3D models in the digital environment, with the implementation of computers and some of many software programs for 3D modelling and printing. 3D printing turns the digital 3D object into the physical one by using a 3D printer and materials for printing (most often for educational purposes plastic strings are used). Both of these processes were united under the name 3D printing in previous educational research (Bosse & Pelka, 2020; Buehler, Easley, McDonald, Comrie, & Hurst, 2015; Kostakis, Niaros, & Giotitsas, 2015), and this term is also used in this research. Experimental research conducted by Kostakis, Niaros, and Giotitsas (2015), indicates that 3D printing could serve as a learning tool for cooperative learning and communication between SWD and SWOD. They concluded that 3D printing has a rich potential for implementation in inclusive education environment, but before definitive recommendation is given more research ought to be done to acknowledge advantages and disadvantages of its implementation.

This research seeks to contribute to the knowledge in this area. We present the results of a one-year research project on implementation of 3D printing in an inclusive classroom. The project was funded by the European Union and the Council of Europe and actualised in Montenegro. The paper is focused on the identification and exploration of nonverbal and verbal socio-affective aspects of collaboration and collaborative learning between SWD and SWOD during their group interactions with 3D printers in schools and production of 3D models.

2. Background and Literature Review

Acceptance, understanding and good relations between students with different needs are core values of inclusive education. Cooperative learning, as a teaching method, supports interaction between the students with different abilities. This teaching approach connects social and academic experiences within the learning environment. Both SWOD and SWD are taught to complete tasks collectively to acquire academic and social skills (Johnson & Johnson, 1986; Slavin, 1996). Implementation of cooperative learning in inclusive education could be one of the possible approaches for increasing opportunities and improvement of learning and social outcomes of SWD (O’Connor & Jenkins, 1996). Dugan et al. (1995) implemented cooperative learning as an approach in fourth-grade social studies classes where students with autism and SWOD participated. The evidence from this research suggests that cooperative learning contributes to both social and academic engagement. Additionally, SWD and SWOD achieved better results on the post-test. The positive impact of cooperative learning on social engagement in inclusive education environment is confirmed by Grey, Bruton, Honan, McGuinness, and Daly (2007). Grey, Bruton, Honan, McGuinness, and Daly (2007) examined the contribution of cooperative learning. This approach was applied in mathematics and English language classes to examine social and task engagement of students who are on the spectrum. Results of the study indicated that cooperative learning influenced the level of social engagement positively for both SWD and SWOD, but no such influence was found in task engagement.

Implementation of digital technologies in education has enacted changes in many teaching approaches and methods. Computer-assisted cooperative learning provides more opportunities to accommodate the different needs of students in group work, than it is the case with the same approach without implementation of technologies (AbuSeileek, 2012). The experimental study of Xin (1999), analyzes the contribution of computer-assisted cooperative learning from the achievements of SWD and SWOD in mathematics. The results of his research indicate that both SWD and SWOD achieved better results when they were taught with the implementation of computer-assisted cooperative learning.

A 3D printer is used in educational research and practice as a potential tool for cooperative learning in an inclusive classroom. 3D printers are suitable for producing individualised learning aids which fulfil different needs of students, as well as enable the students to learn together (Bosse & Pelka, 2020; Buehler, Easley, McDonald, Comrie, & Hurst, 2015). This is supported by the results of several experimental studies. For instance, Kostakis, Niaros, and Giotitsas (2015) examined the potentials of using 3D printers as learning and communication tools in an inclusive educational environment. In their experimental studies, 3D printers were used by SWOD for creation and production of the artefacts for visually impaired students. Based on the results of the study, Kostakis, Niaros, and Giotitsas (2015) concluded that 3D printers could be used as a tool for cooperative learning and communication of SWOD and visually impaired students. A similar research was conducted by Stangl, Jernigan, and Yeh (2015), in which the potential of 3D printing as a learning tool for cooperative learning between visually impaired students and SWOD was examined. The researchers drew attention to the fact that 3D printing contributed to multimodal learning and communication skills of both student groups, but only when the students were provided with clear instructions. They indicated that 3D printing can be used for constructionist learning, which fosters communication in an inclusive classroom. Buehler, Comrie, Hofmann, McDonald, and Hurst (2016) pointed out that 3D printing is a powerful learning tool which could be used in backing the science, technology, engineering, arts, and mathematics (STEAM) cooperative education of students with different needs. It could also be used to produce the learning aids for SWD. Students with intellectual disabilities were included in their experimental study. The obtained results indicate that 3D printing scaffolds cooperative learning not only between SWOD and students with intellectual disabilities, but also between SWD by building relations and promoting engagement. This is in line with experimental research of Jacobs et al. (2016) where 3D assistive teaching aids were produced by 3D printing through cooperative learning between SWOD and students with motor impairments. They concluded that this learning approach contributes to the STEAM knowledge of both groups of students. A recent literature review on the topic of implementation of 3D printing in inclusive education environment has found that this technology could be used for education of students with visual, motor and cognitive impairments and their cooperative learning with SWOD (for review see Ford & Minshall, 2019). As it can be deduced from this literature review, 3D printing is a promising tool for cooperative learning between SWD and SWOD in inclusive education. However, no research has previously examined verbal and nonverbal interaction between SWD and SWORD in 3D printer-supported cooperative learning. This research seeks to contribute to the knowledge in this area and tries to answer the following research questions:

1. Which verbal interactions (positive and/or negative) between primary school SWD and SWOD promote the socio-affective aspect of cooperative learning when they learn together with the implementation of 3D printers?

