Introduction
Education frequently refers to the STEM acronym as the partial or full integration of the separate disciplines of Science, Technology, Engineering and Mathematics, including a focus on twenty-first Century competencies (Koul et al.,
2018; Timms et al.,
2018). Research evidence suggests there is a need to advance STEM education across Australia in order to ensure international economic competitiveness (Education Services Australia,
2018; Hudson et al.,
2015; Office of the Chief Scientist,
2013). One of the key reasons for this drive is the decline in enrolments and performance within STEM education (Education Council,
2015; Education Services Australia,
2018). Caplan et al. (
2016) explain that, whereas studies have been conducted within high schools, research indicates the need to engage students prior to the ages of 11–14 years to ensure longterm interest in pursuing these disciplines.
STEM education and the development of critical STEM skills are essential for Australia’s future of economic success, particularly when faced with unknown working conditions due to innovation and technology (Caplan et al.,
2016; Honey et al.,
2014; Marginson et al.,
2013; Timms et al.,
2018). The Foundation for Young Australians (
2017) identified that occupations requiring these skills have risen by 70% and involve higher pay than those which don’t. Additionally, the World Economic Forum (
2017) highlighted that automation and COVID-19 will have impacts on working conditions and that around 85 million jobs could be displaced by these changes; however, 97 million new roles will replace these jobs, requiring additional sets of skills. The World Economic Forum's (
2017) top 15 skills for 2025 are outlined in Table
1 and refer to the skills that will be essential for the future workforce. Therefore, the development of STEM competencies from an early age is crucial to building a workforce with the capacity to undertake these new roles. These skills are referred to later within the Results and Discussion sections to highlight links between STEM education and the needs of the industry.
Table 1
World Economic Forum (
2017) top 15 skills for 2025
1 | Analytical thinking and innovation |
2 | Active learning and learning strategies |
3 | Complex problem-solving |
4 | Critical thinking and analysis |
5 | Creativity, originality and initiative |
6 | Leadership and social influence |
7 | Technology use, monitoring and control |
8 | Technology design and programming |
9 | Resilience, stress tolerance and flexibility |
10 | Reasoning, problem-solving and ideation |
11 | Emotional intelligence |
12 | Troubleshooting and user experience |
13 | Service orientation |
14 | Systems analysis and evaluation |
15 | Persuasion and negotiation |
Because of the need to inspire the young generation’s enthusiasm for STEM education, it is important to determine effective strategies or circumstances that target engagement across the integrated disciplines. Learning environment research is an extensively researched field that has been built upon for decades. The learning environment can be described as the psychosocial and emotional dimensions of a classroom that are identified from the perspective of a student and/or educator, including relationships, perceptions and attitudes (Fraser,
2012). The use of extensively validated questionnaires to measure perceptions within these environments is an established practice (Koul et al.,
2018). Teachers utilise their learning environments to convey their expectations, directly impacting student perceptions of learning areas (Watt,
2016). Therefore, it is critical for researchers to determine which characteristics of STEM learning environments have positive or negative impacts on these perceptions in promoting engagement and aspiration.
Methods
Design
This research utilised a mixed-methods case study approach to determine the perceptions of upper-primary students on their STEM learning environments and their interactions with their teacher. Upper-primary education was a focus because it is at this early age when children begin to form their career aspirations (Wiebe et al.,
2018). This project firstly utilised the pervasive approach within learning environment research involving using a quantitative questionnaire to measure student perceptions of their classroom emotional climate, their interactions with the STEM teacher, and attitudes to STEM. Fraser et al. (
2020) suggests that combining quantitative and qualitative methods when implementing questionnaires assists with explaining findings and provides greater detail. To elicit further understanding from students, semi-structured focus groups were conducted with a smaller number of students to collect further details about these perceptions, as well as their preferred teacher behaviours. Ethics approval for this study was granted by the Human Research Ethics Committee of our University.
