What children know and want to know about climate change: a prior-knowledge self-assessment

Abstract

Researchers engaging in science communication tend to present information that they find most important. However, understanding the level of prior knowledge and interest of a target audience is key for effective communication of science, especially when dealing with complex and mediatic themes, such as climate change. This study relies on a prior-knowledge self-assessment applied to children between 10 and 13 years old, followed by a mixed analysis under the assumptions of the Grounded Theory. The goal is to understand whether children’s self-perception of knowledge and interests regarding climate change can support the design of environmental education initiatives. Children declared to know about the concept of climate change and specific causes and consequences, but were generally incapable of specifying topics about which they are interested in learning about. And when these were specified, they match the ones they already know about – sea ice changes, pollution and biota. This suggests that relying solely on children’s interests may originate ‘information corridors’. Future initiatives should therefore not only resonate with children’s experiences and reality, but also broaden fields of knowledge by introducing innovative approaches that go beyond researchers’ pre-assumptions.

Graphical Abstract

Keywords:

• Science communication
• SDG 13: Climate action
• prior-knowledge self-assessment
• children
• mixed analysis
• grounded theory

1. Introduction

Science communication and science education have similar general goals, as they both aim to foster students’ scientific knowledge and ultimately promote active citizenship (Kohen and Dori 2019). However, while the science communication community is deeply focused on the process of communicating science to achieve its goals; the education community is largely focused on the gains as a result of learning and engaging with science (Kohen and Dori 2019). Despite this intrinsic disparity, there have been recent efforts to narrow the gap between both communities (Höttecke and Allchin 2020). Nemadziva, Sexton, and Cole (2023) explored strategies from the science communication literature and presented recommendations to improve learner engagement with science. Aurélio et al. (2021) explored the impact of a children’s science communication book to promote environmental awareness and increase knowledge. Trott (2020) used participatory methods with children to facilitate informed climate change action, increasing learning and motivation. Guilherme, Faria, and Boaventura (2016) explored out-of-school settings as a learning context to increase scientific knowledge.

One of the strategies used by the education community is the invitation of researchers to the classroom (Aurélio et al. 2021; Boaventura et al. 2021), which may be beneficial for both students and teachers (Aurélio et al. 2021). However, the risk of cognitive load is real if the instructional materials provided are excessive (Cook 2006), which may easily occur when researchers use themselves as an audience model presenting information that they find most important and interesting (Bruine De Bruin and Bostrom 2013). To guarantee that the informational needs of the audience are being met, it is essential to understand the level of prior knowledge and interest of the target audience, which is one of the key elements of effective science communication (Mercer-Mapstone and Kuchel 2017; Shivni et al. 2021).

In education, it is fairly common to identify baseline skills to prepare a tailored curriculum to help students learn and achieve conceptual understanding as efficiently and effectively as possible (Bergan-Roller et al. 2020). But in science communication, although there are valid examples such as those using mental models (Sleboda and Lagerkvist 2022), the design of initiatives is more channel-oriented (Cinelli et al. 2022) and age-oriented (Aurélio et al. 2021; Trott 2020), rather than based on prior knowledge or interests (Bruine De Bruin and Bostrom 2013). And, frequently, for researchers engaging in science communication, especially those without sturdy communication skills and training, not only is prior knowledge often ignored, but communication tools are often limited to adapting language by simplifying scientific jargon (Rakedzon et al. 2017).

Assessments of prior knowledge and interests are particularly relevant for complex and mediatic themes, such as climate change. These assessments allow defining a more careful and informed communication strategy in situations where the potential for fragmented knowledge (Cook 2006; Jurek et al. 2022), misinformation (Farrell 2019; Maertens, Anseel, and van der Linden 2020; Treen, Williams, and O’Neill 2020) and consequences for mental well-being (Martin et al. 2022; Vergunst and Berry 2022), is high. These assessments can actually benefit from the use of open self-assessments (Falchetti, Caravita, and Sperduti 2007), as they are potentially more relevant and advantageous for understanding the audience’s prior knowledge than if questions are selected and directed by the researchers.

