Impact of outdoor place-based learning on elementary school students’ ability to make unsolicited observations about living organisms over time

ABSTRACT

Our research objective was to investigate whether a place-based educational intervention for elementary school students leads to a better ability to make unsolicited observations about living organisms. We tested the following prediction: Elementary school students who learn about living organisms outdoors in a familiar environment will be able to make more unsolicited observations about living organisms one week after the end of the intervention compared to those two months after the intervention ends. Our study was conducted in the province of Québec, Canada, which resulted in 116 students experiencing the same educational intervention from home. Based on individual telephone interviews, we compared elementary school students’ ability to make unsolicited observations about living organisms one week (Group A, n = 58) and two months (Group B, n = 58) after a place-based educational intervention in their home environment. We developed three emergent categories to analyse students’ place-based unsolicited observations about living organisms. We found that elementary school students who learned about living organisms in their home environment made more unsolicited observations about living organisms leading to factual descriptions two months after the end of the learning experience than those who were interviewed one week after the intervention.

KEYWORDS:

• Biology
• elementary
• observation
• outdoor education
• place-based learning

Introduction

The lack of children’s exposure to nature has been widely documented in recent years, especially since the publication of Louv’s (2005) Last Child in the Woods. In a systematic scoping review, Oswald et al. (2020) found that exposure to nature declined as students engaged more in visual screen-based technologies. Moreover, the COVID-19 lockdown affected outdoor contact with nature, sometimes increasing it (Friedman et al. 2022), but in other cases decreasing it (Moore et al. 2020). Thus, it seems that children should be more exposed to nature since it has been tied with academic achievement, personal development and stewardship (Kuo, Barnes, and Jordan 2019), but that such contact has declined in recent years.

Place-based learning in science education

Place-based learning in science seems to be a valuable avenue for fostering more contact with nature, because it can provide a great opportunity for students to explore and learn about the natural world in a meaningful and contextualised way. As an approach to teaching and learning, place-based learning connects students’ educational experiences with the local environment, community, and culture (Gruenewald 2003). A place can be ‘any area within the local community that supports student learning about their worlds’ (McVittie et al. 2020). Place-based learning is grounded in a situated learning perspective, which recognises that learning is enabled and constrained by contexts (Sadler 2009). And, the context in which learning is situated influences its process (Giamellaro 2017). Through place-based learning, students can gain a deeper understanding of science and cultural understanding within their local environment (Gruenewald 2003), such as through citizen science (Kermish-Allen, Peterman, and Bevc 2019).

Place-based learning can support several intentions associated with science education.

1. Building awareness of the place of nature and living organisms. In 1925, in an article published in The Elementary School Journal, Moseley wrote that students should go out to enjoy ‘the beauty of nature, and to learn something of her secrets’ (p. 38). As this example shows, connecting students to nature isn’t a new idea in science education. The argument can be made that outdoors is particularly conducive to learning associated with the natural sciences (Guerrero and Reiss 2020; Stapleton and Reif 2022), because no matter where you are, there are always natural phenomena to discover and understand (Slingsby 2006). Through direct contact, students are more likely to develop a stronger connection with nature and a deeper understanding of environmental issues (Kuo, Barnes, and Jordan 2019).

2. Increasing achievement in science. Scientific literature has shown that school-based outdoor science education can be beneficial for students’ learning (Faber Taylor, Butts-Wilmsmeyer, and Jordan 2022; Finn, Yan, and McInnis 2018; Guerrero and Reiss 2020). According to Kuo, Barnes, and Jordan (2019), educational interventions that expose students to nature lead to better retention of knowledge, higher scores on standardised tests, and better grades. As an enhanced context strategy, it has the potential to be supportive of student achievement (Arıkan 2021; Johnson et al. 2020; Schroeder et al. 2007) for a broad range of students (Faber Taylor, Butts-Wilmsmeyer, and Jordan 2022), including for children from disadvantaged families (Camasso and Jagannathan 2018; Faber Taylor, Butts-Wilmsmeyer, and Jordan 2022).

3. Providing more authentic learning experiences. An outdoor place-based approach should be considered as providing a third space for science education at school, that is, a space that bridges school practices and students’ lives (Finn, Yan, and McInnis 2018; Stapleton and Reif 2022). Schools’ immediate surroundings can be used as an extension of the classroom to make abstract knowledge concrete through direct observations (Ayotte-Beaudet et al. 2023b; Moseley 1925). For instance, students can engage in science field activities such as soil analysis, sky observations, and species identification.

4. Enabling active and experiential learning. The idea of place-based learning is frequently associated with experiential learning. In his famous book Experience and Education, Dewey (1938, p. 40) wrote: ‘The teacher should become intimately acquainted with the conditions of the local community, physical, historical, economic, occupational, etc., in order to utilize them as educational resources’. Later, Kolb (1984) described experiential learning as a way for students to learn directly through hands-on experience. With outdoor place-based learning, students can engage in scientific investigations such as doing scientific observations, asking questions, or conducting experiments (Ayotte-Beaudet et al. 2023b).