2. Which nonverbal interactions (positive and/or negative) between primary school SWD and SWOD promote the socio-affective aspect of cooperative learning when they learn together with the implementation of 3D printers?

3. What are opinions of SWD and SWOD on cooperative learning with the implementation of 3D printers?

3. Research Methodology

3.1. Setting and Context of the Study

This study is descriptive and exploratory in nature and was conducted at five primary schools in Montenegro. Primary education in Montenegro includes students of ages 6 to 15. All primary schools are inclusive in accordance with the Montenegrin Law on Education. Each primary school in Montenegro employs a Special Needs Education Specialist to support teachers in the implementation of inclusive education. In addition, Montenegrin Ministry of Education offers workshops for teachers in different areas each year, which include the topics from inclusive education.

Our research commenced with a public call for a workshop set. The workshop entitled ‘3D Printing as a Cooperative Learning Tool in Inclusive Education’ was launched by the department of Ministry of Education of Montenegro. It was noted that every selected school would be equipped with a 3D printer and materials for printing. Ten schools were selected based on the following criteria: total number of students, number of students with disabilities, and number of teachers who were interested to participate in the workshops. The workshop instructors team consisted of an international expert in educational technologies − 3D printing, an international expert in inclusive education, and an expert from Montenegro, well-grounded in education and educational policies in the state.

During the workshops, the teachers were taught how to implement 3D printers in cooperative learning between SWD and SWOD. Tinkercad free online 3D modelling program was used for teachers’ workshops and it was recommended to be used in the classroom. This modelling program was selected based on the recommendation of a previous research on using 3D modelling and printing by students, teachers and beginners in this field (Micallef, 2015; Trust & Maloy, 2017). During the workshop section titled 3D printing as cooperative learning between SWD and SWOD, the teachers were presented with the examples from international practices, as well with the relevant research in this area such as Buehler, Easley, McDonald, Comrie, and Hurst (2015), Kostakis, Niaros, and Giotitsas (2015), Jacobs et al. (2016). The teachers were also presented with a simpler program for developing 3D models, such as CookieCutters, as well as possibilities for converting hand-drawn sketches and student drawings into digital form for 3D printing. On the other hand, teachers were briefly familiarised with more demanding software for developing 3D models, such as Blender and FreeCAD. It was suggested to the teachers that the selection of software for the development of 3D models should depend on the digital skills of the teachers and students and it should start with a simple one and gradually move in direction of more sophisticated software which requires higher digital skills. A special part of the workshop was dedicated to instructing the teachers in the ways in which students with various disabilities can be involved in working with 3D printers, which was based on previously published research (e.g. Bosse & Pelka, 2020; Buehler, Comrie, Hofmann, McDonald, & Hurst, 2016; Stangl, Jernigan, & Yeh, 2015; Wang, Laffey, Xing, Galyen, & Stichter, 2017)

At the end of the workshops, the teachers were invited to voluntarily participate in collecting the data from their classrooms, after they implement 3D printing in cooperative learning between SWD and SWOD. The teachers from five schools were willing to collect the data and provide the information to the researchers.

3.2 Data Collection

The data in this study were collected from both the volunteering teachers and their students. The researchers did not participate in the organisation or realisation of classes, all this process was done by the teachers. This means the teachers were totally free to select a suitable topic while teaching 3D printers in cooperative learning, make student groups, create instructions for the students, and realise the classes. This approach was used in order to provide an insight into the real cooperative learning between SWD and SWOD. Therefore, the material could be critically analysed and suggestions for practice and future research provided. A controlled experimental research cannot provide full insight into how teaching practice really works, or how teachers and students feel about an educational approach (Cook, Tankersley, Cook, & Landrum, 2008). Both qualitative and quantitative types of data were collected, which is recommended for research in inclusive education (Artiles, Kozleski, Dorn, & Christensen, 2006; Cook, Tankersley, & Landrum, 2009).

Qualitative data were collected by audio and video taping the classes in which 3D printers were implemented in cooperative learning between SWD and SWOD. All the classes lasted for 45 minutes. When the implementation of 3D printers in cooperative learning was done, and teachers considered it as complete for that teaching topic, a semi structured questionnaire was shared with SWD and SWOD in aim to extract their opinions and experience. The questionnaire contained twelve items organised into three sets. Each questionnaire set contained four questions. The first set of questions contained questions about communication between peers in the group during the class. The second set of questions examined students’ opinions on contributions of cooperative learning with 3D printing to their knowledge in the teaching subject where it was implemented. The aim of the third set of questions was to examine students’ willingness to use the technology, as well as to provide suggestions on the implementation of cooperative learning with 3D printing in the future. The first and second set of questions were close-ended and a five-point Likert scale was implemented. The third set of questions contained three close-ended and one open-ended question. The questionnaire was pilot tested with five SWOD and two SWD, who provided feedback about the clarity of questions and ease of answering. After this part was conducted, the questionnaire was distributed to all the participants in each school. With SWD who had the problem to fill questionnaire by themselves, the answers were collected by asking the question orally and then the students’ answers were recorded. In the development of the questionnaire, similar previous research that examined the application of digital technologies in inclusive education was followed Andjic et al., (2019); Turan and Atila (2021); Rodriguez-Ascaso, Letón, Muñoz-Carenas, Finat, and Papa (2018). The validity of the questionnaire was checked by three teachers who teach in inclusive classes, one expert in the field of inclusive education and one expert in the field of educational digital technologies. Based on their opinion, it was confirmed that the questions in the questionnaire are clear and understandable, easy to answer, that they provide answers that are in agreement with the research questions, and do not violate the privacy of the participants in the research. These answers were later transferred into a questionnaire sheet. The same questionnaire was used with all the participants.

