Key Predictors of Outdoor Science Teaching in Regular Classes

Abstract

Outdoor science teaching provides meaningful, hands-on learning opportunities, yet remains inconsistently practiced in schools. This quantitative study investigated how specific teacher-level predictors associate with the frequency of outdoor science teaching in regular classroom time. A structured questionnaire was developed and administered to 258 Slovenian science in-service teachers, measuring their attitudes, self-efficacy, perceived support, professional development, connection to nature, and inclusion of nature. After reducing items through exploratory factor analysis, variables were then entered into an ordinal logistic regression model of outdoor teaching frequency. Teachers’ attitudes and self-efficacy emerged as the strongest predictors, while their professional development, support, connection to nature, and inclusion with nature were not significant. Supplementary correlations showed that teachers perceiving greater support or stronger connection to nature held slightly more positive attitudes. The correlation indicated a weak but positive link between accessible environments and outdoor teaching frequency. Based on these findings, we recommend actively fostering teachers’ positive attitudes and self-efficacy, as these personal factors are the primary drivers of how frequently outdoor science teaching occurs. To support the development of these personal factors, professional development programs should be systematically designed to build confidence, provide hands-on mastery experiences, and reinforce positive beliefs. Rather than being treated as predictors in their own right, these elements should be understood as structured pathways for strengthening the attitudes and self-efficacy that drive outdoor teaching practice.

1. Introduction

Outdoor education refers to learning that takes place beyond the classroom, ideally in natural or intentionally designed outdoor spaces (Borsos et al., 2022; Novljan & Pavlin, 2022). It includes overlapping concepts—such as outdoor teaching, nature-based education, place-based learning, and learning outside the classroom—each with distinct goals and levels of curricular integration (Falzon et al., 2025). We adopt a widely used conceptualization of outdoor education as a multifaceted framework encompassing diverse environments and purposes (Rickinson et al., 2004). Further, our definition excludes indoor or simulated experiences (e.g., museums, galleries, virtual field trips, zoos), which lack the immersive qualities of natural settings. In line with recent work, school laboratories are also excluded, with the focus placed on outdoor science teaching within school grounds or nearby areas (Ayotte-Beaudet et al., 2023).

Outdoor science teaching, as a distinct subset of outdoor and environmental education, engages students with scientific phenomena through direct observation, exploration, and inquiry (Goldman & Alkaher, 2023; Romero et al., 2022). Outdoor science teaching can also be understood as a form of place-responsive pedagogy, in which learning emerges through teachers’ and students’ sustained engagement with local environments and the social contexts of schooling (Mannion et al., 2013). It offers authentic, real-world contexts that foster not only conceptual understanding but also emotional engagement and social development (Mann et al., 2021; Romero et al., 2022). Such experiences can promote scientific knowledge along with environmental responsibility and sustainability awareness and therefore contribute to students’ holistic development (Haraldsson et al., 2024; Pirchio et al., 2021). From an educational research perspective, outdoor learning is not merely a change in location but a pedagogical approach whose value depends on how teachers integrate it into planned instruction, curriculum goals, and classroom management within formal schooling (Špernjak et al., 2026).

Despite increasing recognition of its benefits, the integration of outdoor teaching into school curricula remains inconsistent and often limited (Falzon et al., 2025). This challenge spans multiple disciplines, including science education—a field that is especially well-suited for outdoor instruction due to its emphasis on experiential methods and real-world relevance (Ayotte-Beaudet et al., 2017). Consequently, the key question is not whether outdoor learning is beneficial, but under which professional and institutional conditions teachers adopt it as a regular instructional strategy.

1.1. The Role of Teachers in Sustaining Outdoor Science Education

Whether outdoor teaching becomes a sustained and meaningful practice depends largely on teachers, who are often expected to invest additional time, energy, and personal resources when implementing approaches beyond conventional classroom routines (Dring et al., 2020). Teachers should not be left to navigate the complexities of outdoor instruction without adequate support (Schilhab, 2021). Strengthening outdoor science education requires a shared commitment from teachers, school leaders, researchers, and policy makers (Klopčič & Torkar, 2025).

A central consideration is not only teachers’ willingness to teach science outdoors, but also their perceived ability to do so within the constraints of the school environment. Beliefs, prior experiences, and confidence levels all shape professional decision-making in this context (Glackin, 2018; Schilhab, 2021). Social Cognitive Theory offers a useful lens here, emphasizing the role of self-efficacy—teachers’ beliefs in their capabilities to organize and execute instructional actions—as a key determinant of behavior (Bandura, 1986). In the context of outdoor teaching, self-efficacy is likely to influence not only whether a teacher considers taking students outside, but also how they plan for, manage, and reflect on such experiences. Those who feel competent in outdoor settings—individuals possessing contextual knowledge—are more likely to incorporate outdoor environments into their instruction. This form of knowledge encompasses disciplinary understanding, pedagogical skill, experience with field-based inquiry, and the ability to manage student behaviors in outdoor settings (Barnett & Hodson, 2001).

In addition, teachers’ personal experiences with nature and a preference for student-centered approaches to teaching may further encourage engagement with outdoor science instruction (Glackin, 2018; Torkar, 2014, 2015).

These interrelated dimensions shape how science is taught across settings, with outdoor instruction as one of the options.

1.2. Science and Outdoor Teaching in Slovenian Compulsory Education

Slovenia’s nine-year compulsory basic education system serves students aged 6 to 15 and is organized into three three-year cycles. The first two cycles (grades 1–6) correspond to primary education, while the third cycle (grades 7–9) is classified as lower secondary education (Eurydice, 2025).

Science is taught across all three cycles, although the subject structure and name changes over time. In the first cycle, science is introduced as the subject learning the environment, emphasizing basic observation and exploration of the natural world. In grades 6 and 7, science is taught as natural sciences (naravoslovje), integrating content from biology, chemistry, and physics. In the final cycle (grades 8 and 9), these disciplines are taught as separate subjects. Across all stages, the curriculum aims to develop students’ understanding of living and non-living systems through observation, inquiry, and real-world experiences grounded in environmental contexts (Skvarč et al., 2011).