School-based outdoor learning

Place-based learning practices and purposes are connected with outdoor education, both taking pace outside of traditional classrooms (Gruenewald 2003). Drawing on characteristics described by Ayotte-Beaudet et al. (2023a), school-based outdoor learning refers to any learning experience taking place outdoors, in any type of environment, geared towards learning in accordance with pedagogical intentions defined by the teacher. While many intentions can be associated with outdoor education, such as increasing physical activity (Coventry et al. 2021; Finn, Yan, and McInnis 2018; Lacoste et al. 2021) or promoting mental health and well-being (Bølling et al. 2019; Gustafsson et al. 2012), two are directly related to place-based learning. Outdoor school-based learning has the potential to connect children with nature (Kuo, Barnes, and Jordan 2019). With an approach integrating nature in the outdoors, both natural environment and elements can become a context for academic learning (Jordan, Sajady, and Faber Taylor 2022). In science education in particular, Faber Taylor et al. (2022) suggested that outdoor nature-based learning, depending on the science learning outcomes targeted, could be more supportive of science learning than classroom-based instruction ‘for a broader range of students’ (p. 1527). Outdoor school-based education also allows teachers to use a meaningful science learning context for students (Amos and Reiss 2012; Guerrero and Reiss 2020). Using a school’s immediate surroundings gives more opportunities for authentic science since students are familiar with these environments and they can collect real data (Glackin 2016). These settings could facilitate the contextualisation process in science; that is, the linkage between learning and the context in which it occurs (Giamellaro 2017).

Methodological challenges can arise when conducting research about school-based outdoor learning. A first methodological challenge of outdoor school-based learning in science is that the learning context is not academic, but rooted in a specific place. Although scientific knowledge and skills are involved, their application remains unique to that setting (Rivera Maulucci et al. 2014). Therefore, standardised assessments and data collection strategies normally used to measure academic learning that occurs in classrooms cannot easily be used to measure the same outcomes of an outdoor learning intervention (Carrier, Tugurian, and Thomson 2013). Data collection instruments that are adapted to this environment are generally required (Braun and Dierkes 2019). For example, it would not be coherent to assess student learning via responses to a questionnaire or a paper-and-pencil knowledge test if the intention of an outdoor period is to develop skills that are applied in a context-specific way. A second methodological challenge is control of variables with students from different schools, and consequently with different outdoor environments. Since variability necessarily exists in the outdoor settings used by the participating students, they are likely to produce a different effect on each student (Norwood, Lakhani, and Kendall 2021) who can have prior interest in, or knowledge of, these places. These methodological challenges show that particular attention should be paid to the study designs and the development of adapted data collection instruments to test new hypotheses or predictions associated with the persistence of learning that is acquired outdoors.

Relationships in ecosystems

School-based outdoor learning provides a unique opportunity to study the relationships within ecosystems, as nature itself can be explored (Slingsby 2006). Places explored by students provide added value for studying the mechanisms of interactions between species, which often present challenges for students (Ben-Zvi Assaraf and Orion 2009; Bermudez et al. 2017) and for teachers (Sweeney and Sterman 2007). Outdoor settings can help students to generate scientific observations and explanations of ecosystems, especially since many students have difficulty recognising how animals interact with plants (Zangori and Forbes 2015).

Studies have examined the effects of outdoor science on learning about relationships in ecosystems as well as utilising scientific inquiry as a strategy for science learning. As an example, in a paper exploring ‘the impacts of a contextualised outdoor science curriculum on what and how elementary students learn when immersed in the local contexts in which the targeted natural phenomena occur’ (Ayotte-Beaudet et al. 2023b, p. 280), students did an urban citizen science project outdoors (which involved observation). In this study, small artificial prey (clay caterpillars) were first installed by students in trees, then collected a week later to observe them for marks that indicate predation. From there, 63 fifth and sixth grade students’ (aged 10 to 12) interviews enabled three categories of impacts on what students learned to be identified: ‘evolution of conceptual understanding about living organisms’, ‘development of scientific investigation abilities’, and ‘evolution of connection to nature’. Two categories of impacts on how students learn also emerged: school-based outdoor learning offered ‘a context that encourages deeper learning’ and ‘a context that promotes engagement’. Based on their findings, this learning context seems to promote more in-depth learning. As Kapon, Laherto, and Levrini (2018) found, students may not deepen their learning much when learning appears to contribute to their school achievement without contributing to their lives. The same was reported for eighth- and seventh-grade students whose field experience helped deepen their learning back in the classroom (James and Williams 2017). In another study, an intervention involving practical work outdoors with plants had an effect on students’ attitudes and knowledge of plants three months later (Fančovičová and Prokop 2011). Considering these results, we believe that school-based outdoor learning about relationships in ecosystems can encourage students to deepen their learning beyond school, including at home and in their communities.

Objective of the study and prediction

We considered the following issues, which are discussed above, from the scientific literature to frame our research objective: a) children need more contact with nature to better learn about living organisms, but contact with nature is declining; b) observing living organisms in familiar outdoor environments should enhance students’ ability to observe these organisms without guidance, but students are almost never assigned to make such observations in a formal educational context; c) while the assumption is widely promoted that learning in context improves the quality and persistence of learning, there is little empirical research that clearly supports this assumption; and d) to better study students’ ability to make unsolicited observations in informal settings, it is essential to develop new data collection tools.

Our research objective was to investigate whether a place-based educational intervention in elementary school improves students’ ability to make unsolicited observations about living organisms.

In a previous study, we explored the effects of a place-based educational intervention – one week after it ended – on what and how elementary students learn and found that ‘learning about living organisms in their school’s immediate surroundings was […] a context that encourages deeper learning’ (Ayotte-Beaudet et al. 2023b, p. 287). In this new study, which complements the previous, the focus is on the effect of a place-based educational intervention over time. We wanted to test a common conclusion, which is that time is detrimental to learning (Cooper et al. 1996; Custers 2010). Therefore, our prediction is that elementary school students who learn about living organisms outdoors in a familiar environment will be able to make more unsolicited observations about living organisms one week after the end of the intervention compared to those two months after the intervention ends.

