A Phenomenography Study of STEM Teachers’ Conceptions of Using Three-Dimensional Modeling and Printing (3DMP) in Teaching

There is a large amount of research that indicates that the use of 3DMP in STEM education improves students’ knowledge, motivation, and participation in the learning process. Nevertheless, despite the existing attempts to market 3DMP in education, its adoption in schools remains low. A number of studies with teachers in secondary schools and colleges indicate that teachers’ perceptions of 3DMP are one of the key factors for its successful use. However, to our best knowledge, there is no research that examined STEM upper primary school teachers’ perception on 3DMP. Through phenomenographic approach, this study is seeking to address the existing gaps. Four conceptions of 3DMP teaching emerged: (1) 3DMP as tools for classroom modernization; (2) 3DMP technical and software characteristics’ impact on implementation; (3) 3DMP as a tool for learning and improvement in teaching; (4) 3DMP and students’ professional orientation, teachers’ professional development. These four categories are connected by five key aspects of variation: impact on students, impact on teachers, classroom activity management, authenticity, subject-curriculum matters. The results of our study indicate that the mathematics and science teachers have a more sophisticated opinion on 3DMP than teachers of technical education, engineering, and informatics who mostly require additional training when it comes to using 3DMP in STEM education. Comparatively, science and mathematics teachers need support with implementation of software and 3D printers as a technical tool. Considering that this study’s teachers were early adopters of 3DMP, any future research should explore conceptions of experienced users.


Introduction
Introduction of digital technologies has a pivotal role in modernization and improvement in the field of education. Development of new digital technologies and their adaptation for application in the classroom have enabled new ways of teaching and learning. One of these technologies is 3D modeling and printing (3DMP 1 ). In the process of 3DMP, a computer (software) model-designed format is created into a 3D object adding materials layer-by-layer. There is a large body of research that indicates that the use of 3DMP in education has certain benefits, especially in the fields of science, technology, engineering, and math (STEM). For example, Berry et al. (2010) concluded that using 3DMP in education could increase student engagement, inspire creativity, and improve attitudes towards STEM subjects. Several studies have documented that 3DMP contributes to student skills in technical drawing, product design, and development (Lütolf, 2013;Steed & Weevers, 2016). Some of the research recognizes the important role played by 3DMP in achieving the STEM learning goals in schools. Research that confirms this was conducted by Grant et al. (2016) and showed that during the modeling of animal parts in biology classes, students implemented knowledge from other subjects such as computer science, mathematics, and engineering, which lead to achieving STEM learning goals. A similar conclusion was reached by researchers investigating the contribution of 3DMP in chemistry (Brooks, 2016), physics (Dumond et al., 2014), mathematics (Bull et al., 2015), computer science (Minetola et al., 2015), and technological education (Martin et al., 2014). Some research indicates that 3DMP contributes not only to the learning process but also to the teaching process and to the teachers who use it. The results from studies of Horowitz and Schultz (2014) and Ford and Minshall (2019) demonstrate that 3DMP contributes to teachers' interest and engagement in STEM and that 3DMP increases their abilities to adapt the content they teach to their students' capabilities. In the latest research on this topic by Arslan and Erdogan (2021), it was found that the application of 3DMP contributes to the development of more positive attitudes of teachers over its application and impact on learning. The teachers stated that 3DMP transforms abstract concepts into concrete visual representations, facilitates learning and enables longer knowledge retention, makes lessons enjoyable, and strengthens interest, creative thinking, and design skills, as well as motivation to create different educational materials for certain contents. Another study establishes that 3DMP represents an ideal supplement technology in teaching, because teachers are able to create original learning materials, which are not easily accessible and available (Karaduman, 2018). However, in opposition to the studies listed above, the findings of Bull et al. (2015) and Kitts and Mahacek (2018) differ. They concluded that teachers are still unable to fully utilize the benefits of 3DMP, due to the lack of adequate guidance on the use and maintenance of 3DMP, as well as the lack of understanding of processes required for the implementation of this teaching tool. This is in line with the results of Nemorin (2017) and Nemorin and Selwyn (2017), who stress that teachers could face serious obstacles in the implementation of 3DMP in teaching, which could further lead to frustration, physical fatigue, mental exhaustion, tedium, and anxiety. For example, research by Arslan and Erdogan (2021), Karaduman (2018), and Maloy et al. (2017) confirmed that the teachers faced challenges in applying 3DMP, because of the lack of multidimensional and creative thinking skills, which are necessary for designing 3D learning objects. In addition, the teachers in these studies were limited in the use of these facilities due to internet disconnects, occasional power outages, and insufficient knowledge of 3D-printer use in practice, as well as insufficient online models available due to their limited number. These are some of the reasons why a number of teachers do not want to apply 3DMP in their practice, which is directly reflected in the lower representation of these models in learning. Additionally, the results from recent studies demonstrate that research examining teacher perceptions, opinions, and conceptions about using 3DMP in education is uncommon. They suggest that research in this area should be intensified, in order to obtain a clearer picture of the application of 3DMP in practice and to take adequate steps to promote, train, and raise awareness of teachers on the importance and benefits of applying 3DMP in teaching and learning (Arslan & Erdogan, 2021Karaduman, 2018Maloy et al., 2017;Simpson et al., 2017). It is especially important to examine the attitudes, opinions, and conceptions of STEM teachers on the implementation of 3DMP in teaching, taking into account that previous research indicates not only the benefits of 3DMP application but also the unsatisfactory experiences of teachers, which encourages them to stop using 3DMP in practice. If we have in mind that the successful introduction and use of technologies in education mostly depends on teachers' conceptions, perceptions, and opinions about them (Kafyulilo et al., 2015), then the importance of examining teachers' opinions and conceptions about 3DMP is clear. To our best knowledge, there is no research that examines conceptions of STEM upper primary school teachers on 3DMP. Our research seeks to contribute to the knowledge in this area and assist in filling the literature gap.
This prospective study was designed to investigate how 3DMP is experienced by STEM teachers in upper primary schools in Montenegro, and uncover teachers' conceptions on 3DMP as a teaching tool, with the focus on identifying potential differences among the categories. In accordance with the aim of the study, research questions were developed. This research seeks to address the following questions: what are the qualitatively different conceptions of STEM teachers who experience 3DMP in teaching and if there are differences in the conceptions on the use of 3DMP in teaching of teachers who teach different subjects within STEM. The remainder of the paper is organized into the following sections: methodology-providing information on how phenomenographic research approach was used in our study, the results showing the conceptions of STEM teachers who participated in this research on 3DMP, discussion, and conclusion.