To explore the power of open prior-knowledge self-assessments within the thematic area of climate change, an open questionnaire, followed by a three-step qualitative analysis and a complementary quantitative analysis, was implemented to understand what students between 10 and 13 years old know and want to know about climate change. The goal is to understand whether students’ self-perception of knowledge and interests regarding climate change can, and/or should, be considered to guide the design of climate change communication initiatives, especially in educational contexts.

1.1. State of knowledge of climate change

The IPCC Sixth Assessment Report (AR6) states that, through emissions of greenhouse gases, human activities have unequivocally caused global warming and led to widespread losses and damages to nature and people (IPCC 2023). The implementation of adaptation and mitigation measures for climate change has been progressing, but maladaptation is happening and financial flows are low. This iscompromising the 1.5 °C limit in temperature increase which will potentially intensify multiple and concurrent hazards. In the current situation, some future changes are unavoidable and/or irreversible, such as sea level rise, but can be limited by deep, rapid and sustained global greenhouse gas emissions reduction with positive impacts on human well-being and health. Securing a livable and sustainable future for all is thus dependent on accelerated and equitable mitigation and adaptation measures to climate change, which are, in turn, reliant on strong political commitments and multilevel governance (IPCC 2023).

1.2. Prior-knowledge assessments

Prior-knowledge assessments are widely applied to gauge how much a specific target group has learned during an activity, course program, or other initiative. A variety of methods exist to assess the participants’ prior knowledge and skills. Some include direct assessment of participant’s capabilities at the beginning of the initiative (e.g. portfolios, pre-tests, auditions) (Redmond et al. 2021; Wammes et al. 2022). Others include indirect assessment methods (e.g. students’ self-reports, inventories of prior courses or experiences). Next, four of these methods are described.

Performance-Based Prior Knowledge Assessments (Performance-based PKA) are considered the most reliable method, highly suitable for diagnostic purposes. With a Performance-based PKA, a task is assigned to participants (e.g. quiz), which should help the evaluator gain an overview of participants’ preparedness, identify areas of weakness, and adjust the pace of the initiative to be applied (Wammes et al. 2022).

Prior-knowledge self-assessments (Prior-knowledge SA) are suitable to calibrate the initiative to be developed or direct participants to supplemental materials that can help them address weaknesses in their existing skills or knowledge. With a Prior-knowledge SA participants are asked to reflect and comment on their level of knowledge and skills across a range of items (Redmond et al. 2021). It is a relatively easy method to apply, but the evaluator must be aware that participants may not be able to accurately assess their abilities (Sitzmann et al. 2010). The work presented here relies on this method.

Classroom Assessment Techniques (CATs) are suitable to provide immediate feedback about a group’s level of understanding of specific themes. With a CAT, the evaluator is frequently looking to identify the most significant aspects participants know about a theme, or to identify areas of confusion (Saeed, Tahir, and Latif 2018).

Concept maps are suitable to understand how participants organize and represent knowledge (Bergan-Roller et al. 2020). With the Concept maps method, participants are asked to provide a graphic representation of a specific theme, establishing relationships between concepts.

1.3. Case study context

MARE-Marine and Environmental Sciences Centre has, since 2015, an Ocean-related educational program targeted to the Portuguese compulsory education cycles. The program is called ‘O MARE vai à Escola – MARE goes to School’ (https://www.mare-centre.pt/). It is a highly successful program and one of the first in Portugal fully developed and led by researchers. From 2015 to 2019, more than 40,000 students from 358 schools benefited from the program, which covers themes such as overfishing, marine litter, biotechnology, marine biodiversity and climate change. Some activities are developed at school, within classes, whereas others are developed outside school through informal educational activities.

Focused on climate change, ‘Who set up the climate?’ is currently one of the most requested ‘MARE goes to school’ activities. From 2017 onwards, it gradually gained more attention from schools and the increased interest was attributed to the general increase in public awareness, at a global scale, materialized by the School Strikes for Climate (Kenis 2021; Variyan and Gobby 2022), higher amount of news in traditional media and social networks (Parks 2020; Sadik, Benedetti, and Gokhale 2022; Variyan and Gobby 2022), and more frequent political discourses calling out the problems associated to climate change (Arezki et al. 2018; Foss 2018; Tykhomyrova 2020).