Literature review

Contextualization through place-based learning

Contextualization is a term that covers a range of approaches, including place-based learning, which helps to understand a context through a scientific lens (Giamellaro et al. 2022). In addition to being a contextualised approach to learning, place-based learning is also an approach to providing outdoor education. Place-based learning ‘considers the importance of connecting learners with their community by anchoring pedagogy within the context of the locally natural, cultural, and social ecosystems’, focusing ‘on a specific physical space which may or may not involve the natural environment’ (Lee et al. 2022, 13).

The term contextualization is rooted in an understanding that students’ learning is inextricably linked to context (Giamellaro 2017). It is generally used in education to emphasise the importance of having students engage in learning that will first and foremost be meaningful to them beyond school (King and Ritchie 2012; Rivet and Krajcik 2008), rather than learning that is predominantly for the purposes of assessment in school. Giamellaro et al. (2022, p. 439) define contextualization as ‘a curricular approach and a learning process in which science content knowledge is intentionally situated within a context where that knowledge can be authentically applied or observed’. To differentiate the extent to which science learning is contextualised in a school context, Giamellaro (2017) proposed four categories: 1) academic contextualisation (learning entirely for school purposes), 2) secondary contextualisation (learning with context), 3) primary contextualisation (learning in context), and 4) over-contextualisation (learning with no possibility of transfer).

This research is based on the primary contextualisation concept, implying that students’ learning takes place in a context where the knowledge can be applied. Primary contextualisation contributes to building explicit bridges between science as experienced in school and everyday life (Demircioğlu, Demircioğlu, and Çalik 2009; Walan, Mc Ewen, and Gericke 2016). In order for students to be able to transfer their learning from one context to another, it is important to expose them to different contexts (Kulasegaram et al. 2017), because students’ transfer of school-based science learning into everyday contexts on their own is not automatic (Custers 2010). In fact, students must learn to transfer their learning from one context to another (Sasson and Dori 2015). Furthermore, developing the ability to transfer learning to non-academic contexts is critical, as Custers’ (2010) literature review showed that students typically perform better immediately after the end of an educational intervention.

Consistent with our research objectives and the framework outlined above, our study is rooted in a conception of learning that is grounded in a place-based approach.

Scientific observation

Science relies on observation of natural phenomena. It is a first step towards defining problems and asking questions (NRC 2012). Scientific observation is an essential process when research is based on a collection of facts that are not subject to experimentation with manipulated factors (Giamellaro et al. 2022). It is different from other forms of observation, such as sensory perception, because it involves the use of a scientific and systematic framework of reference for the observation (Remmen and Frøyland 2020). According to Eberbach and Crowley (2009), scientific observation includes four components: 1) noticing scientifically relevant features of the natural phenomenon under study, 2) interpreting those features from a disciplinary framework, 3) using discipline-specific systematic observation strategies, and 4) making observations across a variety of physical and temporal contexts.

Scientific observation can be used as a teaching approach to help ‘students to leverage their existing knowledge and their investigative skills to find, and internalise, new knowledge and solutions to questions they have formulated’ (Bevins and Price 2016, 19). Scientific observation can also promote learning about an environment (Bang et al. 2018), while bridging disciplinary knowledge, theory, and practice (Eberbach & Crowley, 2017). Although scientific observation can be used to collect data during a scientific inquiry, it can also be used as a way to connect natural phenomena to scientific ideas among students (Furtak and Penuel 2019).

Methods

Context of the study and participants

This study was conducted during the first school shutdown caused by COVID-19 in spring 2020 in the province of Québec, Canada. From 13 March 2020 through the end of the school year in June, students did not return to school. They only came back to school at the beginning of the following school year in late August. The students participating in this study were voluntarily engaged in remote activities as a result of both parental choice and access to digital devices.

While this research was originally intended to take place in an environment near the school, the COVID-19 circumstances resulted in students experiencing the educational intervention described in this research from home. To test our prediction, we compared elementary school students’ ability to make unsolicited observations about living organisms one week (Group A) and two months (Group B) after a place-based educational intervention in their home environment. The decision to compare two different groups of students rather than test the same group after a week and two months was based on our assumption that data collection one week after the intervention would act as an intervention itself.

We chose one week after the intervention for Group A (+1 week), to collect observations almost immediately after the end of the intervention. For Group B (+2 months), we wanted to leave as much time as possible before the start of the new school year. More than two months would have led to the start of a new school year, the first after the several months’ shutdown due to COVID-19. This meant we wouldn’t have had as many participants, and the first few weeks of school could have influenced the results.

We first recruited 20 teachers in grades 5 and 6 (whose students were between 10 and 12 years old) through social media, as all schools were closed. All students were recruited from these 20 classes on a voluntary basis. The students came from urban (a few hundred thousand inhabitants), semi-urban (a few tens of thousands of inhabitants), and rural areas (a few thousand inhabitants). They included Canadian-born students and first- and second-generation immigrant students. Most of the students were French-speaking, but some were English-speaking. All participating students used their first language during the data collection. Students aged between 10 and 12 should be familiar with the characteristics of living organisms, their organisation and their stages of growth (MEES, 2018). We controlled for several variables, including teacher, grade, and gender. In the end, a total of 116 students were included in the study, 58 in each of the two groups (A and B). For each student in Group A (+1 week), there was a student in Group B (+2 months) with the same teacher and gender. Each pair of students was then randomly assigned to either group (A or B). By assigning students to each group this way, we were able to ensure that outdoors environments near their homes, prior knowledge and observation skills were the same in both groups. However, because group B data collection took place during the summer vacation, around ten pairs were assigned based on student availability.