3.3. Participants

Five teachers, from five different primary schools, who agreed to collect the data, implemented 3D printers in the inclusive classroom as a learning tool for cooperative learning between SWD and SWOD. In total 31 SWOD and 7 SWD participated in the study. All the students who participated in this research were in the same class before and after the research, so the students knew each other and had previous experience in group work. Considering that the goal of this research is to examine the effects of 3D printing on social interactions in inclusive classrooms in the real learning environment (without the influence of experimental factors), implementation of 3D printing was the only novelty, all other factors, such as the division into groups and assigning the task to them, were in accordance with the previous experiences and habits of the teachers and students who participated in the discussion. The anonymity and confidentiality were guaranteed to all the participants. Detailed information about school grade, teaching topic, lesson number on which students worked cooperatively with the implementation of 3D printers are provided in Table 1.

Table 1. Specific information about the participants from each school and teaching topic.

School number:	School grade:	Teaching subject:	Teaching topic:	Number of recorded lessons:	Number of SWD:	Type of disability for SWD:	Number of SWOD:	The task given 3D printer task:
1	second	Native language	Writing	5	1	physical disability- motor control	5	Create a model for letter writing.
2	second	Native language	Writing	5	2	one student with vision impairment and one with physical disabilities	5	Create the Braille alphabet model.
3	seventh	Biology	Cell	9	1	vision impairment	6	Create the model of a plant cell.
4	ninth	Mathematics	Geometry	7	1	vision impairment	5	Create the model of polyhedron.
5.1	ninth	Geography	Regions of Montenegro	8	1	autism spectrum disorder	5	Create the model of our country.
5.2	eight	Chemistry	The structure of Benzene	7	1	autism spectrum disorder	5	Create the model of structural formula of benzene.

3.4. Data Analysis

Qualitative and quantitative analyses were undertaken in data processing. Qualitative analysis was used in order to understand the interaction between SWD and SWOD. It included the processing of video and audio data obtained from the class recordings. The recorded material was first observed by the researchers several times in order to get a better insight into it. The material was open-coded with the application of grounded theory approach (Glaser & Strauss, 1967; Strauss & Corbin, 1990). Each group’s verbal conversation was transcribed from the video recording. The transcribed material was supplemented with relevant nonverbal actions (body movements, face expression, engagement in various physical activities) and time stamps at which the actions between SWD and SWOD occurred. The coding process was realised manually. All codes in this research were obtained in an open coding process and were generated exclusively from the SWD and SWOD narratives or body movements. The codes and their names were constructed from the transcribed materials. The categories and subcategories were developed based on constant comparative method (Strauss & Corbin, 1990). The transcribed material was coded individually by each member of the research team. The obtained codes were distributed to the teachers who worked with students (teachers whose classes were recorded), as well as to two experts in qualitative research. They were asked to arrange the codes into subcategories and categories. The aim of this process was to ensure validity of the obtained data (Krippendorff, 2013; Miles & Huberman, 1994). The reliability of obtained data was calculated with application of Miles and Huberman (1994) formula, which indicates the correspondence in coding and code division into subcategories and categories. A similar approach for coding is used and recommended by one of the previous researches in inclusive and special education (Anđić, Cvjetićanin, Maričić, & Stešević, 2019; Anđić, Šorgo, Cvjetićanin, Maričić, & Stešević, 2021; Wang, Xing, & Laffey, 2018). In order to analyse the potential statistically significant difference between SWD and SWOD concerning the occurrence of verbal and nonverbal interaction, the codes were transferred into nominal „dummy variable ‘, (1 = interaction has occurred; 0 = interaction has not occurred). These nominal variables were statistically processed with the Mann Whitney U test. Code subcategories and categories presented in this paper are those which were produced in the interaction of the two groups of students.

The data obtained from the questionnaire were quantitatively and qualitatively processed. Descriptive statistics, median, mode and the standard deviation were used to explore general opinions of all participants on cooperative learning with implementation of 3D modelling and printing. The Mann-Whitney U test was used for analysing differences in opinions among SWD and SWOD on the questionnaire. Whether the survey data set is suitable for factor analysis was calculated with Barlett sphericity test and the Kaiser-Meyer-Olkin test – KMO. The Cronbach Alpha coefficient (α) tested the internal consistency of the factors in the questionnaire. The Cronbach Alpha coefficient (α) for all factors in the questionnaire was greater than .82 which indicates internal consistency i.e. reliability.

4. Results

A total of 95 verbal and 18 nonverbal codes were obtained in the coding process. The correspondence in coding and arranging codes in subcategories and categories was calculated using Miles and Huberman. Consensus between the researchers was of 93%. Consensus between the researchers on one side and teachers and experts in qualitative research on the other side was 87%. Based on these results, the obtained data was considered reliable. In order to achieve distinctness, the presentation of the results was divided into the following subsections: verbal interaction between SWD and SWOD, nonverbal interaction between SWD and SWOD, opinions of SWD and SWOD on cooperative learning with the implementation of 3D printers.

4.1. Verbal Interaction Between SWD and SWOD

SWOD provided the researchers with 95 verbal codes, while SWD provided 90 (Table 2). All verbal codes were classified into 16 subcategories and 2 categories were characterised as positive verbal interactions and negative verbal interactions. The most frequent verbal code I can (f = 376), was classified into the subcategory Answer, and Positive verbal interaction category. The least frequent verbal code we are not interested was classified into Impolite interrupting subcategory and Negative verbal interaction category. Five codes: not with your ideas, bad language, not now, we are not interested and stop talking which were all classified into the category Negative verbal interaction were obtained only from SWOD.