Sample
This project was conducted at a co-educational independent school in Perth, Western Australia within Year 5 classrooms. The context was chosen because of its reputation for successfully implementing STEM education, therefore making it suitable for measuring the a quality of the STEM learning environment and investigating the impacts on student perceptions. Additionally, teachers of the four classes collaborated on their projects, so the students shared similar experiences. While the teachers did not always refer to the acronym ‘STEM’, they utilised the integrated approach as defined for this style of education.
Quantitative methods
The quantitative information was collected through the questionnaire which was comprised of scales from three tools and used in a previous study (Fraser et al.,
2020). The first instrument measuring Classroom Emotional Climate (CEC) was based on the Classroom Assessment Scoring System (CLASS) and the Tripod 7 C’s student perception survey (Ferguson,
2012; Fraser et al.,
2020; Hamre & Pianta,
2007). Its eight scales of Care, Control, Clarity, Challenge, Motivation, Consultation, Consolidation, and Collaboration were adopted from Fraser et al. (
2020). The second section, which measured Attitude to STEM, was adapted from the Attitude to Science scale from the Test of Science Related Attitudes (TOSRA) (Fraser,
1981). The final section measured teacher–student interactions and was based on the positive scales of the Questionnaire on Teacher Interaction (QTI) (Fraser et al.,
2020). The four scales are Leadership, Helpful/Friendly, Understanding, and Student Responsibility/Freedom (Wubbels & Brekelmans,
2005). While originally designed for high school classrooms, both the TOSRA and QTI have been validated within primary school classrooms, making them appropriate for our study (Goh & Fraser,
1997; Koul et al.,
2018). The instrument was firstly piloted through six focus groups, and then validated through the use of Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) in a junior high school setting, with a reasonable sample size of 658 participants in a separate study (Fraser et al.,
2020). Functional validation of this instrument within upper-primary students was then completed through this project.
Prior to students completing the questionnaire, the researchers discussed STEM experiences with the participants, including what STEM is and projects or experiences completed at school. During implementation, to ensure understanding and to allow opportunities for questions, the researcher read each item to the students as they responded using a 5-point Likert scale. Additionally, the reliability estimates suggest that students had the capacity to respond to the items as intended.
The questionnaire was administered to 100 students (46 male; 54 female) across four Year 5 (10–11 years old) classrooms with(student ratio 23:24:26:27). All students who had provided consent, and were present on the day, participated in the study. The sample size was impacted by COVID-19, because some students had not returned to school after it had been shut down during the pervious term.
Quantitative analysis was conducted through a range of strategies. Data from the questionnaire was inputted into an Excel spreadsheet, and then processed using IBM SPSS version 27 Statistics Software. Firstly, descriptive statistics were generated for means and standard deviations for the data set. T-testing was used to compare genders and Analysis of Variance (ANOVA) compared means between classrooms. Correlations and muiltiple regressions were used to determine significant relationships between variables. Additionally, alpha reliability, eta2 and Pearson’s correlation were reported.
Qualitative methods
A sample of 12 students was purposively selected to participate in semi-structured focus groups; an outline of participant selection is shown within Table
2. Five male and seven female students were purposively selected,based on their gender and their responses to the Attitudes to STEM items to ensure representation, to determine any differences of opinion between male and female students. We selected 3 out of 5 male students who indicated positive perceptions of STEM education and that they were motivated to pursue further study in STEM. One male indicated negative perceptions and no motivation to pursue further STEM education. The final male student was undecided about STEM education, and indicated no motivation to pursue it further. Three female students were included who indicated positive perceptions of STEM education, and motivation for further STEM learning. Two indicated negative perceptions and no motivation to the field. The final two female students indicated that they were undecided and had no motivation to pursue STEM further. Guiding questions were utilised to help to prompt the students to discuss their thoughts about their STEM learning environment.