Outdoor learning is explicitly embedded in the national curriculum and is primarily implemented through activity days and outdoor school (Šebjanič & Skribe-Dimec, 2019). Schools are required to organize 15 activity days annually, with three focused on science. While typically conducted outdoors, these may also involve visits to museums or science centers (Eurydice, 2025). Outdoor school, which is also compulsory, takes place at least twice during students’ education and often lasts three or more days in specialized centers, specifically the Centre for School and Outdoor Education (CŠOD—Center šolskih in obšolskih dejavnosti) (Šebjanič & Skribe-Dimec, 2019).

In addition to these structured forms, the curriculum encourages outdoor learning as part of regular subject teaching. Specifically, the natural sciences curriculum mandates that at least 40% of lessons involve student-active learning, including experimental or investigative work conducted indoors or, preferably, outdoors (Skvarč et al., 2011). This recommendation constitutes the central focus of the present study.

To teach natural sciences in grades 6 and 7, teachers are typically required to complete a five-year university degree in education or science, followed by a one-year supplemental certification program, which additionally prepares teachers to employ investigative and authentic practices in their science instruction (Eurydice, 2023, 2025).

In addition, teachers have the right and are expected to engage in regular professional development, either based on their own interests or aligned with the evolving needs and priorities of their schools (Eurydice, 2023).

While the Slovene curriculum refers to the subject as naravoslovje (natural sciences), comparable subjects are commonly termed science in the international literature; this terminology is therefore adopted throughout the article for consistency.

1.3. The Aim of the Research and the Research Questions

Existing research on outdoor science education spans several complementary but unevenly developed fields. A substantial body of work has examined the effects of outdoor teaching on students, with syntheses documenting positive academic, motivational, and socio-emotional outcomes across age groups and subject areas (Mann et al., 2021; Pulido et al., 2025). In parallel, teacher-focused research has primarily explored experiences, perceptions, and contextual conditions shaping outdoor education through qualitative and action-oriented designs, often based on small samples and case-based inquiry (Bertelsen et al., 2024; MacQuarrie, 2018; Patchen et al., 2024).

Other studies have investigated teacher-level characteristics—particularly perceived self-efficacy, outcome expectancy, attitudes, confidence, and preparedness—most often in the context of pre-service education or targeted professional development interventions (Cetken-Aktaş et al., 2025; Fuchs et al., 2026; Wolf et al., 2022). More recent systemic and survey-based research has expanded this perspective by examining a broader range of influencing factors, including school culture, leadership support, perceived costs, and contextual constraints; however, these factors are still rarely analyzed simultaneously within a single quantitative framework (Oberle et al., 2021; Pedonti et al., 2025; Špernjak et al., 2026).

At a broader conceptual level, outdoor and science education has also been discussed in relation to the ecology knowledge or climate and nature emergency, reinforcing its normative relevance (Ayotte-Beaudet et al., 2017; Mannion et al., 2013). Taken together, the literature provides substantial insight into outcomes, contexts, and individual predictors, yet remains characterized by a predominance of qualitative approaches, limited sample sizes, and fragmented examination of influencing factors. Consequently, there is limited empirical evidence on how multiple teacher-level predictors jointly relate to the frequency of outdoor science teaching in regular school practice.

To address this gap, the present study examines how a set of theoretically and empirically grounded teacher-level predictors relates to the frequency of outdoor science teaching among Slovene in-service teachers. A structured questionnaire was developed, grounded in specific predictors of outdoor science teaching practice identified in our previous study (Klopčič & Torkar, 2025) and further supported by findings from the related empirical and theoretical literature (more in Instrument—Section 2.3.).

This study examines which predictors are most strongly associated with teachers’ reported outdoor science teaching, thereby providing insight into how education policy and in-service teacher training can be designed to target these factors and address the limited use of outdoor education in science classrooms. Research was guided by the following questions:

RQ1: How frequently do teachers conduct science lessons outdoors during regular class time?

RQ2: How do specific teacher-level predictors associate with the frequency of outdoor science teaching?

2. Materials and Methods

This study employed a quantitative research design. It represents the first phase of a larger research project on outdoor science teaching in Slovenian schools.

2.1. Data Collection

The statistical population comprised Slovenian science teachers teaching the 6th and 7th grades (students aged 11 and 12). A non-probability convenience sample of 258 teachers was obtained via two-stage outreach.

First, the principals of all Slovenian primary schools were contacted via institutional email addresses retrieved from the Ministry of Education’s public database. They were asked to forward the survey invitation to science teachers in their schools. Following a modest initial response rate, a second reminder email was sent one month later to encourage participation.

In the second phase, a more targeted approach was adopted. We manually searched all Slovene school websites to identify and contact teachers whose email addresses were publicly available online. These teachers were approached directly via personalized emails. This strategy yielded approximately one-third of the total responses, although the process was labor-intensive and constrained by incomplete or unavailable contact information on some websites.

We contacted all 457 primary schools listed in the Ministry of Education’s public database of educational institutions. As no official public data are available on the exact number of teachers teaching natural sciences in Grades 6 and 7, the target population size could only be approximated. Based on the number of schools and the typical organization of science teaching in Slovenian primary education, we estimated the target population at approximately 1500 teachers. With 258 completed questionnaires, this corresponds to an estimated response rate of approximately 17%.

Participation was fully anonymous and voluntary.

Data were collected between 8 April and 19 June 2025 using the electronic survey platform 1KA, an open-source tool developed by the Centre for Social Informatics at the Faculty of Social Sciences, University of Ljubljana (General Description 1KA, 2025). The created data is deposited to the public repository Zenodo (Klopčič & Torkar, 2026).

2.2. Sample Description

The majority of the 258 science teachers identified as female (n = 237, f = 91.9%), with the remainder identifying as male (n = 21, f = 8.1%). The respondents reported an average of 18.7 years of teaching experience. Most taught both the 6th and 7th grades (n = 155, f = 60.1%,), with 18.6% (n = 48) teaching only 6th grade and 21.3% (n = 55) teaching only 7th grade.