Table 2. Codes, subcategories and categories in verbal interaction between SWD and SWOD.

Category	Subcategory	Code	Total code frequency	Code frequency with SWD	Code frequency with SWOD
Positive verbal interaction	Question	Can we start	167	43	124
		Can you help	286	61	225
		Can I help	143	27	116
		Did you understand	312	31	281
		How do you do	254	43	211
		Should we	138	19	119
		Can you explain	179	37	142
		Can you repeat	251	57	194
		Can you do	138	17	117
		Do you agree	312	37	275
		Is it clear	254	11	243
		What is	324	52	272
		Shall I explain	126	14	112
		Can you do it by yourself	294	9	285
		Are you listening	159	28	131
		Are you done	182	39	143
	Answer	I understood	271	51	220
		I did not understand	98	27	71
		I can	376	54	322
		I can’t	83	26	57
		I know that	153	45	108
		I’m confused	75	21	54
		I think	289	67	222
		Why don’t we	256	15	239
		I don’t know	154	31	123
		I got it	204	43	161
	Suggestion	We should	365	59	306
		I suggest	329	43	286
		You could	186	20	166
		Have you thought about	65	7	58
		Let me see	261	63	198
	Informing	I see	176	37	139
		That’s strange	54	8	46
		It’s confusing	139	33	106
		It doesn’t work	249	14	235
		I’ll take notes	357	22	335
		I’m next	286	28	258
		My theory is	73	9	64
		The model is	189	55	134
	Accepting - Agreement	You are right	338	47	291
		Ok	364	61	303
		I like it	265	46	219
		That’s good	219	51	168
		Smart idea	59	13	46
		Good job	153	5	148
		Almost	43	8	35
		I agree	285	75	210
		The same	61	17	44
		Like my opinion	45	9	36
		Yes, that is	128	43	85
	Disagreement	I don’t think so	212	51	161
		You are wrong	97	16	81
		It’s incorrect	126	34	92
		I have other opinion	219	13	206
		That is contrary to	53	7	46
		I disagree	175	21	154
	Instruct	Read please	265	17	248
		Draw it on	269	52	217
		Rotate the model	341	87	254
		Take this part of	275	39	236
		Send a file	189	21	168
		Take a	167	45	122
		Summarize	286	17	169
		Look carefully	214	36	178
		Compare	136	19	117
		Touch the model	289	11	278
		Try the model	219	9	210
		Describe	241	189	52
	Evaluate	You do it well	157	39	188
		You can correct it	135	12	123
		It’s incorrect	192	29	163
		You made a mistake	276	51	225
	Alert	Be careful	275	41	234
		It’s hot	153	34	119
		Don’t touch	104	19	85
	Encouragement	You can do it	286	37	249
		I believe you can	189	17	172
		You are great, proceed	287	11	276
	Compliment	You are smart	178	32	146
		You are great	78	34	44
		You are amazing	32	9	23
	Invite	Come here to see	284	32	252
Negative verbal interaction	Reject	Not with your ideas	45	0	45
		We definitely won’t	51	3	48
	Mocking	Bad language	12	0	12
		Nyah nyah nyah nyah nyah (Teasing)	23	7	15
		Provocative laughing	9	2	7
	Impolite interrupting	Not now	49	0	49
		Later	27	2	35
		We are not interested	8	0	8
		You speak too long	21	1	20
		Stop talking	17	0	17
	Blame	It was your idea	52	16	36
		I told you	37	13	24
		It’s your fault	49	14	35

Statistical processing of codes as a dummy variable by Mann Whitney U test (U = 1203.000; p = .309) demonstrates that there is no significant statistical difference between SWD and SWOD in the number of positive verbal interaction occurrences. Comparatively, the result of the same test indicates that there is a statistically significant difference between SWD and SWOD in number of occurrences of negative verbal interaction (U = 159.000; p < .001). The effect size for this data is calculated by implementation of Rosenthal formula (1991). It shows medium effect (r = 0.50).

4.2. Nonverbal Interaction Between SWD and SWOD

SWOD provided the researchers with 18 nonverbal interaction codes, while SWD provided one code less (Table 3.) All verbal codes were classified into six subcategories and two categories - Positive nonverbal interaction and Negative nonverbal interaction. The most frequent nonverbal code fingers directing (f = 491), was classified into Positive gestures subcategory and Positive nonverbal interaction category. The least frequent nonverbal code back turning was classified into Negative gestures subcategory and Negative nonverbal interaction category. The code eye roll classified into Negative facial expression subcategory and Negative nonverbal interaction category was coded only in nonverbal interaction of SWOD.

Table 3. Codes, subcategories and categories in nonverbal interaction between SWD and SWOD.

Category	Subcategory	Code	Total code frequency	Code frequency with SWD	Code frequency with SWOD
Positive nonverbal interaction	Positive facial expression	Smiling	397	124	273
		Winking	23	4	19
		Uncertain/In thoughts	354	89	256
	Positive gestures	Hand alert	65	7	58
		Fist pump	269	45	224
		Fingers directing	491	153	138
		Hand shapes showing	534	219	315
	Touch	Hands touching	164	40	124
		Touching the shoulders	56	12	44
		Handshaking	41	4	37
		Clap their hands together	143	22	21
	Head movements	Nod	317	77	240
		‘no’ head movement	130	51	79
Negative nonverbal interaction	Negative facial expression	Eye roll	59	0	59
		Frown	17	3	14
	Negative gestures	Angry hand movements	32	10	22
		Back turning	7	2	5
		Withdrawal	38	23	15

The results of Mann Whitney U test (U = 1194.000; p = .341) indicate that there is no significant statistical difference between SWD and SWOD in the number of codes which represent positive nonverbal interaction occurrences. However, the result of the same test shows statistically significant difference between SWD and SWOD in the number of negative nonverbal interaction codes. The results obtained with the implementation of Rosenthal formula (1991), indicate that this effect size is small (r = 0.20).