Table 2
Purposive sampling for qualitative focus groups based on quantitative results
Male | 5 | Positive (≥ 4) | 3 | Indicated | 3 | Not Indicated | 0 |
Negative (2 ≤ | 1 | Indicated | 0 | Not Indicated | 1 |
Undecided (2 > < 4) | 1 | Indicated | 0 | Not Indicated | 1 |
Female | 7 | Positive (≥ 4) | 3 | Indicated | 3 | Not Indicated | 0 |
Negative (2 ≤ | 2 | Indicated | 0 | Not Indicated | 2 |
Undecided (2 > < 4) | 2 | Indicated | 0 | Not Indicated | 2 |
Thematic analysis was applied to the qualitative data. Braun and Clarke (
2006) state that this approach, when applied rigorously, can reveal insightful and valid information, and is particularly useful for highlighting similarities and differences of perspectives between participants. Because this project had a focus on differences between male and female perspectives, this approach was particularly beneficial.
A deductive thematic analysis was used for this project because the researcher brought preconceptions and ideas to the coding which developed through the guiding questions. This also allowed connections to be made naturally between the questionnaire and the responses from the semi-structured focus groups. However, any codes or themes that were not linked to the preconceived concepts were used to ensure that the students had voice and that data collection was flexible to these ideas.
During preliminary data examination, transcripts of the focus groups were created by an online professional company for accuracy. The coding process was comprised of a systematic approach to the transcripts, with data points being highlighted and given their initial first codes. These data were then categorised based on collective meaning. The coded data were rearranged into more-refined themes, with some of the data points that reflected more than one category being placed within their ‘best fit’ categories. This process, and the resulting themes, were then reflected upon and reviewed by two colleagues to ensure accuracy and relevance, as suggested by (Braun & Clarke,
2006). From these final themes, data points were examined alongside the quantitative data to further improve reliability.
Qualitative results
Research objective 4
The semi-structured focus groups gave greater insight into the perceptions of the students collected through the questionnaire. Overall, students outlined many positive characteristics of their STEM learning environment and how it impacted on their attitude towards STEM education. They were also able to provide their preferred perceptions about what could improve their STEM learning environment.
Thematic analysis is a qualitative research method that can be widely used across a range of epistemologies and research questions. Braun and Clarke (
2006) observed that this method is useful for identifying, analysing, organising, describing, and reporting themes found within a data set. All data generated were transcribed and read iteratively to locate concepts being represented by the data. All three researchers in the team independently analysed the data and only themes that were identified in all researchers’ analyses or were accepted for inclusion.
Open coding procedures as delineated by Corbin and Stauss (
2008) were utilised and involved continually asking questions such as “Which category does this incident/word/phrase allude to?” and “What are the similarities or differences between the two emerging concepts?” Words/themes/and other data pieces alluding to a particular theme were colour coded. Processes of bundling, grouping similar units, and deletion of synonymous units were utilised to arrive at final categories as delineated in the research findings.
All three authors independently conducted the thematic analysis to arrive at the data themes. The independently generated themes by the authors revealed a high degree of agreement. The data pieces were revisited collaboratively to discuss the disagreements and to develop a consensus on the themes. Focus group discussion transcripts responses were shared with the school principal, who identified themes emerging from the data that could be matched with themes derived by the research team.
Thematic analysis identified seven themes which assist in the understanding of student perceptions of their STEM Learning Environments, as well as perceived preferred environments (see Table
8). Each theme was then further broken into sub-categories that represented different aspects of the theme. Themes 1–6 all included comments, attitudes and feelings about what students were currently experiencing within their STEM learning environments. The final theme, Preferred Environments, included any data points about possible improvements for STEM learning environments. The following sub-sections report results of the thematic analysis.
Table 8
Sub-categories of themes derived from student focus group interviews
1. Student Freedom | 1. Boundaries |
2. Choice |
2. Peer Collaboration | 1. Grouping |
3. Problem Solving | 1. Teacher Support |
2. Peer Support |
3. Trial and Error |
4. Communication | 1. Noise |
2. Teacher Control |
5. STEM Learning | 1. Emotions |
2. Understanding/Misconceptions |
6. Time | 1. Limitations |
7. Preferred Environments | 1. Hands On |
2. Environment |
3. Choice |
4. Technology |
5. Peer Collaboration |
Student freedom
The first theme that emerged from the semi-structured focus groups was Student Freedom, which is a particularly interesting theme and insight from the students. It was broken into the sub-categories of Boundaries (items related to teacher setting boundaries) and Choice (items relating to student choice, such as choosing how to solve a problem). Students did not respond well when their classroom was free from boundaries and structures related to behaviour and control; however, students discussed positive freedoms when given options about their learning. Students were comfortable when the teacher gave them boundaries but sought choices that sat within these boundaries. For example, if students were given an integrated STEM project to complete, they might have a choice about optional topics within the scope of their learning or about how to present the final product of their learning.