As mentioned earlier, this study was conducted during the first school shutdown caused by COVID-19 in spring 2020, when the end of the school year was officially cancelled in the province of Quebec, Canada. Participation in the project was therefore voluntary for both teachers and students. This is what explains the – unusual for a study – average of 5.8 students in the 20 participating classes.

Students in both groups completed the same tasks. All participating teachers received an online briefing from members of the research team prior to the beginning of the intervention to ensure they understood the materials we asked them to use through the lessons and to ensure consistency in the process. The overall goal of the Clay caterpillar project was to observe plants and animals in the vicinity of the students’ homes in order to gain an understanding of possible relationships between living organisms. During the first lesson with the students, the teacher was asked to lead an online discussion in which students discussed asking questions about, and observing, living organisms in the environment (e.g. What living organisms are on your property or near your house?, What do you hear when you are on your property or near your house?, What do you smell?). During the second online lesson, students took part in a citizen science project that had been developed the previous year to study biodiversity in urban areas (Ayotte-Beaudet et al. 2023b). The students were required to instal 20 small artificial prey (clay caterpillars) on a single tree for a period of one week using a kit sent by our research team (see Figure 1). The full protocol presented to the students was adapted from an authentic ecological research protocol (Castagneyrol et al. 2020; Low et al. 2014; Muiruri, Rainio, and Koricheva 2016) and is available in French on the Citizen Science project website (https://www.chenilles-espionnes.com/réalisation). The following week, the clay caterpillars were removed for analysis of predation marks (e.g. from bird beaks, squirrel or arthropod bites). The caterpillar protocol is used in ecology to test fundamental hypotheses’ basic proposition or assumption that is widely accepted by ecologists and that helps to explain or predict various ecological phenomena (Low et al. 2014; Muiruri, Rainio, and Koricheva 2016) and also in other science education projects (Ayotte-Beaudet et al. 2023b; Castagneyrol et al. 2020). In a final online lesson, the teacher led a discussion with the students, asking them about their unexpected observations about living organisms or questions that scientists should answer based on observations made outdoors near their homes.

Figure 1. A student installs an artificial prey (a clay caterpillar) in a tree.

Data collection

Data were collected through structured 30-minute telephone interviews with each participating student. We conducted a total of 116 interviews: 58 for group A and 58 for group B. We conducted interviews because elementary school students generally have greater difficulty expressing themselves in writing than orally (Angeloudi, Papageorgiou, and Markos 2018; Pekmez 2018). Even though the students had made scientific observations within the context of a citizen science project, the focus was their ability to make novice unsolicited everyday observations that were not directly connected to the citizen science project. The students’ observations about living organisms provided evidence of their self-interest in them. In this study, an unsolicited observation consisted of noticing natural phenomena without being prompted by an educational intervention.

During the interviews, because the caterpillar project was the study context for students, we addressed this project at the beginning of the interview. However, since we were studying students’ unsolicited observations about living organisms, the answers that really interested our research team were the observations that were not directly connected to the caterpillar project. Even though we were interested in the students’ recent observations from their home environment, our interview protocol sought to elicit them. We questioned them about a) their experience with the caterpillar project, b) their conception of living organisms, and c) the observations they had made about living organisms in general since the beginning of the project (see Supplementary Material for full protocol). Examples of questions for the third part of the interviews included: What have you observed about living organisms since the beginning of the project that you had never observed before? What are the most surprising and interesting observations you have made about living organisms? Since the beginning of the project, have you made observations about living organisms that scientists might have made?

During the interviews with the students, we also controlled for the students’ ability to distinguish a living organism from a non-living entity. Biological misconceptions are common among students (Kumandaş, Ateskan, and Lane 2019), which is why we wanted to compare their understanding of living organisms – plants and animals in particular, since they are the ones that are familiar to students between the ages of 10 and 12. Students’ ability to distinguish a living organism from a non-living entity was measured with a list of six words (a table, a tree, a rock, a dog, an insect, a plate) for which the students were asked to say whether or not they were living organisms (yes; no). In addition, we controlled for students’ individual interest in living organisms, since contextualised learning can influence interest (Potvin and Hasni 2014b). Individual interest in living organisms was measured with 5 Potvin and Hasni (2014a) items (a: I am looking forward to the next activities on living organisms; b: Living organisms are fun; c: Living organisms are boring; d: Living organisms are really interesting; e: We should spend more time talking about living organisms in school) using a 4-point Likert scale (strongly disagree, disagree, agree, strongly agree).

Data analysis

The first step in the data analysis was to select excerpts from the student interviews that report unsolicited observations about living organisms. All interviews were transcribed and anonymised before being coded. Two research assistants coded all of the interviews separately and did the review together. A third person (the first author) helped to arbitrate the few cases where there was not an agreement. In order to be consistent, the definition of an unsolicited observation was always strictly applied. Every excerpt was coded in such a way that each unsolicited observation of living organisms that was described was entirely captured, meaning that we felt confident about its significance. We decided that the observations in the excerpts had to be 1) about any living organism(s), plants or animals, except human beings, 2) unsolicited, that is, not related to an educational intervention and not prompted by the caterpillar project, 3) made in an outdoor place frequented by the student, and 4) made during or after the caterpillar project. It is important to mention that we recognise that the human being is part of nature and has a significant impact on ecosystems and we did not ask the students to exclude human beings from their answers during the interviews. However, we decided not to include humans since we wanted to focus on the organisms naturally living outdoors where the observations were made. The criteria for the inclusion and exclusion of excerpts are presented in Table 1.