4.3. Opinions of SWD and SWOD on Cooperative Learning with the Implementation of 3D Printers

The principal component analysis (Barlett sphericity test = 319.113; KMO = 680; df = 59; p = .000), demonstrated that there were four factors that explain 67.19% of total variance. For further analysis, three specific factors were taken: Factor 1: Students’ opinions on communication between peers, which explain 23.39% of total variance; Factor 2: Students’ opinions on contributions of cooperative learning with 3D printing to their knowledge in school subject where it was implemented, which explain 20.96% of total variance; Factor 3: Students’ opinions on using cooperative learning with implementation of 3D printing in the future, which explain 22.84% of total variance (Table 4).

Table 4. Range, average values, dispersion of results and Cronbach Alpha coefficient for factors.

Factors	N	M	SD	α
Factors 1	39	4.19	.71	.82
Factors 2	39	3.85	.52	.89
Factors 3	39	4.38	.68	.86

Most of the SWD and SWOD, selected the answer Strongly agree when related to the claims that cooperative learning with implementation of 3D printers helps them to communicate more, gives them the chance to express their opinions and to be heard, as well as to hear different opinions, Figure 1. The results of the Mann Whitney U test (U = 1213.000; p = .409) indicate that there was no statistically significant difference between SWD and SWOD in opinions on the contribution of cooperative learning with implementation of 3D to their communication.

Figure 1. The difference in opinions on the contribution of cooperative learning with the implementation of 3D printing to their communications.

Most of the SWD selected Strongly agree, while most of the SWOD selected Agree in relation to the claims that cooperative learning with implementation of 3D printers helps them to gain better and lasting knowledge. The difference in the opinions of SWD and SWOD, was also confirmed by the Mann-Whitney U test (U = 1203.000; p < 0.05). However, both groups of students selected Strongly agree in relation to the claims that cooperative learning with the implementation of 3D printers helps them to better understand learning materials and implement their knowledge, Figure 2.

Figure 2. The difference in opinions on the contribution of cooperative learning with the implementation of 3D printing to their knowledge.

Most of the SWD and SWOD selected Strongly agree that they would like to use cooperative learning with 3D printing in the future, would recommend it to other peers, that it should be implemented in schools, Figure 3. The results of the Mann Whitney U test (U = 1151.000; p = .317) indicate that there was no statistically significant difference between SWD and SWOD in opinions on the future use of cooperative learning with 3D printing in education.

Figure 3. The difference in opinions on the future use of cooperative learning with the implementation of 3D printing.

When SWD and SWOD were asked: ‘What should be done to improve cooperative learning with implementation of 3D printers?’ they replied: ‘Allow more time to work with 3D printers’, ‘Provide more 3D printers, ‘Try to make the printers more accurate and faster’, ‘Enable multi-colour filament printing’, ‘It would be useful if programs for model creation could be simplified’.

5. Discussion

The main goals of this study was to determine the verbal and nonverbal interactions promoting the socio-affective aspect of cooperative learning between SWD and SWOD while using 3D printers and to examine their opinions about this learning approach. Our data suggests that 3D printers in an inclusive classroom provide mostly positive verbal and nonverbal interaction between students. In our research, both SWOD and SWD had a positive opinion about 3D printer-supported collaborative learning. By placing the focus on knowing the specifics which 3D printers bring to verbal and non-verbal interactions in the inclusive classroom, our paper contributes to the knowledge about SWD and SWOD collaborative learning. In the continuation, we first discuss the verbal interactions between SWD and SWOD, which arose during the 3D printer-supported collaborative learning, followed by the discussion on nonverbal interactions and discussion about students’ opinions and improvements suggestions for the implementation of 3D printing in education.