2.3. Instrument

Data were obtained using a structured online questionnaire that was purpose-designed for this study and informed by our previous systematic review (Klopčič & Torkar, 2025), complemented by the relevant empirical evidence outlined in the following sections. The original questionnaire was administered in Slovene. For the purpose of this article, an English translation was prepared.

The instrument comprised two parts:

• Demographic and contextual characteristics, including grade levels taught, years of experience, gender, frequency of outdoor instruction, participation in professional development related to outdoor teaching, and availability of natural settings around schools.
• Six thematic sections were designed to capture selected key predictors that may shape the decision to teach outdoors. Except for the final pictorial item, all sections employed five-point Likert-type scales to assess the extent of agreement with each statement:
  • Teachers’ attitudes toward outdoor education (20 items): These items were adapted to assess teachers’ beliefs about outdoor science teaching, emotional engagement, and willingness to implement outdoor instruction (van Dijk-Wesselius et al., 2020).
  • Perceived support and constraints (17 items): Based on the barriers and support factors reported in previous studies (Ayotte-Beaudet et al., 2017; Dring et al., 2020; Dutta, 2023; Novljan & Pavlin, 2022; Scott et al., 2015), this section focuses on external conditions influencing outdoor teaching, including institutional, organizational, and structural constraints.
  • Teaching self-efficacy for outdoor science instruction (9 items): This section was adapted from two validated instruments—the Science Teaching Efficacy Belief Instrument (STEBI) (Riggs & Enochs, 1990) and the Teaching Outdoor Education Self-Efficacy Scale (TOE-SES) (Schumann & Sibthorp, 2013). STEBI assesses teachers’ beliefs about their effectiveness in science teaching, including content knowledge and instructional strategies, while TOE-SES measures confidence in managing field-based educational settings. Selected items were adapted to reflect core pedagogical tasks in outdoor science contexts, such as planning, safety, lesson adaptation, and curriculum alignment (Riggs & Enochs, 1990; Schumann & Sibthorp, 2013).
  • Connection to nature index (cNI) (16 items): Adapted from Cheng and Monroe (Cheng & Monroe, 2012), this scale measures emotional and cognitive engagement with nature across multiple dimensions that are relevant to teaching and personal orientation toward the natural environment.
  • Personal nature experience (3 items): Three supplementary items were included to capture aspects of teachers’ personal background: namely, time spent outdoors at present, time spent outdoors during childhood, and preference for rural versus urban living.
  • Inclusion of nature in self (INS): A single-item pictorial measure assessing perceived self–nature integration, adapted from Schultz (2002).

2.4. Pilot Testing and Instrument Validation

The questionnaire was pilot-tested on a sample of 15 in-service science teachers to assess clarity, structure, and interpretability. To ensure broader validity and linguistic inclusiveness, feedback was also obtained from a diverse panel of professionals, including language teachers, school psychologists, and science educators from related disciplines.

Revisions were made to improve the questionnaire structure or item clarity. Multilevel response options were used to increase the instrument sensitivity, and thematic groupings were clearly distinguished to improve the content validity.

2.5. Ethical Considerations

The study received approval from the Ethical Review Board of the University of Ljubljana, Faculty of Education. All procedures complied with institutional and national ethical guidelines for research involving human participants.

2.6. Statistical Analysis

Analyses were performed in August 2025 using IBM SPSS version 28. Descriptive statistics were produced for numeric variables (means, standard deviations, modes, and total frequencies), whereas nominal and ordinal variables were summarized by category frequencies and percentages. Not all procedures were executed on the full dataset (n = 258), as some of the respondents did not finish some of the questions in the questionnaire.

For bivariate correlations, Spearman’s rank correlation coefficient was applied due to the absence of a normal distribution. As the data were primarily measured on five-point scales, this procedure was deemed appropriate. To reduce the number of observed variables to a more manageable set of latent constructs, Exploratory Factor Analysis (EFA) was conducted, employing Principal Axis Factoring (PAF) with oblique (promax) rotation. The resulting factors were subsequently entered into an ordinal logistic regression model to predict the frequency of outdoor science teaching. Detailed values and procedures are reported in the Results Section.

3. Results

This section presents the results addressing the two research questions. First, descriptive findings illustrate how frequently teachers conduct outdoor science lessons during regular class time. Second, inferential analyses examine the associations between selected teacher-level predictors and the reported frequency of outdoor science teaching.

3.1. RQ1: How Frequently Do Teachers Conduct Science Lessons Outdoors During Regular Class Time?

Of the 248 teachers, 31.0% reported conducting up to two outdoor science sessions per year and 27.4% reported three to four sessions; frequencies declined sharply beyond 10 sessions. Notably, about one in six teachers indicated never conducting outdoor lessons (Figure 1).

3.2. RQ2: How Do Specific Teacher-Level Predictors Associate with Frequency of Outdoor Teaching?

Predictors included inclusion with nature, professional development, and a block of Likert-type items. Inclusion with nature was measured with the seven-point inclusion of nature in self (INS) pictorial scale. Professional development was assessed by self-reported participation in outdoor education seminars or training, rated on a five-point descriptive scale ranging from never (1) to more than three times (5).

In addition, 65 five-point Likert-type items were included. Six of these were excluded due to low factor loadings (<0.25) or weak correlations (<|0.30|), leaving 59 items for analysis. Six reverse-worded items were recoded (R) for reliability analysis, although five were later excluded for the reasons noted above. The remaining items were treated as indicators for latent constructs.

All predictor variables were converted to standardized values (z-scores) to facilitate comparability across scales and variable types (Petz et al., 2012). The final set of 59 Likert-type items was subjected to exploratory factor analysis (EFA) using principal axis factoring (PAF) with oblique (promax) rotation, given the assumption of correlated factors and that this approach would provide the best fit to the data (n = 191). The sample size was reduced from 258 to 191 due to a listwise deletion of cases with missing responses across the items included in the factor analysis. The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy was 0.827, indicating meritorious adequacy, and Bartlett’s test of sphericity (χ²(1711) = 7220.19, p < 0.001) confirmed that the correlation matrix did not form an identity matrix, supporting its suitability for factor analysis. Explained variance and scree plot inspection were then used to determine the number of factors to retain.