Table 1. Criteria for inclusion and exclusion of an unsolicited observation meaning unit.

Included is an observation …	Excluded is an observation …
… about any living organisms, except the human being.	… strictly about the human being.
… that is unsolicited, meaning that it was not related to an educational intervention and not prompted by the caterpillar project.	… that is elicited, meaning that it was related to an educational intervention or that it was elicited by the caterpillar project.
… based on an outdoor place frequented by the student.	… that was not made on-site by the student.
… made during or after the caterpillar project.	… made before the caterpillar project.

The second step was to develop categories, or levels of observation, for analysing the excerpts. To do so, we used the conceptual framework of both place-based contextualisation and observation and Hattie’s (2009) three learning phases. These three phases of learning are defined by Frey, Fisher, and Hattie (2017, 570) as follows: surface learning refers to the ‘acquisition and consolidation of initial knowledge base, deep learning to the ‘interaction with skills and concepts, and transfer learning to ‘organizing, synthesizing, and exending conceptual knowledge’. Five members of our team finally stabilised the definitions of three emergent categories for analysing the unsolicited observations about living organisms (1- unsolicited observation about living organisms leading to a factual description, 2- unsolicited observation about living organisms leading to context-specific inferences, and 3- unsolicited observation about living organisms leading to inferences in other contexts). We then proposed definitions for each of these three categories, grounded in our data. The definitions are presented in Section 5.1.

The third step was to assign an observation category to each of the excerpts identified in step one of the data analysis. Two team members separately coded all of the excerpts and then compared all of the coding in order to reach an agreement about each of the units. A third team member then reviewed the segments on which the first two coders could not agree.

The final step was to conduct statistical analyses to test our prediction. We conducted an exploratory factor analysis to determine the adequacy of the individual interest scores. The percentage of each student’s correct answers showing their ability to distinguish a living organism was then averaged across all students to obtain an overall index. The dependent variable (unsolicited observations about living organisms) for the three levels of observation was dichotomised (yes/no). Since the dependent variable for the three levels was dichotomous, we performed a logistic regression for each level of unsolicited observations about living organisms with the group (A or B) as the predictor and the students’ situational interest score and their ability to distinguish a living organism score as control variables.

Results

Emergent categories for analyzing place-based unsolicited observations about living organisms

As described in the Data Analysis section, we developed and characterised three levels of unsolicited observations about living organisms to analyse the students’ observations that emerged during the interviews. Here we present a definition and two illustrative student interview excerpts for each level of unsolicited observations. Table 2 shows more examples of student interview excerpts for levels 1 and 2 of unsolicited observations, for readers to have a better understanding of the observations made.

Table 2. Examples of student interview excerpts for levels 1 and 2 of unsolicited observations.

(Level 1) Unsolicited observation about living organisms leading to a factual description	(Level 2) Unsolicited observation about living organisms leading to context-specific inferences
‘I saw a bird’s nest in a low spot in a tree few days ago’. (Group A, 08JM10)	‘There are a lot more mosquitoes because it’s summer. I think in winter, they’re probably there, but hidden. They wait for summer, like when we wear shorts and short-sleeved sweaters, because with our snow overalls and coats, they can’t really bite us’. (Group A, 07JM08)
‘We have a garden and it’s really grown. At first, we had baby lettuces, but now they’re ready, they’re big and we’re going to pick them’. (Group A, 07JM08)	‘I watched a groundhog up close. It came into our garden. I got a good look at its face, but all rodents have the same face! It’s got a snout, long teeth, and eyes’. (Group A, 20JM24)
‘I also found some eggs in a bird’s nest in a bush’. (Group A, 08JM10)	‘I’d never seen ants on our peonies and trees before. They come and feed. In fact, I wonder if they’re pollinating insects, because they don’t have wings’. (Group A, 20JM24)
‘Squirrels often throw pits into our inflatable pool. I thought squirrels kept the pits’. (Group A, 26JM34)	‘In front of our house, our tree is growing, but it’s sick. It never occurred to me that maybe it’s the bad soil that’s made it sick’. (Group A, 28JM36)
‘There’s a tree in my backyard. I’ve lived in this house for 9 years. Before the project, I’d never noticed it growing fruit’. (Group A, 37JK08)	‘I thought an animal could be quiet anywhere. But it turned out not. A little further from my house, there were a lot of animals going to one side of the street, but they seemed to avoid the other side where there was construction, maybe to be safer’. (Group A, 35JK05)
‘This summer I’ve been watching locusts, bees, and honeybees on the sunflowers’. (Groupe B, 08AM03)	‘I learned that trees and plants were alive, as we’d already learned at school. I also saw that there are lots of animals that settle on them. It’s as if the tree were a house’. (Group B, 01AK16)
‘My mother made a garden. Every day, we get big tomatoes and cucumbers. It goes fast’. (Group B, 30AM24)	‘The life cycles of my tomatoes; there was a little seed and a little plant and then it made flowers and then the flowers were tomatoes’. (Group B, 08AM03)
“I have lots of little rabbits in my yard. There’s a family that walks around there, so we give them food sometimes. I’ve noticed that you can get very close to them, and they’re not shy at all. (Group B, 33AM07)	‘I had seen a caterpillar cocoon in my field. I thought their transformation was a form of adaptation’. (Group B, 14AK09)
“I always saw spiders in houses, but I didn’t know they could also be in trees. (Group B, 37AK28)	‘A lot more animals come out in the summer. Such as in my neighbourhood, there’s a lot more of skunks, birds, and squirrels’. (Group B, 49AM11)
‘I didn’t know that there could be black squirrels. I thought there was only grey and brown ones’. (Group B, 54AM18)	‘Living beings are much more intelligent than we think. I always thought a bird was less intelligent than a human. But it has a way of life and can recognise caterpillars. They’re a lot smarter than you think’. (Group B, 52AM20)

(Level 1) Unsolicited observation about living organisms leading to a factual description. A practice that consists of noticing only visible characteristics of living beings in an environment and does not lead to the development of knowledge about the environment in question. This is a necessary initial step in the student’s learning process.