In our study, 3D printer-supported collaborative learning was a fertile ground for mostly positive verbal interaction between SWD and SWOD. In the positive verbal interaction category, those verbal acts which are classified under the Question subcategory are the most represented and have the highest frequency in SWD and SWOD. Most of the questions which students asked each other occurred while they worked cooperatively on the task. Students asked each other for an opinion on ideas, asked questions to check whether the members of their group understood their suggestions, asked or offered help in task realisation. In addition, there were questions which were not related to the task such as: ‘We are a super/the best team. Do you agree?’. Our research reveals interesting occurrences, SWD were more willing to ask a question when they interacted with 3D printed models than while modelling. For example, after exploring the 3D printed model, a student with an autism spectrum disorder (from 5.1 school- Table 1) asked her peers about the colour of the model and about the position about the capital city where she lives. However, the same student did not raise this question during the 3D modelling process. We offer two possible explanations for this result. The first physical 3D model provides a multisensory experience (by seeing, touching, moving) to the students with an autism spectrum disorder which probably draws their attention and keeps them interested. The results of Simpson and Taliaferro (2021), and Zhang and Griffin (2007) research support this assumption. Those researchers indicated that using printed 3D models could appeal to the students with an autism spectrum disorder, providing sensory interest and increasing their active participation in teaching. Secondly, the explanation could be that there is a lack of the experience of students with an autism spectrum disorder to interact with the 3D digital environment which occurs during 3D modelling. This contextual experience of the students with 3D digital environments may be useful for developing hypotheses for further research. All of the students with vision impairments who participated in our research were more willing to ask a question when they had the chance for exploring the models by touching than in the process of modelling, even though their peers explained to them the modelling process. This is in line with previous research (Anđić, Cvjetićanin, Maričić, & Stešević, 2019; Anđić, Šorgo, Cvjetićanin, Maričić, & Stešević, 2021), in which the students with visual impairments felt more confident when they perceived teaching content based on their sensory experience, which boosted their self-confidence when communicating with their peers, in comparison to the moments when they received oral explanations. Verbal acts which occurred between SWD and SWOD which are classified into the category Questions in our research were mostly followed with verbal acts from the Answer, Suggestion, Informing, Agreement, and Disagreement category. SWD who participated in our study were less confident than their peers without disabilities in undertaking verbal acts classified into the above-mentioned categories. The most frequent verbal act of SWOD from the Answers category was I can, which they used to express that they could do something requested by their peers (read, model, create, repeat, etc.). However, when SWD were in the same situation, they usually replied with verbal acts under the code I think (I can-I can’t). Our study provides additional support for the results of Johnston, McDonnell, Nelson, and Magnavito (2003) and Williams (2011) who claim that SWD could be less confident in using new learning and communication tools in an inclusive classroom than SWOD. The most surprising finding is that the younger SWD in our research were more willing to provide suggestions to their peers than the older SWD. Although the SWD from the second, third, and fourth schools (Table 1) had the same disability – visually-impaired, the students from the first school provided much more suggestions to SWOD, than students from the two other schools. It is very likely that teachers’ task from the second school „Create Braille Alphabet model” was appealing, well designed, and adapted to fit the needs of both SWD and SWOD, and this enabled more communication between the students. The teachers from the third and fourth schools in the tasks did not specify that the student-created model should contain some Braille symbols, or some adaption for visually impaired students which in turn possibly reduced the suggestions from SWD during the 3D model design. Our assumptions are supported by the conclusions of Ramirez and Gordy (2020) and Simpson and Taliaferro (2021). They concluded that adaptation of 3D modelling and printing tasks to SWD characteristics contributes to their active participation in the process of modelling, learning, and communication with the peers. In our research, some verbal interactions which occurred between the SWD and SWOD are classified into subcategories Informing, Agreement, and Disagreement. According to Eggins and Slade (1997) and Mavrou (2012), this type of verbal interaction leads to a structural-functional conversation and increases the integration of SWD in inclusive classrooms. SWD who participated in our research, generated more verbal acts within the subcategories Informing, Agreement, and Disagreement, insofar as they had previous experience with the task they needed to model, than SWD without the experience in modelling topic. Thus, for example, SWD from school 5.1, (Table 1) addressed more verbal acts to their peers than students from school 4 who were the same age and school 5.2 who had the same type of disorder. In generating these verbal acts, the student from 5.1 school used the previous knowledge of the geography of his country, but also his life experience and informed his friends where to position certain parts of the model. He further approved or disagreed with the suggestions of his friends. In our research, positive verbal interactions classified into subcategory Instruct contain verbal acts which developed between SWD and SWOD in relation to directing each other on how to perform a task. The most frequent instructional verbal act „Touch the model” occurred when SWOD directed SWD, whereby it created an opportunity for SWD to sensory perceive the physical form of the 3D printed model. Conversely, the most frequent instructive verbal act used by SWD, when addressing SWOD was „Describe”, and it was used in the process of digital modelling. With this verbal act, SWD usually instructed SWOD to describe the part of the model which they modelled or which was not visible because of the digital 3D perspective during the modelling process. „Describe” as a verbal act was not only the characteristic of students with visual impairments but of all SWD and some SWOD as well. To give an example in the first school (Table 1), when a SWD rotated the digital model in the modelling process, one part became invisible because the angle changed. He then asked the peer „Describe where the letters A, B, C disappeared” where the instruction was supported by some of SWOD. It is interesting that SWOD reacted to this instruction by providing the instructional direction „Rotate the model”, which changed the perspective and made the whole model visible again. The process of 3D modelling supported in this way highly supported the socio-affective aspect of cooperative learning between SWD and SWOD. The results of another similar research indicate that this type of interaction could be beneficial for both SWD and SWOD (Bond & Castagnera, 2006; Mavrou, 2012). They concluded that the students who receive instruction and support in learning from their peers in an inclusive classroom, and those who provide instructions achieve better academic and social scores. Verbal acts classified into subcategories Evaluate, Alert, Encouragement, Compliment, Invite were also present in the communication of both groups of students. In our research, both SWD and SWOD provided an evaluation-feedback to each other in the process of modelling and on the final 3D printed model. Remarkably, SWD in our research were more engaged in the evaluation of 3D printed models produced by themselves and their peers, than when the models were digitally processed. By way of illustration, SWD from the second school pointed out mistakes in the height of Braille points to one of the SWOD who insisted that they should be increased. After examination of the 3D printed model, the SWD from the fifth school provided positive feedback to their peers. The SWOD provided evaluation – feedback, and encouraging comments to the SWD during the modelling as well while handling the 3D printed model. These findings are significantly different from previous results reported by Mavrou (2012), in which verbal acts, such as Evaluation, were an exclusive act of the SWOD in computer-supported cooperative learning. Implementation of 3D printing as a learning tool in our research could be the reason for this difference. It is very likely that 3D printing supported cooperative learning placing SWD and SWOD in role creators and product producers, not only as consumers and recipients, and this triggered verbal interaction involving evaluation of the work and product. Besides our results, this assumption is grounded on the results of Dukuzumuremyi and Siklander (2018) research which concluded that the capacity of digital technologies for increasing communication between students in an inclusive classroom is significantly greater when they have the chance to create and produce educational content, information, and multimedia. This is in good agreement with the results of Buehler, Easley, McDonald, Comrie, and Hurst (2015), Kostakis, Niaros, and Giotitsas (2015), and Ulbrich et al. (2020), which conclude that the implementation of 3D modelling and printing in cooperative learning provides possibilities for interactions with computer screens and printed models, which can benefit students’ interactions. Our research indicates that the process of 3D printing in an inclusive classroom develops possibilities for verbal interactions classified into subcategories Alert. These verbal interactions were presented by SWD and SWOD, and they used it in the process of 3D printing and printer handling. Similar verbal interaction is not found in similar research which examined computer-supported cooperative learning (Dukuzumuremyi & Siklander, 2018; Mavrou, 2012). Together with, verbal acts classified into subcategories Encouragement and Compliment these interactions positively impacted relations between students and made them feel more comfortable in cooperative working, which was manifested by students smiling, thanking their peers, and similar. Siklander, Kangas, Ruhalahti, and Korva (2017) indicate that this type of interaction positively contributes to the development of a sense of belonging to the group and students’ wellbeing. These verbal interactions highly promote the socio-affective aspect of collaborative learning, acceptance, and active participation of all students in an inclusive classroom.