Based on the total variance explained, Cattell’s scree plot, and its inflection point, a four-factor solution was retained, accounting for 46.73% of the variance. A threshold of 5% explained variance was additionally applied as a secondary criterion to support interpretability.

The six omitted items were TA11, TA12, TA13, TA19, TA20, and PB3 (see

Table A5. The six omitted items TA11, TA12, TA13, TA19, TA20, and PB3.

TA11: During outdoor classes, students require more supervision to stay focused on learning.

TA12: Compared to classroom teaching, outdoor science teaching requires more preparation.

TA13: I find it more challenging to teach science outdoors than in the classroom.

TA19: Less content is covered in outdoor science classes compared to classroom teaching.

TA20: Outdoor science teaching is more exhausting for the teacher than classroom teaching.

PB3: If I could choose, I would prefer living in the countryside over the city, because I can have more contact with nature there.

in Appendix A). The factors reflected the predictors outlined in the research questions and were labeled teachers’ attitudes, support, self-efficacy, and connection to nature (with personal background). The factor analysis largely confirmed the theory-based grouping of items developed from the literature, indicating that the questionnaire’s structure was consistent with the intended predictors. Reliability was assessed using Cronbach’s alpha and McDonald’s omega, with all coefficients exceeding 0.90, indicating excellent internal consistency (see

Table A1. Items for connection to nature (CNI) (with personal background (PB)) with descriptive statistics and reliability coefficients (Cronbach’s alpha and McDonald’s omega).

Factor	Statement (Scale)	N	M	SD	Mo	Cronbach’s α	McDonald’s ω
Connection to nature (CNI) (with Personal Background (PB))	CNI8: I like to see wild animals living in a clean environment.	206	4.58	0.551	5	0.906	0.902
	CNI10: Taking care of animals is important to me.	205	4.54	0.606	5
	CNI4: I like to see wild flowers in nature.	206	4.49	0.590	5
	CNI7: I like to garden.	205	4.48	0.638	5
	CNI3: I like to hear different sounds in nature.	206	4.42	0.611	5
	CNI2: I feel sad when wild animals are hurt.	206	4.44	0.612	5
	CNI1: Collecting rocks and shells is fun.	206	4.39	0.572	4
	CNI9: I enjoy touching animals and plants.	206	4.24	0.717	4
	CNI13: Being outdoors makes me happy.	205	4.84	0.418	5
	CNI6: Being in the natural environment makes me feel peaceful.	206	3.90	0.950	4
	CNI11: Humans are part of the natural world.	205	4.89	0.340	5
	CNI5: When I feel sad, I like to go outside and enjoy nature.	205	3.98	0.982	4
	CNI12: People cannot live without plants and animals.	205	4.92	0.303	5
	CNI15: Picking up trash on the ground can help the environment.	205	4.68	0.581	5
	CNI14: My actions will make the natural world different.	205	4.45	0.782	5
	CNI16: People do not have the right to change the natural environment.	205	4.40	0.789	5
	PB1: As a child, I used to spend a lot of time in nature.	205	4.30	0.745	5
	PB2: I spend a lot of time in nature.	205	4.58	0.693	5

Note: N = number of respondents; M = mean; SD = standard deviation; and M_o = mode.

,

Table A2. Items for teachers’ attitudes (TA) with descriptive statistics and reliability coefficients (Cronbach’s alpha and McDonald’s omega).

Factor	Statement (Scale)	N	M	SD	Mo	Cronbach’s α	McDonald’s ω
Teachers’ Attitudes (TA)	TA8: Students learn more outdoors than in the classroom.	226	3.51	0.910	4	0.916	0.917
	TA16: I believe that outdoor science teaching should be a mandatory part of the school curriculum.	216	3.66	1.146	4
	TA5: Outdoor science classes encourage students’ curiosity.	227	4.11	0.779	4
	TA15: Students are more motivated to learn during outdoor science classes.	217	3.67	0.834	4
	TA6: Outdoor science classes enhance students’ innovation in solving scientific problems.	226	3.89	0.820	4
	TA3: Outdoor science classes contribute to students’ stronger connection with nature.	227	4.28	0.745	4
	TA4: Outdoor science classes strengthen the teacher–student relationship.	227	3.72	0.820	4
	TA1: Outdoor science classes improve collaboration among students.	227	3.87	0.896	4
	TA17: I wish I could teach science outdoors more often.	215	4.06	0.918	4
	TA7: Students are more focused during outdoor science classes.	225	2.93	0.970	3
	TA2: Outdoor science classes reduce behavioral problems in students.	227	3.10	1.039	3
	TA14: Outdoor science teaching increases students’ interest in the subject.	217	4.03	0.841	4
	TA10: In a real environment (e.g., a meadow), certain principles of nature are easier to demonstrate than in a classroom.	227	4.19	0.795	4
	TA9: Students look forward to outdoor science classes more than indoor ones.	227	4.37	0.731	5
	TA18: I prefer teaching science indoors to teaching it outdoors (R).	214	2.86	0.941	3

Note: N = number of respondents; M = mean; SD = standard deviation; M_o = mode; and (R) = reversed coding employed.

,

Table A3. Items for support in outdoor settings (S) with descriptive statistics and reliability coefficients (Cronbach’s alpha and McDonald’s omega).