Peer collaboration
The second theme to develop was Peer Collaboration, which closely relates to a range of STEM skills, and was only given the one sub-category of Grouping. Students described how they were given ample opportunities to work in teams and the ways in which they formed their learning groups. They explained that their teacher would sometimes choose their groups, allow them to choose groups, or occasionally group them at random. Interestingly, students were able to explain why they thought that their teachers did this, because they knew they were more productive when grouped by an adult. Additionally, they also noted that they performed better when in mixed-gender groups, even though they said that they wouldn’t choose these groups if given the option:
2.1.2.8: Sometimes our teacher picks our groups. If you do not like that. But sometimes she just lets us cause um.
Interviewer: So, do you think that is most of the time? She lets you choose?
2.1.2.10: Yeah.
2.1.2.11: Um, yeah, she – [teacher] gives us relative, like, um. In our groups, we can really choose who we want. But yeah, she normally says there has to be like, a split gender. ‘Cause otherwise, you just have whole groups of all girls and whole groups of all boys.
Interviewer: And what would you prefer?
2.1.2.13: Um, I like the split. It gives different perspectives normally.
Interviewer: Perspectives coming from the?
2.1.2.15: I would probably prefer to have an all-girl group but I think you do work better when you have like, different genders.
Problem solving
Problem Solving was the third theme, with the sub-categories of Teacher Support, Peer Support and Trial and Error. Again, this theme has many connections to a range of essential STEM competencies required for a successful future workforce. Teacher Support was discussed within all three of the focus groups, with students being able to outline how they felt very supported during the problem-solving process. They were able to describe how their teachers created a balance between support and challenge through prompting (as opposed to giving answers) and how difficult the learning was. Peer Support was also discussed and highlighted the cooperative nature of the learning environment where students were able to seek answers from each other prior to approaching a teacher for support. While this was the case, they mentioned that they knew that, if they couldn’t find an answer from a peer, their teacher would be happy to guide them. Regarding Trial and Error, the final sub-category for the Problem Solving theme, students discussed being given multiple opportunities to solve problems. Students also indicated that they felt happy that they were able to trial a range of methods, including their own ideas, to try and solve problems. They noted that it allowed them to experience success and improve:
3.1.1.5: You can get other friends down or the teacher might just let it try and, try and let it [work] out itself, and if you still cannot get it, she will come down and help.
3.1.1.6: They [the teacher] give us a sudden urge to like try to find another idea and go around the problem and find a new solution.
Interviewer: Great. And how does that make you guys feel?
3.1.1.10: Um, better because we know we have something to work with.
3.3.1.1: Like, I cannot give up so I can revise how it could be better and what can you – you can improve.
Interviewer: How does it make you feel when you do that?
3.3.1.3: Uh, happy I guess.
Interviewer: Makes you happy?
3.3.1.5: Makes us more inspired, so it can like, make more, uh, ideas and better ideas.
3.3.1.6: Ah yeah. Just giving advice.
3.3.1.7: Just like you are [inaudible], some actual help.
3.3.1.8: Yeah, that you help you and use the result to generate ideas.