I saw a bird’s nest in a low spot in a tree few days ago.

(Group A, 08JM10)

This summer I’ve been watching locusts, bees, and honeybees on the sunflowers;. (Groupe B, 08AM03)

(Level 2) Unsolicited observation about living organisms leading to context-specific inferences. A practice that uses one or more informal observations leading to a factual description to develop knowledge about the specific environment. This is a step in the learning process that allows students to discover relationships between their observations as well as concepts or theories.

I thought an animal could be quiet anywhere. But it turned out not. A little further from my house, there were a lot of animals going to one side of the street, but they seemed to avoid the other side where there was construction, maybe to be safer.

(Group A, 35JK05)

The life cycles of my tomatoes; there was a little seed and a little plant and then it made flowers and then the flowers were tomatoes.

(Group B, 08AM03)

(Level 3) Unsolicited observation about living organisms leading to inferences in other contexts. A practice that uses one or more informal observations leading to a context-specific inference to identify whether it is possible to systematise these observations in settings other than the one where the initial observations were made. In addition to being the ultimate step in the learning process, this is similar to the concrete process of reflection that constitutes a scientific investigation.

The predators near a house are different than in a forest. Maybe it is because there’s fewer trees and some predators need more space to go high and move around.

(Group B, 12AK24)

Students’ ability to make unsolicited observations about living organisms

For the five items measuring individual interest in living organisms, we conducted an exploratory factor analysis; the third item was inverted because it was phrased negatively. The Kaiser-Meyer-Olkin measure verified the sampling adequacy for the analysis, KMO = .695, which is above the recommended threshold of .5 (Field 2013). Bartlett’s test of sphericity, X = 67.07, p = 0, showed that the sample was adequate. A principal component analysis indicated two components as did the eigenvalue graph, but the second one was borderline. Since the purpose was to arrive at a single score, the single component solution was explored. All five items had a factor loading of at least .4 (Stevens 2009), indicating that it was appropriate to create an individual interest in living organisms score.

For students’ ability to distinguish a living organism from a non-living organism, we created a score between 0 and 1 (e.g. 5 correct answers out of 6 = .83). The participants had an average score of .99, which means that most of them had a perfect score; only 12 participants failed (only) one question (Table 3). Thus, we confirmed that the students were good at distinguishing a living organism.

Table 3. Descriptive statistics for the individual interest and students’ ability to distinguish a living organism score.

Scores	N	Mean	SD	Median	Min	Max
Individual interest	116	3.50	.44	3.6	2.20	4
Ability to distinguish a living organism	116	.99	.04	1.0	.86	1

After the two research assistants analysed the excerpts independently, we established an interrater reliability score of 92% for their classification into the three levels of unsolicited observations about living organisms’ categories. In a meeting with a third team member, the remaining 8% were categorised. For each of the three levels, each participating student was assigned presence (1) or no presence (0) of that level. For instance, 18 students in group A (+1 week) talked about an unsolicited observation about living organisms that led to a factual description (level 1) compared to 32 students in group B (Table 4).

Table 4. Descriptive statistics for the three levels of unsolicited observation about living organisms.

Scores	N	Level 1		Level 2		Level 3
		0	1	0	1	0	1
Group A (+1 week)	58	40	18	43	15	57	1
Group B (+2 months)	58	26	32	41	17	56	2

To add the control variables, logistic regressions were performed for level 1 and 2 observations only, as there were not enough participants who made at least one level 3 observation (Table 5).

Table 5. Logistic regressions for levels 1 and 2 of unsolicited observations about living organisms.

	Level 1 (n = 116)			Level 2 (n = 116)
Predictors	OR	CI	p	OR	CI	p
Intercept	0.01	0.00–63.80	0.303	0.64	0.00–5117	0.924
Month (June/August)	2.70	1.27–5.90	0.011	1.18	0.52–2.72	0.686
Individual interest	1.62	0.68–4.06	0.289	1.76	0.68–5.02	0.266
Distinguishing a LO	12.94	0–258 969	.586	0.070	0.00–1257	0.570

Our results showed that the predicting variable for level 1 is the group (p = 0.011): participating students from group B (= 2 months) were 2.7 times more likely to make unsolicited observations about living organisms leading to a factual description than students from group A (= 1 week). There was no difference between group A and group B for Level 2.

Discussion

The purpose of this study was to investigate whether a place-based educational intervention in elementary school would lead students to develop a better ability to make unsolicited observations about living organisms one week after the end of the intervention compared to those two months after the intervention ends. Through student interviews, we developed three levels of unsolicited observations about living organisms that were grounded in our data: (level 1) observations leading to factual descriptions, (level 2) observations leading to context-specific inferences, and (level 3) observations leading to inferences about other contexts. We also found that elementary school students who learned about living organisms in their home environment made more unsolicited observations about living organisms that led to factual descriptions two months after the end of the learning experience than one week after the intervention.