Factor	Statement (Scale)	N	M	SD	Mo	Cronbach’s α	McDonald’s ω
Support (S)	S14: Available funding for outdoor science teaching equipment.	207	4.33	0.852	5	0.912	0.909
	S13: Cooperation with external institutions (e.g., borrowing equipment, help with activity preparation).	208	4.14	0.909	5
	S12: Training for unexpected situations (e.g., injuries, behavioral issues).	208	3.91	1.032	5
	S11: Availability of additional professional development possibilities for teachers.	206	4.09	0.928	4
	S4: Acknowledgment of additional preparation time for outdoor lessons.	208	3.98	1.090	5
	S6: Flexible scheduling (e.g., easy rescheduling, possibility of teaching multiple periods in a row).	208	4.45	0.838	5
	S9: Possibility of student assessment in this form of instruction.	208	3.95	1.013	5
	S8: Clear pedagogical guidelines for comprehensive outdoor science teaching.	208	4.07	0.951	5
	S10: Prepared activities and materials intended for outdoor science teaching (worksheets, lesson plans, manuals).	208	4.29	0.859	5
	S15: Easy access to natural areas (e.g., lawns, gardens, parks, ponds).	208	4.38	0.770	5
	S3: Praise or reward for innovative approaches.	207	3.43	1.120	3
	S5: Option to divide the class into smaller groups (e.g., like in technology/home economics or ability grouping).	208	4.54	0.785	5
	S17: Precautions for student safety.	207	4.35	0.760	5
	S7: Curriculum that supports outdoor teaching.	208	3.96	1.030	5
	S2: Support from colleagues (willingness to accompany the group, help with activity preparations).	208	3.88	0.985	4
	S16: Favorable weather conditions.	207	4.09	1.001	5
	S1: Support from school leadership (encouragement for outdoor teaching).	206	3.52	1.062	4

Note: N = number of respondents; M = mean; SD = standard deviation; and M_o = mode.

and

Table A4. Items for self-efficacy (SE) with descriptive statistics and reliability coefficients (Cronbach’s alpha and McDonald’s omega).

Factor	Statement (Scale)	N	M	SD	Mo	Cronbach’s α	McDonald’s ω
Self-efficacy (SE)	SE3: I am familiar with teaching methods and strategies that are appropriate for outdoor instruction.	206	3.88	0.855	4	0.913	0.915
	SE4: I can adapt science content for outdoor teaching.	206	4.02	0.777	4
	SE5: I feel capable of handling challenges related to outdoor teaching (e.g., weather, safety, logistics).	206	3.79	0.891	4
	SE2: I believe that planning and organizing outdoor science lessons is (or would be) manageable for me.	206	3.98	0.817	4
	SE9: I know how to use outdoor spaces and materials from the environment for teaching.	206	4.02	0.820	4
	SE6: I know how to motivate students to actively and constructively participate in outdoor science activities.	206	3.93	0.739	4
	SE7: I can achieve many science subject objectives through outdoor instruction.	206	3.78	0.814	4
	SE1: I have a good understanding of science content that is suitable for outdoor teaching.	206	4.27	0.680	4
	SE8: I feel physically fit enough to conduct outdoor science lessons.	205	4.06	0.903	4

Note: N = number of respondents; M = mean; SD = standard deviation; and M_o = mode.

in Appendix A).

After extracting four factors, the composite variables were computed by averaging the scores of their component items. For inclusion with nature and professional development, the original values were transformed into standardized ones as described above. These predictors were then tested for their ability to explain the frequency of outdoor science teaching. Given the ordinal nature of the dependent variable—with ordered categories but unequal intervals—ordinal logistic regression was applied.

The overall model was statistically significant (χ²(7) = 32.56, p < 0.001), indicating that the set of predictors effectively distinguished between the levels of outdoor teaching frequency. A goodness-of-fit test (Deviance, χ²(1212) = 633.786, p > 0.05) suggested an adequate model fit, as the predicted and observed values did not differ significantly. Pseudo-R² statistics (Cox & Snell = 0.148; Nagelkerke = 0.153) indicated that approximately 15% of the variance in teaching frequency was explained by the predictors, reflecting modest explanatory power and suggesting that additional factors that were not included in the model may also influence outdoor teaching practices. Finally, the test of parallel lines (χ²(30) = 4.87, p = 0.999) indicated that the proportional odds assumption was not violated; however, the exceptionally high p-value warrants cautious interpretation, as it may reflect the limited sensitivity of the test, potentially indicating some degree of underfitting.

From the perspective of individual predictors shown in

Table 1. Ordinal logistic regression results for six predictors of teachers’ frequency of outdoor science teaching.

						95% CI
Predictor	B	SE	Wald	df.	p	Lower Bound	Upper Bound
Connection to Nature	−0.265	0.234	1.275	1	0.259	−0.724	0.195
Teachers’ Attitudes	0.697	0.212	10.795	1	0.001	0.281	1.113
Support	−0.103	0.223	0.213	1	0.644	−0.540	0.334
Self-efficacy	0.538	0.195	7.600	1	0.006	0.156	0.921
Inclusion with nature	0.161	0.137	1.389	1	0.239	−0.107	0.429
Professional Development	0.106	0.128	0.685	1	0.408	−0.145	0.357

Note: B = regression coefficient; SE = standard error; Wald = Wald chi-square statistic; df = degrees of freedom; p = significance level; CI = confidence interval (95%).

, the most significant is Teacher’s Attitudes (B = 0.697; SE = 0.212; Wald = 10.795, p = 0.001), which indicates that as attitudes turn more favorable, outdoor teaching sessions become more prevalent throughout the school year. Self-efficacy was also found to be statistically significant (B = 0.538; SE = 0.195; Wald = 7.600, p = 0.006) and considering the positive value of the coefficient, can be interpreted in such a way that as their self-assessment improves, the frequency of outdoor teaching increases. Other predictors were not statistically significant in this model.

3.3. Further Exploration of Significant Teacher-Level Predictors

In the regression model, several potential predictors of outdoor science teaching frequency were examined. Among these, two variables—teachers’ attitudes and self-efficacy—emerged as statistically significant. To better understand their influence, we further examined these predictors separately. In addition, descriptive and correlational analyses of environmental accessibility were included to provide contextual insight into how the availability of suitable outdoor settings may interact with these teacher-related factors.

3.4. Teacher’s Attitudes

Bivariate correlations using Spearman’s rank correlation coefficient (

Table 2. Spearman’s rank correlation coefficient of teacher’s attitudes with four selected predictors.