Methodology
We prepared our study in the form of exploratory and descriptive research. Our study implements a qualitative approach, based on phenomenography with the aim to examine teachers' conceptions about the usability of 3D modeling and printing in STEM teaching. For data gathering, individual and focus group interviews were conducted. The teachers in our study were divided into four focus groups. Having two or more focus groups in research ensures different perspectives, reduces possible threats, and increases reliability of the results (Lundy-Allen et al., 2004;Patton, 2002).

Procedures and Interviews
All of the teachers who participated in our study were primarily part of workshops about implementation of 3D modeling and printing in teaching as volunteers. The workshop lasted 8 h and was divided into the following parts: theoretical introduction about 3D modeling; theoretical introduction about 3D printers and printing; introduction to the implementation of 3DMP in education; practical 3D modeling for teachers; practical 3D printing of teachers' models, discussion. The participating teachers' schools were equipped with a 3D printer each after the workshop, for which the project funds were previously allocated to. The workshop trainers provided their contacts to the teachers to contact them in case of obstacles which they could not overcome on their own. One year after the workshops, and after the teachers acquired experience in the practical implementation of 3DMP in teaching, individual and focus group interviews were conducted.
The interviews took place in schools where the teachers worked. The moderator was skillful in group discussions and used interview guide and pre-determined questions. The interviews were first conducted individually with teachers, then these teachers were grouped into focus groups, which were also interviewed. During the focus group interviews, the moderator allowed the participants to lead the conversation in a direction that was of interest to them, but always attentively guided the interview to the main topic of 3D printers. At the beginning of their interviews, the participants were told that they do not have to agree or support someone's opinion. It was emphasized that it is important to express their opinion based on personal experiences. The interviews lasted between two and two and a half hours. Three focus groups were made out of 9 teachers, and one was made with 10 teachers participating in it. With the implementation of four focus groups with this number of participants, we tried to ensure diversity and reliability of our data, which is recommended by Cohen et al. (2017) and Patton (2002). The interviews were audiotaped. This procedure provided the researchers with the possibility to collect the qualitative data, but also to get insight into the participants' feelings, emotions, contradictions, and tensions, which are connected with the topic (Lundy-Allen et al., 2004). The interview guide consisted of 20 open-ended questions. The interviews started with three general questions, followed by more specific ones. Some of the examples of general questions are as follows: What is your experience of using a 3D printer in teaching? What do you think about the added value of using 3D printing in teaching? and examples of specific questions are as follows: Describe one learning activity in which you implemented 3D printing, and everything went well? How did you organize and prepare exercises where you used 3D printing in your lessons? The full list of questions used in this research is provided in Appendix A.

Participants
Thirty-seven upper primary school teachers from Montenegro participated in our study. STEM subject does not exist in the Montenegrin national education curriculum. In the education system of Montenegro, students in primary school (age 6 to 14) are taught scientific, technical, and informatics content in an integrated form (integrated biology-physicschemistry-technics and informatics content) in the first five grades as a part of the teaching subject nature. After that, STEM content is taught in the separate subjects of biology, physics, chemistry, technical education and engineering, and informatics for the next 4 years. Mathematics is a separate teaching subject from the first to the final grade of primary school. After completing upper primary education, STEM content is taught in secondary schools (age 15 to 19) through separate teaching subjects. However, in the curriculum, as well as in teacher handbooks, suggestions are given for achieving an interdisciplinary, inclusive, and future-oriented STEM approach in teaching. These suggestions in the Montenegrin curriculum were given to try to make connections between the following teaching subjects: biology, chemistry, physics, technical education and engineering, informatics, and mathematics. Teachers who teach these subjects in upper primary schools were included in our research. The characteristics of the participants related to the teaching subject are presented in Table 1.
A phenomenographic research (Åkerlind, 2005;Green, 2005) stresses the importance of participants' variety in age, gender, and experience for this kind of research. These suggestions were followed when the participants for this research were selected. As it can be seen in Table 2, the age range of the participants in this research is from under 30 to over 60, while the range of years of teaching experience ranged from under 10 to more than 30 years of teaching experience. All of the teachers who participated in the research taught students from 6th to 9th grade (age 11 to 14) in an upper primary school in Montenegro.