While implementing the activity, the researchers realized that most of the students had, in fact, heard the concept climate change before, but the activity plan was not prepared to evaluate the students’ knowledge about the theme, neither before nor after the activity. Moreover, the researchers realized that the activity could be improved, but there was a lack of information regarding what students actually knew or wanted to know about climate change, that could inform the team on the path to follow for improvement.

The researchers decided to focus on this last challenge, implementing a prior self-assessment method, to a sample of students, in which they were prompted to reflect on their level of knowledge about climate change. Ultimately the results will support the improvement of communication activities within the Climate Change thematic area.

2. Methodology

A non-structured prior-knowledge self-assessment method was defined to allow participants to reflect on their level of knowledge and interest regarding climate change. As such, an open-question survey was prepared with two questions: What do you know about climate change? (knowledge) and What do you want to know about climate change? (interest). For each question, students could provide a maximum of four items, i.e. responses.

Questionnaires were applied to middle school students, from the ages of 10 (5th grade) until the ages of 13 (8th grade) (Table 1). This is a non-interventional study, for which no identity information was collected and stored. As such, an ethical approval is not required. However, because the study participants were involved in a school programme named ‘Clube Ciência Viva na Escola’, broad consents were obtained from children and parents by the school.

Table 1. Number of questionnaires per type of session (individual / group), per school grade.

		School grade
	Type of session	5th	6th	7th	8th	Total
Number of questionnaires	Group	12	16	13	19	60
	Individual	0	25	39	42	106
Total		12	41	52	61	166

The survey was applied to students both individually (n = 106) and in group (n = 60), encompassing a total of 417 students (Figure 1). However, individual questionnaires for the 5^th grade were not collected due to the lack of available classes in the participating schools, after collecting the group questionnaires. The lack of further available classes was the result of miscommunication with the school and nothing could be done to overcome it. For individual sessions, each student had 10 min to answer the survey, without any discussion with their peers. For group sessions, each class was divided into four groups of students, where the number of members depended on the number of students per class. In Portugal, classes’ sizes usually range from 20 to 28 students; thus, each group had, in general, five to seven members. Each group had 10 min to discuss and answer the survey. Sessions were held in two schools, at Escola Básica e Secundária Dr. Pascoal José de Mello and Escola N° 2 de Avelar, from the municipality of Ansião in the central region of Portugal, during February and November of 2020.

Figure 1. Photo of a questionnaire delivered to children. The questionnaire page shows the What we know about climate change question.

2.1. Data analysis

A qualitative and quantitative data analysis was implemented to analyse the responses to the questions What we know about climate change? and What we want to know about climate change?

For the qualitative analysis, a three-step methodology was applied: (1) Open coding; (2) Axial coding; (3) Selective coding. For the first step – Open coding – all responses, from all questionnaires’ registry forms, were transcribed to spreadsheets and an inductive code approach was applied, i.e. the codes were interpreted from the students’ responses, and not defined a priori (Corbin and Strauss 1990). Co-occurring codes were accepted, meaning that the same responses can have more than one code. An example would be the response ‘Causes global warming, deforestation, drought’ for which three codes were assigned: (1) Global warming within the category Concept of Climate Change, (2) Deforestation and (3) Natural Hazards, both within the category Climate Change Consequences.

Once the codes were assigned to the responses, they were refined and grouped into subcategories and then into a more abstract and higher level, the category level (Axial Coding Step). To guarantee that the categories’ development process was fully explored, the Corbin and Strauss (1990) Coding Paradigm was used. The Coding Paradigm defines six subcategories – phenomenon, causal conditions, strategies, consequences, context, and intervening condition – which were accounted to support the identification of relationships among codes. However, because this work is focused on the participants’ perception of the phenomenon and not on how the participants deal with it, the coding subcategories (Table 2) and the coding diagramming (Figure 2) were adapted to fit the study. In particular, it is important to notice that the Corbin and Strauss (1990) Coding Paradigm considers that people’s actions and the pursuit of their strategies have consequences (Corbin and Strauss 1990). Although this is also true for CC, one cannot ignore that CC itself has consequences that need to be accounted for when analysing students’ perception and that strategies are implemented to deal more with the consequences and the causes, rather than the phenomenon itself.