		Support	Connection to Nature	Inclusion with Nature	Professional Development
Teachers’ Attitudes	ρ [CI]	0.265 [0.13; 0.39]	0.259 [0.12; 0.38]	0.136 [−0.01; 0.27]	0.030 [−0.10; 0.16]
	p	<0.001	<0.001	0.053	0.653
	N	208	206	204	227

Note: ρ = Spearman’s rank correlation coefficient; CI = 95% confidence interval; p = significance level; and N = sample size. Confidence intervals are shown in brackets.

) indicate that the relationship between teachers’ attitudes and other predictors is significant in the case of support (p < 0.001, ρ = 0.265) and connection to nature (p < 0.001, ρ = 0.259); however, albeit positive, the strength is rather weak (ρ < 0.30).

Inclusion of nature was measured with the seven-point (A–G) inclusion of nature in self (INS) pictorial scale, which depicts overlapping Venn diagrams to capture the perceived human–nature relationship (Schultz, 2002). Responses were skewed toward the high end: 76.0% fell in E–G, none in A or B, 2.9% in C and 21.1% in D; the modal category was E (31.9%), indicating generally high inclusion with nature. However, its association with teachers’ attitudes was not statistically significant (ρ = 0.136, p = 0.053).

Professional development participation was low: 56.6% reported no outdoor (science)-related training. Among the remainder, 15.6% attended one session, 11.0% two, 2.5% three, and 14.3% more than three. Professional development was not associated with teachers’ attitudes (ρ = 0.030, p = 0.653). Frequently cited courses included workshops at the annual science conference, school-based study groups, CŠOD field programs, and Erasmus courses.

3.5. Self-Efficacy

There was no statistically significant correlation between teaching experience (measured in years) and self-efficacy (ρ = 0.088, p = 0.210).

3.6. Access to Outdoor Environments

Although the regression model focused on teacher-related psychological and experiential predictors, a supplementary analysis explored the role of environmental accessibility using Spearman’s rank-order correlation. While teachers’ perceptions of the importance of suitable environments were already captured within the support factor, the actual availability of specific outdoor sites represents a distinct contextual condition, rather than a teacher-level attitude, and was therefore examined separately. The number of suitable outdoor settings available to teachers was positively associated with the frequency of outdoor science teaching (ρ = 0.314, p < 0.001). Teachers with access to a greater diversity of nearby outdoor sites tended to conduct outdoor lessons more often.

The most commonly reported accessible settings included lawns or meadows (70.5%), forests (51.6%), concrete playgrounds (51.6%), school gardens (42.2%), and aquatic ecosystems (41.9%). Less frequently mentioned were raised or flower beds (29.1%), beehives (22.8%), and parks (19.8%). A small proportion of respondents (1.9%) also described other available spaces such as olive tree groves, fruit orchards, or purposefully designed outdoor classrooms equipped with benches and tables. Notably, 4.7% of teachers reported having no access to any suitable outdoor teaching environments.

Among those who reported conducting outdoor science lessons at least once (221 teachers), the most frequently used environments were lawns or meadows (75.5%) forests (65.2%), and school gardens (46.2%). Other commonly used spaces included aquatic ecosystems (41.3%), concrete playgrounds (29.9%), and raised or flower beds (26.1%). Less frequently used settings included parks (21.2%) and beehives (17.4%). A small proportion of teachers (3.3%) reported using other natural or semi-natural settings, such as olive tree groves, fruit tree groves, sea shores, botanical garden and learning trails.

3.7. Teaching Experience and Outdoor Teaching Frequency

Teaching experience was analyzed separately as a background variable and was not included in the regression model. The regression analysis was conceptually limited to the set of teacher-level psychological predictors. Teaching experience functions as a demographic descriptor, so it was examined descriptively and through correlation for contextual insight. A Spearman’s rank-order correlation showed no significant association between years of experience and the frequency of outdoor science teaching (ρ = 0.052, p = 0.415), indicating that years of teaching experience were not related to how often teachers taught outdoors.

4. Discussion

This study examined how frequently teachers conduct science lessons outdoors (RQ1) and which teacher-level predictors are most strongly associated with this practice (RQ2).

Overall, outdoor science teaching occurred with moderate frequency, suggesting that while many teachers are positively disposed toward outdoor learning, regular implementation remains uneven. In addressing multiple teacher-level predictors within a single analytical framework, this study extends previous research that has often focused on isolated factors or predominantly external barriers. The results suggest that internal teacher characteristics, particularly attitudes and self-efficacy, are more strongly associated with outdoor teaching frequency than structural constraints.

In the following sections, these findings are discussed in greater detail.

4.1. Attitudes and Self-Efficacy as Key Drivers of Outdoor Science Teaching

The findings indicate that psychological factors—particularly teachers’ attitudes and self-efficacy—play a more decisive role than structural or experiential variables in explaining variation in outdoor science teaching.

Teachers’ attitudes toward outdoor learning emerged as the strongest predictor of how frequently science lessons were conducted outdoors. This finding aligns with prior research emphasizing the central role of teachers’ beliefs, values, and pedagogical orientations in shaping instructional practice (Rickinson et al., 2004; Waite, 2011). Many studies identified teachers’ confidence and attitudes as recurrent enablers or barriers, often outweighing structural constraints such as time or resources. Even when schools provide access to outdoor spaces or administrative support, outdoor teaching ultimately depends on whether teachers perceive it as educationally meaningful and worth the associated effort (Dillon et al., 2006; Rickinson et al., 2004).

Although teachers in our sample generally expressed favorable attitudes—mean scores for most items exceeded the scale midpoint—variation within this positive range still mattered. Teachers with the strongest and most deeply internalized attitudes were those who most consistently implemented outdoor science teaching. Another recent study among Slovenian lower secondary science and biology teachers likewise reported broadly positive perceptions of outdoor education, suggesting that favorable evaluations of outdoor learning are widespread within the national context (Špernjak et al., 2026). This supports attitude–behavior theories suggesting that the strength of an attitude, rather than its mere presence, determines whether intentions translate into sustained action (Ajzen, 1991; Kaiser et al., 1999). Similar patterns have been observed in science education, where positive beliefs about inquiry-based learning are associated with greater use of experiential and place-based approaches (Carrier et al., 2014). Thus, attitudes appear to function as a motivational filter through which teachers evaluate whether outdoor teaching is instructionally legitimate and compatible with their everyday classroom practice.