Data Analysis
The phenomenographic approach (Marton, 1981) was employed for analyzing the teachers' interviews. All of the interviews were transcribed at first and after that left aside for 2 weeks in order to create the distance for the researchers' minds and enable them with a more open-minded data analysis (Trigwell, 2000). These transcripts were processed using seven-step phenomenographic analysis of data Han and Ellis (2019), Dahlgren andFallsberg (1991), andMcCosker et al. (2004). This means that the data were processed in the following way: Step 1-familiarization: The transcribed interviews were read and reread several times by the researchers in aim to be familiar with the data and its details, and to create personal notes about the information; Step 2-condensation: To reveal data patterns, the most representative sample units were marked in the transcript.
Step 3-comparison: The representative sample units were compared to find sources of variation or agreement; step 4-grouping: Sample units with similar traits were allocated and grouped.
Step 5-articulating: In this step, similarities within each category of sample units (statements) were described.
Step 6-labeling: The meaning of the categories was expressed linguistically.
Step 7-contrasting: A contrastive procedure was used for comparison between the obtained categories in aim to reveal their individual meanings and similarities, as well as the differences between them. Structural relations were used for establishing the hierarchy between the developed categories. In terms of offering an explanation of relationships between categories, multiple ways of experiencing the same phenomena, the structure of the "outcome space" was used (Åkerlind, 2005). Hierarchy establishment was based on the evidence that some categories were intertwined with others, as it is suggested by Åkerlind (2005). The same authors claim that the confirmation of the hierarchy could be supported by logical and empirical perspectives, and both should be confirmed by the transcribed data. Based on this suggestion, we concentrated our transcript revision on the important ideas and claims and their level of manifestation in transcripts. For example, one of the participants said: From this part of the transcript, "follow modern technology trends" was extracted as an important part to summarize the key meaning. Another teacher provided the following explanation: ST: Modern educational games and applications enable today's students to build entire cities and their inhabitants in a virtual space. Today, students spend a lot of time in such an environment. Then imagine the disappointment of one contemporary student, who in biology class then learns about heart from a poster or picture, like it was 100 years ago. The application of 3DMP enables the student to acquire knowledge in a modern and interesting way by combining virtual environment in which he spends a lot of time with the physical world and teaching materials.
From the passage above, "aligning with the needs of contemporary students" is formed as the main idea representing the meaning of the paragraph. Following the same principle, "key meanings" were extracted from all the parts of the teachers' narrative, which were then compared with the aim of developing categories. Based on these principles, the above quotes and the others, which are described in the results section, are classified under category 1: 3DMP as a tool for classroom modernization. In the process of developing categories and their structural relationships, three rules suggested by Marton and Booth (1997) and Han and Ellis (2019) were followed: each category should reveal some distinctness from other categories about explored phenomena; categories should be parsimonious and present most important data; the categories in hierarchy relations should be clearly and logically specified. In the aim to explore and understand the potential difference in conceptions about 3DMP of  0  3  5  1  0  1  6  0  2  Technical education and engineering  3  6  0  1  6  1  1  0  7  1  1  Informatics  7  2  1  5  0  3  0  2  5  2  0  Mathematics  6  4  0  2  5  2  1  2  5  2  1  Total:  21  16  1  11  16  7  2  5  23  5  4 teachers from different STEM subjects, the total frequency of the conceptions in the focus groups and frequency by individual participants were used. With this approach, we tried to adapt the recommendations about distribution of general, main, and achieved conceptions of teachers given by Tsai (2009) and Hsieh and Tsai (2017), as well as the recommendation of Irvin (2006), who suggests using phenomenography to examine the conceptions of a collective group of participants instead of analyzing the understandings of individuals. In our research, the main concept category was determined based on the highest frequency in the narrative of the teacher focus groups and individual interviews. The category with the highest achievement was the one that was recorded in the teachers' narratives and was regarded as the most sophisticated. Following the suggestions of a previous similar phenomenographic research (Marton & Booth, 1997;Åkerlind, 2005), the main concepts have been presented with categories that are logically related to each other. In terms of increasing complexity (from less to more complicated) from basic to sophisticated categories of concepts, the relationships between categories are posited to form a structural hierarchy of inclusiveness (Åkerlind, 2005). It is important to stress that the hierarchy is based on the evidence of some categories being inclusive of others, rather than on a value judgment (good or bad) (Åkerlind, 2005). This means that the highest category (the most sophisticated) is the most complete one, which includes the elements of less ranged categories (Khan et al., 2019). For example, the fourth category-3DMP and student professional orientation, teacher professional development is hierarchically above the three lower categories. In this category, 3DMP is not only seen as a medium for modernization of the classroom, or as a teaching tool that improves technological information skills and the learning and teaching process, which is the case in the three lower categories, but also as contributing to students' and teachers' professional orientation and professional development. According to this principle, a hierarchical relationship was made for all the categories in this research. Trigwell (2006) suggests that the reliability of data in the phenomenography research could be achieved by easy recognition of the categories by others. The internal logic of the categories and their easy representation indicate the validity in the phenomenographic research (Marton, 1986). Validity and reliability can be achieved by applying the opinions of experts in this field, which are related to the developed categories based on the transcript (Cope, 2004). In this research, three colleagues with the experience in qualitative research and their opinions about developed categories were used for checking the reliability. A high level of agreement between the researchers and external experts (84%) indicates that the data in this study could be considered reliable (Säljo, 1988).

Results
In this part of the paper, we are presenting the qualitatively different categories of teacher conceptions of using 3DMP in STEM education. Through the phenomenographic analysis, four qualitatively different categories were discovered: • Category 1: 3DMP as a tool for classroom modernization; • Category 2: 3DMP technical and software characteristics' impact on implementation; • Category 3: 3DMP as a tool for learning and teaching improvement; • Category 4: 3DMP and student professional orientation, teacher professional development.
Along these different categories, five key aspects of variation were marked: impact on students, impact on teachers, classroom activity management, authenticity, and subject-curriculum matters.