Contribution of our emergent categories for analyzing unsolicited observations in future research

While we initially intended to create categories of analysis based on contextualisation, we did not know whether the focus would primarily be on content or on context. We believe that we were successful in developing categories of analysis that consider both content and context, as well as students’ ability to transfer their learning to other contexts. Here we drew on the three categories of learning presented by Almarode et al. (2018): teaching for surface learning (which inspired level 1), teaching for deep learning (which inspired level 2), and teaching for transfer learning (which inspired level 3).

The first important attribute of our categories for assessing the students’ scientific knowledge and skills is that they were adapted to the uniqueness of the outdoor environments in which the learning experiences took place. Given that there is a lack of standardised outcome measures adapted to outdoor nature-based learning (Miller et al. 2021), our categories contribute a solution to this problem, which is a key issue for more robust advancement of research in the field (Jordan and Chawla 2019). The second important attribute of our categories is that, even though they were developed to collect data about students’ unsolicited observations about living organisms, they can be used for any scientific topic. The phrase living organisms can be replaced by any object of scientific knowledge under investigation, and our definitions will remain relevant. For instance, to study a learning experience in geology, the first level of observations could be changed to unsolicited observation about rocks and minerals leading to a factual description. The fact that our categories are designed for the outdoors, can be used in any location, and can be applied to a broad variety of scientific knowledge leads us to believe that they will be of genuine value for the advancement of research about place-based outdoor science.

Contribution of our results regarding the use of outdoor place-based science education

We rejected our prediction that elementary school students who learn about living organisms in their home environment will make more unsolicited observations about living organisms leading to a factual description (Level 1 observations) one week after the end of the learning experience than after two months. We also rejected the prediction that elementary school students who learn about living organisms in their home environment will make more unsolicited observations about living organisms leading to context-specific inferences (Level 2) two months after the end of the learning experience than after one week. Although we rejected the prediction for the levels 1 and 2, the reasons are different for Level 1 and for Levels 2. For Level 2, there was no significant difference between group A and group B. However, for Level 1, results were reversed from what was expected. This suggest that the null hypothesis is unlikely to be true for unsolicited observations about living organisms leading to a factual description.

While students typically perform better immediately after the end of an educational intervention (Custers 2010), our results instead indicate that students were better at making Level 1 observations two months after the event than they were after just one week. Even though this may be the case, since our study was limited to one context, we cannot conclude that the outdoor learning context caused more observations two months later. Still, we think our results were influenced by context. While in school, students generally learn in a context that remains academic (Giamellaro 2017), the educational intervention they experienced at home in this study allowed them to continue applying their learning without the intervention of a teacher. Because schools typically focus on student achievement in academic contexts (Blatt and Patrick 2014; Carrier, Tugurian, and Thomson 2013), schools commonly emphasise close transfer; that is, a context that is similar to testing contexts. This could be part of the reason why students have difficulty transferring academic learning to out-of-school contexts (Khishfe 2019). Thus, while transferring learning autonomously and outside of school is generally much rarer than one would like (Custers 2010). In contrast to most school experiences, students had the opportunity to observe living organisms in the same context during and after the intervention. As context influences learning (Giamellaro 2017), students may have been able to consolidate their learning autonomously within the context of the original intervention. Additionally, participating students may have benefits from spaced repetition, which is recognised to allow learning to be consolidated and more easily transferred to different contexts (Cepeda et al. 2008; Roediger and Butler 2011). Although our study design does not allow us to conclude that the outdoor education settings at home are responsible for our significant results for level 1, there are some reasons to believe that they might have played a role to explain them. We believe our research opens up new avenues for understanding outdoor learning’s longer term effects on students’ ability to make observations about living organisms in a familiar environment.

The results for the Level 1 observations also shed light on the theoretical assumption that learning is inseparable from the context in which the learning occurs (e.g. Liljeström, Enkenberg, and Pöllänen 2013). Several researchers have argued that the more scientific learning is applied in contexts that can be encountered by students beyond the academic context, the more likely they will be able to engage that learning in non-academic contexts (Braund and Reiss 2006). Indeed, Giamellaro (2017) suggests that this also promotes the ability to transfer scientific academic learning in the longer term. Although this assumption seems to be generally accepted in the scientific literature, little empirical research seems to exist that has investigated this prediction. Our results for Level 1 observations tend to support this claim.

It is also important to discuss the results for students’ ability to make Level 2 and Level 3 observations, which were not significant. What we can understand from these results is that students in Group B were no more able to associate their unsolicited observations with concepts or theories or to infer them for other contexts. These results are not very surprising, as understanding ecosystems is a complex operation for non-experts (Ben-Zvi Assaraf and Orion 2009; Bermudez et al. 2017; Hmelo-Silver, Eberbach, and Jordan 2014), even for teachers (Sweeney and Sterman 2007). Understanding complex systems requires abstract thinking (Hmelo-Silver and Pfeffer 2004). In addition, explanations of natural phenomena are generally nonlinear, while students tend to have very simple causal explanations (Perkins and Grotzer 2000). Experts are more able than novices to have a deeper understanding of the complex systems in which living organisms evolve in an ecosystem (Hmelo-Silver and Pfeffer 2004).

Perhaps more crucial to understanding these results in a broader perspective is that deep learning does not occur automatically, nor does learning transfer (Ayotte-Beaudet et al. 2023b). It takes time for students to become familiar with the surface or factual knowledge and concepts or principles of a discipline before they can consolidate and deepen their knowledge (Frey, Fisher, and Hattie 2017). To help students do so, teachers have an important role to play (Biggs 1999). This potentially explains why almost no students in either group were able to make Level 3 observations. To increase students’ ability to make Level 2 and Level 3 observations, we therefore believe that the support of a more expert person, which in a school-based situation is a teacher, is essential.