Teachers’ self-efficacy also emerged as a significant predictor of outdoor science teaching. Self-efficacy reflects teachers’ beliefs in their ability to plan, manage, and successfully deliver instruction (Bandura, 1986). Extensive research has shown that higher self-efficacy is associated with greater openness to innovation, persistence in the face of challenges, and instructional flexibility (Jerrim et al., 2025; Schumann & Sibthorp, 2013; Skaalvik & Skaalvik, 2010; Tschannen-Moran & Hoy, 2001).

Within science education, self-efficacy has been operationalized through instruments such as the Science Teaching Efficacy Belief Instrument (STEBI), which distinguishes between personal teaching efficacy and outcome expectancy (Riggs & Enochs, 1990). Studies using this instrument show that teachers who feel capable of teaching science effectively are more inclined to engage students in inquiry-based and experiential learning (Deehan, 2017). These results suggest that self-efficacy functions as a key motivational resource, supporting teachers in translating pedagogical intentions into consistent outdoor science practice.

The findings underscore the importance of promoting high-quality professional development for both pre-service and in-service teachers, with a particular emphasis on outdoor education programs that intentionally build teachers’ positive attitudes and self-efficacy. Such programs should be intensive (Holden et al., 2011) and experiential in nature (Barrable et al., 2020), as short-term or isolated training experiences rarely lead to sustained improvements in teaching practice (Moseley et al., 2002).

Described recommendations are logical, but often not feasible for in-service teachers, for whom engaging in intensive outdoor education programs is difficult with all their school and family obligations. A sensible strategy would be to support science teachers by partnering them with specialized outdoor education practitioners—such as environmental education experts in nature parks, geoparks, bird-watching associations, and botanical gardens. These specialists could work alongside teachers in delivering outdoor education programs, gradually building teachers’ competencies for teaching in outdoor environments. This collaborative model mirrors the practice of “nature schools” in Finland, where outdoor education experts support local schools in implementing learning activities in nature and fostering teachers’ confidence and skills in outdoor pedagogy. Such programs should explicitly target both “what” and “how” subject-specific competencies (e.g., in science, ecology, and environmental sciences) and the practical skills required to plan, implement, and evaluate outdoor teaching (Sjöblom & Svens, 2019).

One effective way to promote teachers’ positive attitudes toward outdoor education and strengthen their self-efficacy is to engage pre-service science teachers systematically and longitudinally in outdoor education programs. Although outdoor education is widely regarded by practitioners as an important if not essential component of higher-education curricula, including science teacher education, there is strong evidence that fieldwork experiences have been declining in many countries, and this has been attributed to a range of academic, financial, and societal pressures (Fleischner et al., 2017; Scott et al., 2012).

Previous analyses indicate that the frequency with which teachers implement outdoor education is most strongly associated with their attitudes toward outdoor teaching and their sense of self-efficacy. These core psychological determinants do not develop in isolation, but may relate—sometimes only weakly—to a range of supporting factors. To further clarify the mechanisms underlying the regression model, correlations were performed for those two strongest predictors. This analysis aimed to explore how these key psychological variables relate to other teacher-level factors, such as connection to nature, perceived support, teaching experience, and prior training (professional development). A weak yet statistically significant association emerged between connection to nature and attitudes toward outdoor teaching. Teachers with a stronger sense of environmental connectedness tended to express slightly more positive attitudes, although the effect was modest. This finding is consistent with earlier research suggesting that emotional or affective affinity with the natural world, while important, rarely translates directly into sustained pedagogical change without complementary efficacy beliefs and systemic support (Ernst & Tornabene, 2012). Thus, valuing nature may foster openness toward outdoor teaching, but it does not automatically translate into practice (Dring et al., 2020).

A similarly weak positive relationship was observed between perceived support and teachers’ attitudes. This pattern suggests that while structural or administrative encouragement can facilitate outdoor science teaching, its influence remains limited unless teachers themselves possess the motivation and agency to explore available opportunities. Prior studies have noted the same conditional nature of institutional enablers—schools may offer access or flexibility, but uptake depends largely on teachers’ self-perceived competence and internal commitment (Waite, 2011, 2022).

Surprisingly, no meaningful correlations were found between professional development and teachers’ attitudes. In our sample, most respondents reported attending only short courses or occasional lectures, rather than extended training programs, which likely limited the depth of impact. This pattern aligns with longstanding evidence that brief or isolated professional development opportunities are insufficient to change instructional beliefs or confidence. Effective professional growth in science and environmental education typically requires longitudinal and sustained engagement supported by mentoring and collaborative reflection (Desimone & Garet, 2015; Luft & Hewson, 2014).

Finally, teaching experience showed no association with self-efficacy. This finding underscores that confidence in outdoor teaching does not simply accumulate with years of service. Rather, as Bandura (1997) conceptualized, self-efficacy is built through mastery experiences—through successful practice and positive feedback that reinforce teachers’ belief in their capability to act effectively in novel settings. Consequently, outdoor teaching competence is not acquired passively through years of service, but through specific teaching experiences and reflective practice. This highlights the importance of sustained, practice-oriented professional learning opportunities that allow teachers to plan, implement, and reflect on outdoor lessons. The relatively low levels of self-efficacy observed among teachers may be linked to their limited participation in professional development. Opportunities for mastery experiences, observing other teachers, and receiving feedback are critical for developing confidence. Teachers who engage less in professional development may therefore have fewer of these reinforcing experiences, which helps to explain the observed variation in self-efficacy.