Category 1: 3DMP as a Tool for Classroom Modernization
This category represents teachers' opinions on the use of 3DMP in order to keep up with trends in technology. For example, TET: Every profession changes with time, so if a teacher wants to do their job in a good and contemporary manner, he/she should follow the trend in that area.
Teachers' expressions indicate that this category is not based on teaching subject matter requirements, rather it was based on teachers' aspiration to modernize the teaching process. This can be seen in the following quotations: ST: Generally, I don't hesitate to explore new modern teaching materials and tools. That made my teaching more interesting for myself, as well as for my students. MT: I like to experiment with new teaching methods, materials, and technology and observe the student reactions to this innovation in the learning process.
The experience of teachers presented in this category is connected with their personal innovativeness and its implementation in the teaching process. An illustrative quotation, for this case, was given by an IT teacher: In both business and private life, I like to research and use new technologies and digital tools.
The experience described by the teachers in this category relates to teachers' need to modernize their teaching and follow trends in technology and innovation. As could be seen in quotations, teachers like to experiment with new teaching materials, and methods as well as with new educational technologies, and this helps them to modernize their teaching.

Category 2: 3DMP Technical and Software Characteristics' Impact on Implementation
In this category, teacher views about technology and software characteristics of 3DMP and its use in educational purposes are considered. It affirms the potency that 3DMP brings to a classroom, as well as some of the shortfalls which should be remedied in aim to increase the productiveness of 3DMP as a teaching and learning tool. For example, ST: With my students, I assembled a 3D printer in our school. It is an easy process, students almost did it on their own. MT: It is really good and useful that students can observe the working process of 3D printers and model building. On the other hand, this process is too long for school class time.
As the main technical strengths of 3D printers, teachers considered the ease of assembling printers, the possibility to fix a printer part by part if that is necessary, ease of use, security, and the ability for students to observe the printing process itself. An illustrative quotation given by a TET: 3D printers are really easy and safe to use, and a huge advantage is the machine could be fixed part by part if due to use some of the parts get damaged.
Additionally, teachers considered accuracy and precision in model printing, long time of printing, sensitivity to shaking, and requiring maintenance, as some of the weaknesses of the technical characteristics of 3D printers, which should be improved for better efficiency of this tool in educational purposes. One of the IT participants said: Printers on their own are sometimes inaccurate in the printing process, commonly because of small quakes that can occur due to students touching the table on which the machine is standing.
Other teacher aspects were concentrated on the characteristics of software for 3D modeling. Free software that is adapted to different levels of digital skills, possibilities to modify the models and possibilities for group-collaborative modeling are software characteristics that teachers considered beneficial to the classroom. Some examples of illustrative quotations-a TET: In the beginning, I was using Cookie Cutters, when I become more comfortable with it, I started using other more sophisticated software, such as TinkerCAD and Blender. The simple to more complicated approach, which I implemented with students, was a winning combination. MT: A huge advantage is the students can participate in the modeling process together, model in groups and collaborate through the software for 3D modeling.
However, teachers highlighted some of the software disadvantages that they experienced. The long time required to master the modeling process, which requires high knowledge in informatics, as well as the long time needed for modeling, is the main software disadvantage according to the teachers.
For example, ST told us: The software requires a lot of time for teachers and students to learn the skills to use it with ease. Also, the modeling process takes a long time, which can jeopardize the realization of planned learning outcomes and activities.
The second category is more sophisticated than the first because in it, 3DMP is not only seen as a means of modernizing the classroom but this category also considers the technical and software characteristics of this teaching tool that enable or hinder its application in teaching. Therefore, relative to the technical and software characteristics, teachers consider the possibilities of applying 3DMP in teaching.

Category 3: 3DMP as a Tool for Learning and Teaching Improvement
In this category, using 3DMP is viewed as a way to increase the learning activities of students, as well as the teaching skills of teachers, or generally as a tool for improvement of the educational process. 3DMP was conceptualized as a tool that scaffolds the implementation of new or adapted teaching approaches and methods on the teachers' part, as well as promotes active participation in the classroom on the students' part. For instance, an IT noted: There are a lot of opportunities, which could be conducted when using 3DMP in classrooms, which provide students with the opportunity to explore, construct, and correct their mistakes in this process. In addition, teachers regarded the implementation of 3DMP as providing multiple learning content representations, which contribute to students' motivation to participate in the learning process and gain knowledge at higher cognitive levels. Teachers focus on implementing 3DMP as a learning tool for teaching improvement that shifts from using new teaching methods and approaches towards student groups and collaborative work to providing different roles to the students where this increases their knowledge at a higher cognitive levels. MT: 3DMP provides concretization of the mathematical tasks, students are involved in practical problem solving, which leads to knowledge acquisition at a higher cognitive level. One more interesting aspect of this category is promoting student practical activities in the learning process inside and outside of school. Students are seen as active participants in knowledge building through 3D modeling from home or other places, not only in schools. For example, an ST said: I used 3DMP in the teaching topic "Cell". Aiming to include 3DMP in my lesson, I used some new teaching methods and approaches. For example, I used flipped classroom, where students were supposed to learn about cells from textual and video material that I provided to them after which they modeled the cells and printed them.
The third category is more sophisticated than the second one because the attitudes of the teachers in this category are not focused mainly on the technical and software characteristics of 3DMP, but on the possibility of using the capacity of this teaching tool for delivery, teaching approaches, methods, and instructions. The teachers pointed out the possibilities which 3DMP provides for the improvement of teaching and learning and the contributions that are achieved in the application of 3DMP.