Communication
The fourth theme was Communication, which was broken into the sub-categories of Noise and Teacher Control. Within this research project, communication was seen as an essential element of skills within STEM learning environments, because of the importance of having the capacity to spread and share new knowledge competently to others. The focus within this theme was the opportunities students were given to communicate and present their knowledge. The Noise sub-category related to the control that the teacher had over the learning environment. While students frequently noted how they had opportunities to work cooperatively or collaboratively to problem solve, they also noted that their teachers could ask them to reduce their noise, even when it was relevant to their projects. It is interesting to note the student perspective on this situation. It is likely teachers ensure that noise stays at a productive level so that there are still supportive boundaries in place for the students to prevent a disruptive environment. The Teacher Control sub-category related to teacher decisions about how students would communicate their knowledge. Students generally reported that they were given guidelines, and this usually meant that all groups were presenting in the same way, but the presentation styles changed for different projects. Students were able to list posters, iMovie, performances, websites, typed reports and speeches as different ways in which their teachers asked them to communicate their knowledge to others.
STEM learning
The fifth theme was STEM Learning, which was divided into sub-categories of Emotions and Understanding/Misconceptions. This first sub-category centred around emotional experiences within SLEs, whereas the second focussed on student understanding of STEM education. The data indicated, overall, that students were experiencing positive and engaging learning within their SLEs and that their perceptions of STEM were positive. Interestingly, students were not always able to define exactly what STEM education is, but they were able to name learning projects that they had completed and explain how these were STEM learning tasks.
Time
The sixth theme was Time, with the sub-category of Limitations. This related primarily to the balance between giving the students enough opportunities to problem solve and complete tasks and ensuring that the mandatory curriculum requirements were being met. This is frequently a dilemma for teachers and can be a difficult balance. One of the key points made by students was that they identified that sometimes projects were spread out over weeks or a term, and other times they could be spread out over a single day. They remarked that they preferred projects that went over a day rather than spread between other curriculum subjects, which is interesting but not always possible. Additionally, while they explained that they were given ample opportunities for problem solving, they still felt they didn’t always have enough time to complete their work. As time is a complex factor within classrooms, it can be difficult for both students and educators to navigate this delicate balance.
Perceived preferred learning environments
Perceived Preferred Learning environments was the final theme, and it was divided into sub-categories for the different characteristics that the students believed would improve their STEM learning environments. For the first sub-category of Hands On, students discussed wanting more physical experiences during which they could create and play to learn about concepts. One child explained that being involved in the learning was far more effective than “just like, watching videos or writing stuff down”. The second sub-category of Environment was related to ideas that the students had for their physical environment, including having more opportunities for flexible seating, including beyond the classroom where they could have meeting spaces or gaming rooms. Some schools develop Makerspaces, laboratories or technology laboratories to support their students, and this idea connects with those spaces. The third sub-category of Choice contained points that the students made about being able to have more agency with their learning. They discussed selecting with whom they work, having a choice of presentation style and being able to co-construct the curriculum for projects. The fourth sub-category of Technology related primarily to having access to more technologies. Specifically, the students listed Minecraft, Micro:bits, bee bots, and more opportunities for making and testing that were related to technologies. The final sub-category of Peer Collaboration involved students discussing how they prefer to work together. This included larger learning spaces and having access to people beyond their own class, or even beyond the school. This could connect to other classes or year levels, or to special guests and excursions.
Overall, students indicated that they believed that their STEM learning environment was positive and that generally students felt engaged. Students were able to describe many positive characteristics that their teachers implemented to support their STEM education and described several changes that they believed would improve their learning environment. Because student perceptions and perceived preferred perceptions provide valid and valuable insights into high-quality STEM learning environments, seeking these perceptions from multiple contexts could provide further insights for improving the engagement and aspirations of students within STEM education. As Fraser (
2012) states, it would be a positive step to change learning environments, where possible, to suit the preferred perceptions of our students.
Discussion
The drive to improve STEM education to meet future workforce needs for Australia’s economic success is critical (Education Services Australia,
2018; Hudson et al.,
2015; Office of the Chief Scientist,
2013). High-quality STEM education that develops STEM skills will be essential for unknown roles which are necessary for Australia’s future (Caplan et al.,
2016; Honey et al.,
2014; Marginson et al.,
2013; Timms et al.,
2018). Learning environment research has demonstsrated that student perceptions of their Classroom Emotional Climate and their interactions with their teacher can impact their academic achievement and motivation (Reyes et al.,
2012) and, therefore, it is important to consider student perceptions when designing learning environments (Fraser,
2012).