Results of our study indicate that between SWD and SWOD, negative verbal actions were also present. They include initial rejecting of peer ideas without reasoning, mocking, impolite interrupting, and blaming for poor results or similar. Negative verbal acts were mostly initiated by SWOD, while in SWD they appeared in reaction to the negative behaviour of their peers. Some of the negative verbal acts such as bad language, stop talking, we are not interested, not now, later, and not with your ideas were exclusively an act of SWOD. The results of our study show the occurrence of negative verbal interaction directed from SWOD to SWD in two particular cases: 1) when they were time-limited for task performing and 2) when their interaction with SWD required greater effort. By way of illustration for the first case, in the 5.2 school, a SWD tried to give suggestions before the model was sent for 3D printing, where one of SWOD interrupted him with „Not now with your ideas, I want to finish this and show it to the teacher, I want a plus”. As it could be seen in this particular case this student placed his own success before the success of the group. Similar negative verbal interaction between SWD and SWOD was presented in the earlier research by JJohnson and Johnson (1992), explaining that this type of negative verbal interaction occurs because the SWOD tend to individually benefit from the activity, they lack patience, or they have anxiety to complete the task. As one illustration of negative verbal interaction, we present a situation from the third school. When one of the SWD asked his peers to explain to him one part of the digital 3D model, one of the SWOD replied with „Not now”. However, another SWOD provided a polite and clear explanation to the SWD. Madden, Ellen, and Ajzen (1992) indicate the likelihood that some SWOD, even if they have a positive opinion of their classmates with disabilities, avoid communication with them if it requires extra effort. It is to be noted that these negative verbal acts in our research most often came from several SWOD and do not represent a characteristic of all SWOD. Rejection behaviour in peer interaction is related to personal characteristics of some students’ personalities and backgrounds (Hendry et al., 2005). However, Roberts and Smith (1999) and Thousand and Burchard (1990), indicate that teachers need to recognise forms of negative verbal interaction and contribute to their suppression in inclusive performance.

Li, Kidziński, Jermann, & Dillenbourg (2015), stressed that successful technology-supported collaborative learning required good understanding of both verbal and nonverbal interactions. The results of our research show that in 3D printer-supported collaborative learning between SWD and SWOD, positive and negative acts occurred in nonverbal communication. The implementation of 3D printers has enabled a wide variety of student nonverbal communication acts, which promote socio-affective aspects of cooperative learning. Our results are consistent with previous results (Wang, Laffey, Xing, Galyen, & Stichter, 2017; Williams, 2011; Xin, 1999) where suggested collaborative learning between SWD and SWOD supported with technologies (computers, laptops, virtual reality) caused following positive nonverbal interactions: smiling, finger-directing, hand-touching, handshaking. These researches indicate these types of nonverbal interactions contribute to students’ well-being and their feelings of belonging to the class and school community. The results of our study provide additional information about nonverbal interactions which promote the socio-affective collaborative learning between SWD and SWOD when supported by 3D printing and modelling, which are: hand shape showing, hands touching, hand alert. Positive nonverbal acts hand shape showing was the most represented nonverbal act and occurred in every part of the cooperative learning process. To illustrate, in the fourth school (Table 1), a SWOD tried to make different geometrical shapes with his own hands and hands of visually impaired students with the aim to explain the student the shape which they digitally modelled. The same approach was used by a visually impaired student when he was examining the physical 3D printed model. He took the hands (hands touching nonverbal act) of his peers and directed them on which part on the model to touch in aim to provide feedback to each other. Hand alerts as a nonverbal act occurred between students in the process of 3D printing, and they were used in order to inform each other about the hot parts of 3D printer, or some good or not so good part of the printed model. To our best knowledge, these non-verbal interactions were not found in other research (Mavrou, Douglas, & Lewis, 2007; Wang, Laffey, Xing, Galyen, & Stichter, 2017; Xin, 1999) that examined the contribution of different technologies to the collaborative learning between SWD and SWOD and they could be considered as a benefit to what 3D printing brings to an inclusive classroom. Based on the video data observations, the nonverbal acts precede verbal acts such as evaluation, compliment, and blame between SWD and SWOD. 3D printers provide opportunities for interactions between students with different educational needs, which leads to better interaction between them (Buehler, Comrie, Hofmann, McDonald, & Hurst, 2016; Kostakis, Niaros, & Giotitsas, 2015). Nevertheless, some negative non-verbal interactions between SWD and SWOD were also detected in our research. For example, negative nonverbal act eye roll (this act signifies boredom in the cultural environment in which the research was conducted), was registered only in the behaviour of SWOD. For example, this non-verbal act was registered with the SWOD in the third school and occurred during an SWD suggestion. Our result indicates SWOD were more prone to negative nonverbal acts than SWD. Similar to negative verbal acts, negative nonverbal acts were caused by individual SWOD and do not represent the actions of the whole group. These individual characteristics are caused by the personality and social development of students (Dyson & Rubin, 2003). Although, the negative nonverbal act Withdrawal occurred more frequently in SWD and it occurred most often as a reaction to negative verbal or nonverbal SWOD acts. For example, in 5.2 school, SWD idea presenting was interrupted with „Not now with your ideas”, by one of the SWOD. As a result, SWD reacted by withdrawing and starting to draw an optional drawing in the notebook and he returned to the task participation on the call of another SWOD. Even though these are examples of individual behaviour, it is very important that the teacher is aware of the possibility of such nonverbal acts, as well as the reasons for their occurrence and to do the best to prevent them. McGregor and Forlin (2005) suggest that teachers can prevent the exclusion of SWD by individuals from SWOD by raising students’ awareness of accepting diversity and disability and developing their communication skills.