Category 4: 3DMP and Student Professional Orientation, Teacher Professional Development
The respondents in this category perceive 3DMP in terms of student professional orientation and teacher professional development. When the teachers discussed various 3DMP enhancements to teacher professional life, they pointed out that it contributes to their digital skills, language skills, and communicational skills with colleagues from different teaching subjects. For example, a TET noted: When I was attending a workshop on 3DMP, I learned a lot of new English words, specific to 3DMP, which means I didn't only increase my digital but language skills as well. From their professional development, teachers' focus was shifted to the student professional orientation and choosing the future occupation. The teachers expressed their view of 3DMP as a learning tool in a manner that 3DMP allows the students to get a closer insight into some of the future occupations by placing them in different roles, as well as developing their abilities to perceive the importance of transdisciplinarity in everyday-life problem solving, and developing their affinity for STEM occupations. One example of an illustrative quotation from this perspective given by an ST: During the application of 3DMP in chemistry classes, students apply mathematics, language, and computer science knowledge and they understand that in real life all knowledge permeates and makes one knowledge. I think that in this way students indirectly develop affinities for choosing a STEM profession as their future occupation.
The teachers considered 3DMP as a tool that connects students with different professions-allowing them to understand the importance of transdisciplinarity, critical thinking, and creativity as crucial twenty-first-century skills for their future occupations. For example, an MT said: Students get not only the skill to create or do something but as well to critically analyze and evaluate their work. This really contributes to their development of 21st-century skills and abilities to choose their future occupation easier.
Some aspects of this category ended up being similar to those of category 3, where the teachers considered 3DMP as a tool for learning and teaching improvement. However, these two categories are different, because in category 3, the main focus is on the learning activities and their outcomes, while in category 4, the focus is on skills required for a twenty-first-century job market. This reflects the higher hierarchical position of the fourth category compared to the third because in it, 3DMP is seen as a teaching tool that can provide skills and experiences that could potentially influence the choice of future occupation, which is not the case within category 3.

Distribution of Categories
The categories described indicate the concepts which all the STEM teachers who participated in this research had about 3DMP. In order to diagnose differences in concepts between teachers who teach different subjects within STEM, the frequency of concepts in teachers by subjects was calculated. Conceptions of teachers from different STEM teaching subjects about 3DMP, based on the frequency of concept keywords and phrases in the focus groups, are presented in Table 3. For example, as can be seen in Table 3, the science teachers who participated in this research had concepts which spread over four categories. However, category 1 is the least frequently mentioned in the total narrative of this group of teachers (72), category 3 is considered as the main because it is the most frequently mentioned theme (193), the "highest achieved" was category 4 appearing 78 times in total teachers' narrative. Similar approaches for determining the main categories were used in previous research on a similar topic (Tsai, 2009;Hsieh & Tsai, 2017). The main and achieved conceptions of 3DMP are marked in the table. Conceptions about 3DMP from science, informatics, and mathematics teachers are present in all four categories; however, conceptions of technical education and engineering teachers are only present in three.
For the science and math teachers, the main conception is category 3-3DMP as a tool for learning and teaching improvement. For the technical education and engineering teachers that is category 2-3DMP technical and software characteristics' impact on implementation. For the teachers from these three focus groups the highest achieved category is 4-3DMP and student professional orientation, teacher  Table 3). The results of our research indicate that parameters such as gender, age, and technology experience did not affect teachers' perceptions of 3DMP and their use for educational purposes. On the other hand, as can be seen in Table 3, teachers of science, informatics, and mathematics had more sophisticated concepts on the application of 3DMP in teaching than teachers of technical education and engineering, as indicated by the highest achieved category. A detailed analysis of the board concludes that mathematics and science teachers had more sophisticated opinions than informatics teachers, because their main concepts were concentrated in the third category while the concepts of informatics teachers were mostly concentrated in the second category. As the results show (Table 3), the mathematics and science teachers had the most similar concepts about 3DMP. The concepts of these two groups of teachers are concentrated mostly in category 3, while the fourth category is the highest one achieved.

Relationship Among the Categories
The four categories, which represent teacher conceptions about using 3DMP in education, are interrelated with five key aspects: impact on students, impact on teachers, classroom activity management, authenticity, and subject-curriculum matters. Table 4 presents a brief summary of the key aspects and their relations in the categories.

Key Aspect 1: Impact on Students
Along with this aspect, the impact of the conception of using 3DMP with students extends from access to the new learning tool and experience, through the possibilities to be used by students with different technical skills and placing students in different roles, to developing skills and affinity for a future occupation. In category 1, 3DMP is seen as an enabler of new experience to the students, and students are seen as the recipients. For example, an MT said: It is good to implement at least once a year 3DMP in classroom because it is a totally new learning tool for students and provides them with a new experience. In category 2, the impact of 3DMP on a student shifts from experience perceiving to experience creating in accordance with students' abilities. As one example we can take the statement of an IT: 3DMP provides the opportunity for students with different computer skills to model and print something by using different software in this process. In category 3, the students are seen as active learners who can undertake different roles of investigators, constructors, and evaluators during the using 3DMP in the learning process. An illustrative quotation made by an ST: When using 3DMP, students are placed in different roles, which contributes to their engagement in learning, but they also get to know different professions such as researcher, constructor, and engineer. 3DMP in category 4 is seen as a learning tool that provides knowledge to students and skills that are required in different professions. Additionally in this category, the students are seen as active learners who choose their future occupation based on knowledge skills and experience. Another illustrative quotation given by an ST: 3DMP helps students to create a realistic picture of some occupation and to choose what they like.
Key Aspect 2: Impact on Teachers Within this aspect, teachers' conceptions of 3DMP are expanded in all four categories, from the reaction of teachers to new technologies to the professional development of teachers and their training. In the first category, teacher conceptions indicated that teachers view 3DMP as a modern technology that needs to be incorporated into teaching in order to follow trends. Actually, the inclusion of 3DMP in the classroom was the teachers' response to technical modernization. For example, an IT said: We are living in a modern digital environment most of the time, however, schools are still stuck in using projectors only, and using new technologies such as 3DMP enables teaching to be enhanced with modern technology trends. However, the second category reflects the teachers' view about the effort and time required for mastering the skills for using the 3DMP in teaching. The teachers viewed 3DMP as a teaching tool that requires some effort and time on their part to be able to use it in the best possible way in teaching. As one of the MT participants stated: I believe that it takes a lot of time and effort for a teacher to acquire the knowledge and skills one needs in order to be able to use all the benefits of 3DMP in teaching. In category 3, this aspect shifts to observing the 3DMP as a tool that stimulates teachers to modernize and adapt their teaching practice. In this aspect, there is an emphasis that teachers believe that 3DMP is not only a new technology but also a trigger for designing new tasks for students, new activities, as well as modernizing and adapting approaches and teaching methods. In an example given by an ST: The application of 3DMP in teaching requires that the whole approach to teaching be adapted -from the tasks for students to the approaches and methods of teachers. In the last category, teacher emphasis shifts from seeing 3DMP as a teaching tool to the opportunities for professional development and improvement of digital and communication skills through using 3DMP. The teachers consider the implementation of 3DPM in teaching as an opportunity to increase their digital competence and communication between colleagues from other teaching subjects, but as well as with their students. The following statement given by an ST is one example of it: While practicing 3DMP in chemistry classes, I did not only increase