Using a mixed-methods approach, a students were able to express their perceptions about their STEM learning environments, including their perceived preferred perceptions. This insight might assist in further understanding the characteristics of STEM learning environments that are conducive to high-quality education and student engagement as needed by industry.
Key findings from this study include that students within this context did not associate well with behavioural freedom and, in fact, preferred environments that have structure and also promote agency linked to curriculum and communication choices. Similarly, in a recent study, Koul et al. (
2021) found that students had relatively positive perceptions of teacher control during STEM learning, with females scoring more highly. Similar conclusions were reached in other studies, such as research with middle-school students in which girls had more positive perceptions of order, involvement and organisation (Waxman & Huang,
1998).
Peer Collaboration was also an important aspect of STEM learning environments, with students having the opportunity to learn and grow with each other through problems, without simply being given answers by their teachers. This essential concept links closely with several of the World Economic Forum (
2020) Top 15 Skills for 2025, potentially including Complex Problem-Solving (3), Critical Thinking and Analysis (4), and Emotional Intelligence (11).
Another key finding included Problem Solving, with students being given opportunities to trial a range of solutions to problems. Closely related to Peer Collaboration, in the sense that students liked how their teachers didn’t simply give them the answer, this characteristic also develops a number of the Top 15 Skills for 2025 (World Economic Forum,
2020), potentially including Analytical Thinking and Innovation (1), Complex Problem-Solving (3), Critical Thinking and Analysis (4), Resilience, Stress Tolerance and Flexibility (9), and Reasoning, Problem-Solving and Ideation (10).
Communication was a theme linked to noise control within the classroom and was related to opportunities to discuss problems, as well as methods for communicating knowledge. Interestingly, communication skills are not explicitly stated within the World Economic Forum (
2020) Top 15 Skills for 2025, though they might sit within categories such as Leadership and Social Influence (6) or Persuasion and Negotiation (15). Students also explained that they would like more agency regarding choosing how to present their information, which would then also promote the development of skills such as Creativity, Originality and Initiative (5).
Time was an interesting key theme that was brought up by the students through the focus groups. It relates to several of the other key findings because, without enough time, problem solving are more difficult to implement effectively. As discussed previously, it could be hard for children to understand why teachers need to limit time within the classroom to ensure that they are balancing curriculum requirements and high-quality teaching.
The perceived preferred environments findings highlighted concepts such as hands-on learning, physical environments, choices relating to agency, the frequent use of technology, and varied opportunities to collaborate beyond the classroom. These potential changes to the learning environment would also promote the development of a range of the World Economic Forum (
2020) skills, including but not limited to Active Learning and Learning Strategies (2), Technology Use, Monitoring and Control (7), and Technology Design and Programming (8).
The development of these skills within STEM learning environments is crucial, and the use of integrated approaches to develop these competencies through authentic contexts is an effective method (Nadelson & Seifert,
2017; Rosicka,
2016). Positive experiences with these disciplines will also assist in improving student attitudes and aspirations (Murphy et al.,
2019), which can lead to further engagement with these fields. Therefore, this study utilised the extensively researched field of learning environments to measure student perceptions of their context, to identify positive aspects of high-quality STEM learning environments, and identify potential deterrents to engagement.
Conclusion
This research is significant in that it measured student perceptions of their STEM learning environment to determine how these perspectives impacted their engagement within their context. Additionally, we collected student perceived preferred characteristics of STEM learning environments and identified potential ways to improve engagement. This involved the functional validation of a questionnaire for upper-primary classrooms that can be utilised by other researchers and schools to assess students’ perceptions within other contexts.
The research also identified a potential deterrent which could negatively impact student perceptions of STEM education. Relating to the Student/Responsibility Freedom scale of the questionnaire and discussed multiple times within the semi-structured focus groups, it was shown that the students did not associate positively with freedom that related to behaviour issues and a lack of structure. As a child’s perception is their reality, it is crucial that we take into consideration their thoughts, feelings and attitudes when designing learning environments.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.