By comparative analysis of survey numbers, it can be noticed that both groups of students have a positive opinion on cooperative learning with the use of 3D printers. The data indicates that the implementation of cooperative learning with the use of 3D printers has contributed to possibilities for SWD and SWOD to express and to hear different opinions. This is in line with the results of Xin (1999), which show that after the realisation of cooperative learning, SWD and SWOD have a more positive opinion on communication and friendship with peers. The conclusion from the second set of survey questions is that SWD had more positive opinions than SWOD students. SWOD has a less positive opinion than SWD on the contribution of cooperative learning with the implementation of 3D printing to their knowledge. One possible reason for the slightly less positive opinion of SWOD can be found in the students’ answers to the open question, in which they suggested that increasing the number of printers, simplifying and facilitating the use of modelling programs could improve this way of learning. This assumption is supported by the results of Roberts and Smith (1999) who pointed out that if there are obstacles that require an increase in work from SWOD or SWD, they may provoke a more negative student opinion. However, both groups of students would like to use cooperative learning with the implementation of 3D printers with their peers in the future and believe this learning approach should be implemented in each school. When appropriate and supportive behaviour develops in SWD and SWOD then the students are more likely to have positive opinions on using that interaction in the future (Malone, Fodor, & Hollingshead, 2019). Both groups of participants were of the opinion that technical improvements of 3D printing were necessary, which would be reflected in faster and more accurate printing, as well as software improvement of 3D modelling software, in order to easily model the models. These student suggestions can be very important inputs for developers working in the field of 3D printer development and 3D modelling software for their usability in education. Improving these features of 3D printers and 3D modelling software would probably increase their potential to be used as effective collaborative tools in learning. This assumption is supported by recent researches by Dukuzumuremyi and Siklander (2018), who indicated that students’ perceptions about the easiness and efficiency of using technologies are one of the most important factors that influence 3D printing implementation as a collaborative learning tool in an inclusive classroom. Comparing the results of our research with previous research dealing with CSCL in inclusive classrooms, it can be concluded that the use of 3D printers contributes to social interactions in inclusive classrooms through the following: the development of group creative ideas of students and the practical creation of digital and physical 3D models; enabling interactions between SWD and SWOD in process of model design, interactions of SWD and SWOD with a computer; SWD and SWOD with 3D printer; SWD and SWOD with a 3D printed model as a product of student work. This research indicates that the use of 3D printers in an inclusive classroom fosters positive verbal and non-verbal communication between SWOD and SWD through interactions with learning content displayed in a digital and physical environment (multiple representations of learning content) which is missing in CSCL that is not supported by the use of 3D printers.

6. Conclusions and Implications for Practice

Implementation of 3D printers in cooperative learning between SWD and SWOD contributes to the verbal and nonverbal interaction and communication between peers. It fosters communication between SWD and SWOD through working on modelling, interacting with the computer while modelling, the process of printing and interaction with a printed model. The probability of interaction occurring is greater than when 3D printers are not included. The students have a positive opinion on this way of learning. All the students who participated in this research would like to use this approach in future learning and consider this way of learning should be present in all schools. However, there are some challenges that teachers who implement 3D printers in an inclusive classroom should bear in mind. Some of the SWOD is prone to negative verbal and nonverbal interactions with their disabled peers. In this research, this was registered as individual conduct of a small number of SWOD. Nevertheless, this behaviour should be prevented by raising awareness of these students on the acceptance of peers with different needs and increasing their communications skills. The results of our research indicate that SWD are more involved in the process of 3D modelling insofar as the task is related to their previous experience or when the task is tailored to these students, which teachers should keep in mind when creating tasks. Additionally, the students believed that the implementation of 3D printers would be more efficient if technical obstacles to implementation were solved. According to the student’s opinions, increasing the easiness of use of 3D modelling software and accuracy in 3D printing should be improved. These aspects should be considered in the future development of 3D modelling software and 3D printers for educational purposes. Our research set out several assumptions which should be examined in further research. We propose that further research should be undertaken with the aim to provide an answer to the following question: How does the design and adaptation of the tasks and instructions for SWD and SWOD influence the promotion of socio-affective aspects of 3D printer-supported cooperative learning.

Disclosure Statement

No potential conflict of interest was reported by the authors.