Key Aspect 3: Classroom Activity Management
In this aspect, the view of the role of 3DMP in teaching differs from changing interaction and individual work to the group work in the classroom, over practical students learning activity to the developing students' abilities for self-evaluation and corrections of their works. Categories 1 and 2 are associated strongly with communication and the type of student work (individual, pair, and group). A TET example: When using 3DMP in teaching, students are more focused on each other, discussing and finding solutions in the group, rather than individually asking questions and waiting for information from the teacher. An ST example: When using 3DMP in learning, students have the opportunity to present their ideas to their peers in a digital environment, but also in physical, both oral and non-verbal communication over the model they have created. This contributes to better cooperation between the students. In these two categories, 3DMP is considered a learning tool for providing different type of communication: verbal, non-verbal, face to face, and online, which in teachers' opinion enrich students learning experience and contribute to the learning process. 3DMP is also seen as a tool that increases collaborative pair and group student activities. In category 3, 3DMP is seen as having an impact on student practical learning activities in school and beyond. In fact, a new opportunity is created for students' practical learning activities in school, but also at home or in the library, or in any other place where they can have access to the Internet and 3D modeling software.
As an ST pointed out: While using 3DPM in my classes, I did not anticipate that students could access modeling from home, or any other place that suits them when it suits them. This significantly accelerated the activities, so we had more time for presentations and discussions in class. Teacher colleagues should keep this in mind when planning classes. Also, students organize their learning activities independently, by themselves, while the teacher only coordinates them. Some of the teachers stressed the possibility for team modeling enabled by the software used, so students can learn in pairs or groups, as if they are in their homes, even if they are physically separated or in different time zones. This provides an opportunity for teachers to expand the teaching activities they plan beyond school classes. Category 4 provides an extended view of the influence of 3DMP on student abilities to provide feedback to each other, as well as to self-evaluate their works. An illustrative quotation from an IT: When they print the model, students see the quality of their work, and can assess where they went wrong. This changes the students' opinion about their work, because if they are just reading the text they usually have no idea how well it is done but when they see the model, they have a clear picture of their work. This promotes self-evaluations. These skills or habits could be really useful not only in school but in everyday life and work. But this also puts us teachers in a position to think about evaluating students' ability to spot a mistake and want to correct it. The teachers consider these as very useful skills which are important for future student activities in learning and working. In the teachers' opinion, this also encourages themselves to change the ways of evaluating student work. The new way of evaluation would take into account more aspects-such as students' awareness of where the mistake was made and the desire to correct it.

Key Aspect 4: Authenticity
This aspect presents different teacher views of the authenticity brought by the implementation of 3DMP into the classroom. Authenticity as an aspect starts in category 1 and continues to category 4. The authenticity displayed in category 1 is related to the possibilities for the teachers to design and for the students to work on problem-solving tasks which are solved based on practical (non-fictive) and realistic student working activities and examples. As one of the TET participants said: I usually ask students to solve a practical problem through modeling and printing, for example a holder for a broken umbrella wire, or wrench. In this way, my task and students activities are real, non-fictive and this is of great importance in teaching in learning. The students know why they learn something. In category 2, the uniqueness of 3DMP as a teaching tool is seen from the perspective of a variety of interactions that it brings to the classroom. According to the STEM teachers who participated in this study, 3DMP in the classroom provides a wealth of different interactions, such as interactions of students with computers, software for modeling, 3D printers as a machine, and with a 3D printed model. For example, ST: When using 3DMP during the teaching subject matter 'Cells', the students had the opportunity to interact with a textbook, computer, software, printer and of course the model... Aspects in category 2 are similar to the aspects in category 3 where teachers consider the implementation of 3DPM in teaching as a possibility for multiple representations of learning content. However, this aspect in category 2 is related to the student activities, and in category 3 to the representation of the learning content. An ST illustrative quotation: ...while working, students had the opportunity to see what a cell looks like in an image, as a 3D representation in software and as a printed model. Each of these ways provided students with new information and contributed to their knowledge. The teachers used 3DMP to provide the students different representations of the same learning content. They see 3DMP as a learning tool that enables the transformation of learning content from two-dimensional presented in the textbook, to 3D in digital software version and 3D in printed version. Some of the teachers emphasized that 3D modeling has a huge contribution to students' abilities to visualize the learning content and if there is no time for printing in the classroom, modeling should be conducted anyway. In category 4, the teacher conceptions are focused on increasing teacher skills and developing student skills and abilities for working in digital and physical environment. We can take a MT's opinion as an example: During 3DMP, the students develop their abilities for learning and working in digital, but as well as in the physical environment. During this process their skills in this area are upgraded as well. The teachers consider this especially important for students as this is one of twenty-first century requested skills.