Limitations and further research

To further understand the meaning of our findings, it is important to recognise some limitations of this research. First, it is important to recognise that it was conducted in the context of COVID-19, during the first lockdown at the end of the 2019–2020 school year. Thus, the context of the research was not the usual school setting. Students experienced the instructional intervention in their home environments. The teachers led the educational experiences only remotely. It was therefore more difficult for our team and for the teachers to ensure that the students followed the entire sequence without the intervention of a parent or tutor. Participation in this research and the teacher-led learning situations was also voluntary, and not all students in each group logged in to do their online learning. Typically, the students who participated were those who had technological devices at home and who were encouraged by a parent or tutor to do online schooling. Thus, it is reasonable to assume that students from higher socioeconomic backgrounds were overrepresented, which limits the generalisability of our results. Nevertheless, we believe that our efforts to control for important variables and to create quasi-equivalent groups with respect to teacher, grade, and gender mitigated these effects to a certain extent.

Another limitation we faced in this study was the method for analysing the student interviews. It was difficult to isolate the unsolicited observations made by a single student. Our team therefore concluded that it was more reasonable to characterise the highest level of unsolicited observation each student achieved rather than to count the number of unsolicited observations for each of the three levels. Despite this limitation, we were able to see a significant difference between the two groups for Level 1 observations.

One more limitation of our research is related to the interview protocol. The way the questions were asked implied that students had observed living organisms. We did consider asking them if they have observed any living organisms before assuming they have. However, our pilot interviews revealed that some students may opt to answer no, even if they have made valuable observations. In framing the question as if they observed living organisms, we believe we maximised the richness of the answers. And, despite this formulation, a few students stated that they had not seen living organisms. Of course, interviewers had to ensure that the details provided were sufficient to convince us, when analysing the students’ transcripts, that the answers weren’t made up. This direct manner of asking questions was also based on our interview experiences with 5th and 6th graders, which showed it is sometimes difficult to elicit detailed answers from students of this age if the questions are dichotomous and do not require them to elaborate (Ayotte-Beaudet et al. 2023b).

For future research, we suggest collecting data in a more systematic way so that the levels of the unsolicited observations can be properly segmented. For example, students could be specifically asked to identify 5 unsolicited observations so that the research team would be able to clearly distinguish each observation, and thus produce more detailed results. Another idea would be to ask students to take 5 photographs in a given environment and then ask them to verbally explain each of their observations. These two (non-exhaustive) options show that our framework can be used in a wide variety of applications depending on the research design.

Our results also allow us to point out further avenues for research. This study was conducted in students’ home environments, not the typical school setting. It would therefore be of interest to see whether the results would be similar if the pedagogical intervention occurred in a more structured school context. Also, the results of our research suggest that it may be very difficult for students to make unsolicited observations that are in-depth or that lead to transfer to other contexts. Therefore, another important suggestion for future research would be to develop a pedagogical intervention that explicitly integrates situations in which the teacher supports the students in in-depth learning and helps them experience explicit transfers. In this case, we believe that students would be more likely to make such observations autonomously, and this prediction certainly deserves to be tested in future research. Another limitation of our research is that we cannot conclude that it is specifically the contextualisation through outdoor place-based learning that led to a difference between the two groups. We do not know whether more formal classroom instruction would have led to the same results. Therefore, we suggest that future studies compare results for students who learned outdoors with those who learned in the classroom. Finally, we believe that it would be relevant to study scientific concepts other than living organisms to verify whether the results are replicable. Such results could allow us to better understand how to enable students to achieve learning in a school context that they can mobilise autonomously in their daily lives, beyond the academic and assessment contexts.

Conclusion

We conducted this research because although the assumption that learning in context improves the quality and persistence of learning is widely accepted, there is little empirical research that clearly supports this assumption. While more research is needed to better understand this assumption and its nuances, our research has enabled progress in the understanding of the contextualisation of learning, especially in outdoor place-based learning settings. Our prediction was that elementary school students who learn about living organisms in their home environment will make more unsolicited observations about living organisms one week after the end of the learning experience than two months after the experience. Our results rejected this prediction for the two first levels of unsolicited observations about living organisms, but the null hypothesis is unlikely to be true for unsolicited observations about living organisms leading to factual descriptions (level 1 observations). These results lead us to believe that observations that demonstrate deep learning and the ability to transfer learning to other contexts are less easily made by students without the support of an expert person. Our results enable several research designs that introduce variations of our design to better understand the subtleties of the impact of outdoor place-based contextualisation of learning on students’ ability to autonomously transfer these learnings beyond the school context (e.g. by modifying the pedagogical intervention sequence, the data collection moments, or the scientific learning). It is also important to highlight that our research allowed the development of categories of analysis that go beyond the measurement of surface learning and that are adapted to the outdoor context, which is an important contribution for the development of research in this field. We therefore hope that these categories will be used and adapted and that they will inspire other data collection instruments.

To conclude, we would like to stress how important it is to us that school serves first and foremost as a place for scientific learning that is relevant to students beyond the academic context. We believe that even at school, learning should not be primarily for academic and assessment purposes, which is too often the case. Additionally, although we did not emphasise it in this paper, there are reasons to believe that outdoor place-based education promotes equity in science education, helping all learners regardless of age, gender, and socioeconomic backgrounds. We thus hope that other research teams will build on our results and join our research efforts to understand the potential impact of outdoor place-based contextualisation on learning.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research project has been approved by the Research Ethics Committee – Education and Social Sciences of the Université de Sherbrooke [SRef. No. 2019-1953]. A free and informed consent form was signed by all parents of participating students.