Sustained change depends less on structural provisions alone and more on fostering self-efficacy through guided practice, long-term professional learning, and school cultures that value teaching beyond the classroom. This interpretation is consistent with broader evidence that intrinsic motivation and the personal value teachers assign to innovation are stronger predictors of behavioral change than external incentives or administrative mandates (Kottmann et al., 2024). To support these processes, professional development should focus on reflective, mastery-based learning that strengthens both self-efficacy and pedagogical conviction. At the same time, schools should ensure that safe and easily accessible outdoor spaces are available for regular use. Coordinating these psychological and structural supports may offer the most effective route to increasing both the frequency and quality of outdoor science learning.

4.2. Accessibility of Outdoor Learning Spaces

Access to suitable outdoor learning environments in the immediate vicinity of schools is widely regarded as a practical prerequisite for the regular implementation of outdoor science education. Proximity to natural or semi-natural spaces reduces logistical barriers, lowers time demands, and increases the feasibility of integrating outdoor activities into everyday teaching (Ayotte-Beaudet et al., 2025; Yamanoi et al., 2021). To examine whether such physical opportunities are associated with actual teaching practice, the relationship between the availability of accessible outdoor environments and the frequency of outdoor science teaching was explored. The variable accessible environments—representing the number of nearby outdoor areas—was analyzed separately to capture the physical opportunity.

A weak but positive association emerged, suggesting that easier access to outdoor spaces may facilitate more frequent teaching outside the classroom. However, both the present findings and previous research indicate that environmental access alone is insufficient to ensure sustained outdoor teaching. The recent Slovenian study already cited above likewise reported that many teachers have access to suitable natural areas near their schools, yet outdoor education remained less frequently implemented in everyday classroom practice.

Whereas that study primarily documented patterns of practice and perceived obstacles, their findings suggest that environmental availability alone does not explain how frequently outdoor teaching occurs (Špernjak et al., 2026). Instead, outdoor science teaching appears to emerge from the interaction between environmental affordances and institutional conditions: supportive school leadership legitimizes outdoor learning, while the proximity of suitable sites makes such support feasible in everyday practice (Beames et al., 2012; Dillon & Dickie, 2012; Waite, 2022).

Viewed in this way, environmental access should be understood as part of a broader educational ecosystem in which outdoor science teaching is embedded. Beyond its immediate pedagogical benefits, outdoor science education is increasingly recognized as a key component of environmental and sustainability-oriented education, reflecting wider European policy efforts to promote creativity, innovation, and responsible engagement with socio-ecological challenges (European Commission, 2021). From this perspective, the successful integration of outdoor spaces into science teaching depends on collective and coordinated engagement among teachers, school leaders, researchers, and policymakers (Bianchi et al., 2022).

While the present data did not confirm a direct effect of the support or environmental accessibility, this pattern is consistent with the ecological affordance perspective (Gibson, 1979; Nicol, 2014), which views outdoor teaching as the outcome of interacting personal, institutional, and environmental conditions. In this view, access to natural spaces represents only one dimension of opportunity—it enables but does not determine engagement. Similar conclusions have been drawn in research on children’s experiences of nature, where access alone did not guarantee meaningful interaction; instead, social and institutional contexts mediated how natural environments were used and valued (Adams & Savahl, 2017; Nicol, 2014).

4.3. Limitations and Implication for Further Research and Stakeholders

This study has several limitations that should be acknowledged. First, the use of a convenience sample may have introduced self-selection bias, potentially overrepresenting teachers with a prior interest in outdoor teaching. In addition, the sample size was relatively modest, which may limit the generalizability of the findings. Moreover, the length of the questionnaire may have contributed to the dropout and non-response bias.

Second, although the constructs demonstrated high internal consistency, some were refined based on exploratory factor analysis, which may limit the comparability with previous studies. Furthermore, while the statistical model met the required assumptions, its ability to detect more subtle patterns may be limited, meaning that some relationships could remain unobserved.

Third, the findings are based on quantitative modeling and therefore reflect general patterns rather than the full complexity of real classroom contexts. While these results provide insight into relationships among teacher-level factors, they may not fully capture the practical conditions in which teaching decisions are made.

Finally, although this study identifies the key predictors of outdoor science teaching, the findings reflect patterns derived from quantitative modeling. While these patterns provide insight into relationships among teacher-level factors, they might not fully account for the practical context in which teaching decisions are made. Because of the nature of the study, we cannot definitively conclude that positive attitudes and high confidence directly cause sustained outdoor teaching; other unmeasured factors may also contribute. Future research could therefore invite teachers to act as an expert group, reflecting on whether the identified patterns align with their professional experience and highlighting additional predictors—such as school culture, professional norms, perceived risks and challenges, or prior mastery experiences—that are not readily captured through our quantitative data.

Despite these limitations, several implications can be drawn for different stakeholder groups. As our findings suggest, sustained outdoor science teaching is primarily driven by internal teacher factors, particularly self-efficacy and attitudes, rather than structural support alone. For practicing teachers, this highlights the importance of strengthening confidence through reflective practice and collaborative professional learning to support implementation. For school leaders, the results indicate that structural provision is necessary but insufficient without a school culture that encourages experimentation, creates outdoor opportunities and actively fosters teacher confidence. For policymakers and teacher education programs, the findings emphasize the importance of embedding outdoor teaching systematically into pre-service and in-service training professional development programs, with a focus on developing self-efficacy through practical, mastery-based experiences.

5. Conclusions

Taken together, the results of our study indicate that sustained outdoor science practice is most likely when teachers hold strong positive attitudes toward outdoor science teaching and high confidence in their ability to teach outdoors. This aligns with evidence that internal drivers shape instructional behavior more reliably than external incentives or mandates (Kottmann et al., 2024). Once this psychological readiness is established, institutional support—through accessible outdoor spaces, practical assistance, and encouragement from colleagues and leadership—can effectively reduce the remaining barriers, helping outdoor teaching to evolve from an occasional enrichment activity into a regular and integrated part of science instruction.

In contrast to the fragmented nature of previous research on teacher-level predictors, our study integrates multiple diverse predictors within a single model and compares their relative importance in shaping decisions for outdoor science teaching, thereby providing a more comprehensive understanding of the determinants of outdoor teaching, with implications for teachers, school leaders, and policy development.