Key Aspect 5: Subject-Curriculum Matters
This aspect permeates through categories 1 to 4, in regard to how 3DMP could be included in STEM education and how to meet all the learning outcomes in the subjects provided for in the curricula. In category 1, the teachers face difficulties and challenges to meet learning outcomes indicated in the curricula. As can be seen from the following ST example: The realization of classes with 3DMP takes more time than usual, which can jeopardize the realization of other teaching topics. Given that the curriculum is rigid, it would be good to create software for 3D modeling targeted at science or STEM education that would contribute to the effective application of 3DMP in teaching. The reason for this is that if they include 3DMP within one teaching topic, they usually process that teaching topic for more classes than the curriculum allows for, which in turn leaves fewer classes for other topics. The teachers pointed out that this could be improved if some of the software for 3DMP would be designed for the specific purpose of science education. Both categories 2 and 3 are associated to subject learning content. In category 2, 3DMP is seen as a tool, which provides the transfer of knowledge from one subject into the others. Quoted from an MT: The good side of 3DMP is that it enables the students to apply informatics knowledge to math or other subjects. In this category, the teachers mostly consider the implementation of knowledge in informatics, technical education, and engineering into other STEM subjects. On the other hand, in category 3, this teacher perspective changes into a more sophisticated one, where the teachers considered that through the implementation of 3DMP in one class, the learning outcomes from different subjects could be achieved, which leads to the "real STEM." For instance, given by a TET: I believe that 3DMP and robotics are the two best teaching aids that enable transdisciplinarity in learning and actually help in achieving real STEM in schools. In category 4, perceived benefits from 3DMP are directed to development of students' affinity for future STEM occupations. The teachers believe that if students have the opportunity to learn through "real" STEM activities, they will be more interested in future STEM occupations. As an example, we can take the opinion of one of the TET participants in the research: If students learn STEM subjects this way, then they will be more interested in them, and choose them as future occupations. And if teachers teach in this way then they get new ideas, perspectives and skills for teaching STEM. Also, the practice of STEM classes with the application of 3DMP would improve the skills of teachers to plan and implement them.

Discussion
This study revealed four teacher conceptions, held by upper primary school STEM teachers, about 3DMP and relations between them. Firstly, overarching features of the output of this research are discussed from the perspective of all STEM teachers, and afterwards, the differences in perceptions about 3DMP between teachers from different STEM subjects are discussed. The two first conceptions of STEM teachers (categories 1 and 2) which consider 3DMP as a novel technology for classroom modernization and technical and software characteristics of 3DMP are considered simple or less complicated conceptions. The results obtained in this study are supported with those of Stein et al. (2011) and Khan (2015); there, surface-level conceptions related to some educational technologies are focused more on the tool and equipment than on the students. As explained in methodology section, the teachers who participated in this study were voluntarily included in the workshop about 3DMP in teaching. They had 1 year of experience in this domain. Taking this into consideration, the participants could be considered early adopters (Rogers, 2003). The desire and ability to examine new technologies, software, and teaching techniques and instructions, as well as to pay attention to their characteristics, are some of the common characteristics of early adopters (Hixon et al., 2012). This could be one of the possible reasons why the two first categories of STEM teacher conceptions about 3DMP were connected to classroom modernization and 3DMP technical and software characteristics.
From these simple conceptions, the move is towards more complex and sophisticated ones, related to students, and better learning and teaching. The more sophisticated conceptions (categories 3 and 4) considered 3DMP as a tool for learning and teaching improvement and its influence on student professional orientation and teacher professional development. This concept trajectory, from technology and subject content orientation to the learning process and learners' orientation, is supported by the previous research, which employs phenomenographic analysis to explore teacher conceptions about technologies (Shah et al., 2020). The findings of this study broadly corroborate earlier research on teacher conceptions of various types of technology-enhanced teaching, along with 3DMP (Hsieh & Tsai, 2017;Holzmann et al., 2020).
On the other hand, the present research reveals some new and very important conceptions of teachers about 3DMP. The most striking conception among participants in this study is the following: the application of 3DMP would contribute to students' ability to understand the importance of transdisciplinarity and choose STEM occupations as their future job, as well as, develop critical thinking, self-evaluation skills, and abilities to work in an intricate environment, both physical and digital. This teacher conception is especially important if we consider the fact that the most secondary and upper primary school students lack basic interest in pursuing STEM careers (Eccles & Wang, 2016;Blotnicky et al., 2018). In our research, the teachers thought that the 3DMP provided students with the right STEM learning experience, which could lead to the development of student affinities for these occupations. The results of our study can be very useful for educational policy creators in designing educational and promotional activities to promote STEM professions to students. The results of this study show that 3DPM could be used for promoting STEM occupation as a professional option to the students. Our results are supported by Nugent et al. (2015), Zhang and Barnett (2015), and Mangu et al. (2015), which concluded that the student affinity for STEM occupation is related to their experience and knowledge in this area. The teacher conceptions that indicate that the use of 3DMP in learning and teaching can contribute to student affinities to choose a STEM field for future occupation are supported by the results of previous research (Ali et al., 2019;Lin et al., 2021). However, our findings disagree with the results reported by Cheng et al. (2020) who indicated that integration of 3DMP does not affect student STEM careers. Unexpectedly, the teachers considered 3DMP as a useful teaching tool for their professional development, as it increases their abilities to teach STEM together with their digital, communication, and evaluation skills. This finding has important implications for developing prospective workshops about using 3DMP for teachers. Teacher trainers should pay special attention to the possibilities of using 3DMP for achieving STEM and transdisciplinarity principles in schools, but also the possibilities for professional development of teachers when holding workshops on the use of this technology in teaching. This was confirmed by Asempapa and Love (2021), who indicate that 3DMP could contribute to teacher abilities to implement STEM in the classroom or achieve transdisciplinarity between science subjects and mathematics, together with increasing collaboration between teachers. This is a specifically interesting finding, especially if we are taking into consideration that the collaboration and communication skills of teachers are considered important twenty-first century learning and innovation skills (Bedir, 2019;OECD, 2018), and encouraging their development is one of the main goals of the action "Education 2030." All five conceptions contained elements of student and teacher-oriented contribution. The conceptions about teacher-oriented contribution are divided into: 3DMP enhancing teaching and 3DMP promoting teacher professional development. The conceptions about student-oriented contribution areas are split into two parts: 3DMP enhancing the learning process and 3DMP developing student affinity to choose a future occupation. This is in the line with Weller et al. (2015) and Holzmann et al. (2020), who claim that 3DMP requires fundamentally different logic and incentives to develop new knowledge and skills. In our study, we did not find any intermediate orientation between student and teacher-oriented contribution. These results are similar to those reported by Samuelowicz and Bain (2001) and Khan (2015), which indicate that teachers "conceptions of educational technologies" can be student-oriented or teacher-oriented.
As can be seen in Table 2, there are differences in the main and achieved categories between the teachers who teach different subjects within STEM. For the teachers of sciences, informatics, and mathematics, the highest achieved category is fourth, 3DMP and student professional orientation, teacher professional development. However, the highest achieved category for the teachers who teach technical education and engineering was the 3rd one-3DMP as a tool for learning and teaching improvement. In our research, no difference was found in the concepts of the teachers that can be related to the gender age, or years of teaching experience of the participants in the research. Our results are supported by Bell (2016), who used a phenomenographic approach to explore technical education and engineering teachers' perceptions about STEM. One of the conclusions is that technical education and engineering teachers have conceptions about STEM which are not at a high level of sophistication. In our research, one more difference in STEM teachers' conceptions regarding the 3DMP is revealed. For science and mathematics teachers, the main conceptions were classified in the third category (placing students in investigator, constructor, and evaluator roles), while for informatics, technical education, and engineering teachers, these conceptions were classified in the second category (possibilities to be used by students with different levels of computer and technological skills). This indicates that the main conceptions of science and mathematics teachers about 3DMP are more sophisticated than the conceptions which informatics, technical education, and engineering teachers have, considering that the second category covers the attitudes of the teachers related to technical and software characteristics of 3DMP, which coincides with the narrow specialty of teachers of informatics, technical education, and engineering. Upper primary school teachers of TET and IT in Montenegro teach students about the basic technical characteristics of different machines, as well as basic software characteristics and programming processes. This may be one of the reasons why their main conceptions about 3DMP are in the technology and software-oriented second category. This assumption is supported by Abrami (2001), Huberman (1992), and Lin et al. (2018) who indicate that teachers of technology and computer-related subjects are oriented towards the technical and software characteristics of technology, more than teachers of other subjects, because it is a shared part to the official curriculum. On the other hand, our results show the opposite of the results of Instefjord and Munthe (2016) and Lavicza et al. (2020), who concluded that the teachers' concepts about digital learning tools are related to general teaching and digital skills and they are not related to the teaching subject of teachers. Taking into consideration, that on the basis of our best knowledge, there is no available research that explores the concepts of upper primary and secondary school teachers on 3DMP, it is difficult to establish strong connection between the concepts that TET and IT teachers had about 3DMP in this research and the subject content they teach. There is abundant room for further progress in determining connections between STEM teachers' concepts on 3DMP and the subject and content they teach. This is an important question for future research.
Based on these results, it can be concluded that teachers of informatics and technology would need to be provided with additional training including improvement of their skills to implement STEM teaching in schools. Similarly, teachers of science and mathematics need to be provided with training on the application of technical and software tools that would improve the application of 3DMP in teaching.

Conclusion
This study aimed to examine upper primary school STEM teacher conceptions about using 3DMP in their teaching. A phenomenographic research method was used to examine teacher conceptions about 3DMP. The results of this investigation show that upper primary school STEM teacher conceptions about 3DPM are classified into 4 categories: 1. 3DMP as tools for classroom modernization, 2. 3DMP technical and software characteristics' impact on implementation, 3. 3DMP as a tool for learning and teaching improvement, 4. 3DMP and student professional orientation, teacher professional development.
The second category pair of teacher conceptions encompass more sophisticated teacher attitudes about 3DPM compared to the first two categories. The teachers who participated in this research were of the opinion that the implementation of 3DMP in teaching contributes to the realization of STEM learning goals in practice and that it can influence the development of students' affinity to choose one of STEM professions as a future occupation. This study has identified that mathematics and science (biology, chemistry, and physics) teachers had a more sophisticated opinion about 3DMP than teachers of technical education, engineering, and informatics. Therefore, the results of the study can serve as inputs to program and workshop designers on teachers' professional development programs, aiming to develop the training, which will contribute to the more sophisticated conceptions of teachers of technical education, engineering, and informatics about STEM. The results of this study lead us to the conclusion that teachers of science subjects and mathematics need to be provided with training on the technical and software part of 3DMP, while teachers of engineering and informatics need to be provided with training on the principles of STEM teaching with 3DMP, in aim to achieve the full effect of 3DMP in STEM education. At the end of this paper, we would like to list a few shortcomings of this study, which would be preferable to be addressed in future research. The first limitation is related to the lack of similar studies examining the conceptions of STEM teachers about 3DMP, so the results of this study could not be compared directly with similar studies. Another drawback is that teachers in this study can be considered beginners in 3DMP in teaching. There is a possibility that the conceptions of teachers who have used 3DMP in teaching for a long period of time would differ from those of the teachers who are early adopters. Further investigation could shed light on these possibilities, and provide knowledge on conceptions about 3DMP of STEM teachers, who have many years of experience in their application.