Figure 2. Coding diagram. Above, is the original Corbin and Strauss (1990) coding paradigm diagram. Below, the coding diagram used, adapted from the Corbin and Strauss coding paradigm to take into account the fact that strategies (social responses) are implemented to deal with both the causes and consequences of CC, rather than the phenomenon itself.

Table 2. Comparison between the original subcategories of the Coding Paradigm of Corbin and Strauss (1990) and the adapted categories of the current study.

Original Subcategory	Original subcategories’ description	New subcategories’ description
Phenomenon	The central idea of a category and supports exploring the what	The central idea of the category and supports exploring the ‘what’. Equal to the original Coding Paradigm.
Causal conditions	Why does this phenomenon happen	Why does this phenomenon happen? Equal to the original Coding Paradigm.
Consequences	What happens as a result of the strategies?	What happens as a result of the phenomenon? In this work, consequences are an outcome of the phenomenon (i.e. climate change), and not of the strategy.
Strategies	What actions have been taken, or what potential actions can be taken, to deal with the phenomenon?	What actions have been taken, or what potential actions can be taken, to deal with the causes and consequences? In this work, the Strategies are a means to deal with the causes and the consequences, and not with the phenomenon itself. Moreover, they may not reveal participants’ strategies, but society’s strategies.
Context	Details about the phenomenon or circumstances in which the strategies take place	Details about the phenomenon or circumstances in which the phenomenon, its causes and consequences take place. In this work, circumstances are related to the phenomenon, its causes and consequences, and not to the strategies.
Intervening Conditions	Describe attributes that may influence the strategies	Describe attributes that may influence participants’ perceptions about the phenomenon. This work is focused on the participants’ perception about the phenomenon and not how the participants deal with it. It also describes attributes that hinder the capability of the analyst to understand the relevance of the students’ response within the climate change context (unintelligibility).

At the end of the axial coding step, a three-level hierarchical scheme, with increasing detail was obtained, where level 1 stands for the categories. Level 2 for the subcategories and level 3 for codes (Tables 3–6).

Table 3. Subcategory ‘Phenomenom’.

Level 1	Level 2	Level 3
Cat.	Subcat.	Code	Notes		Quotes’ examples
Concept of Climate Change	Phenomenom	Acid rain	References to acid rain as a climate change characteristic	What we know	‘One of the characteristics of these changes is acid rain’
				What we want to know	n.a.
		Climate changes	References to long-term shifts in temperatures and weather patterns	What we know	‘Climate change: when the temperature changes’
				What we want to know	‘How is it made up or formed’
		Date framing	References to time dates: when it all started or until when will it end	What we know	‘Climate change started at the end of the thirteenth century’
				What we want to know	‘How long it takes for this process to increase’
		Extreme weather events	References to extreme weather or extreme weather events include unexpected, unusual, severe, or unseasonable weather; weather at the extremes of the historical distribution	What we know	‘In a cold season can come heat’
				What we want to know	n.a.
		Global warming	References to global warming. Global warming can be defined as the long-term heating of Earth’s climate system observed since the pre-industrial period (between 1850 and 1900) due to human activities.	What we know	‘It causes global warming’
				What we want to know	‘When and how global warming started’
		Greenhouse effect	References to the relationship between climate change and the greenhouse effect	What we know	‘It’s a type of greenhouse’
				What we want to know	‘What is the greenhouse effect’
		Radiation	References to radiation	What we know	‘Radiation’
				What we want to know	n.a.
		Weather forecast	References to the relationship between climate change and weather forecasting	What we know	‘Weather forecast’
				What we want to know	‘How does it get cold and hot’

Three-level hierarchical scheme encompassing categories, subcategories and codes at the end of the axial coding step and used for qualitative and quantitative analysis. This table includes the subcategory ‘phenomenom’. Other subcategories are in Table 4 ('causes’), Table 5 ('consequences’) and Table 6 ('strategies’, ‘context’, ‘intervening conditions’). For each code it is provided a brief description and a quote of a typical response per question (What we know about climate change? and What we want to know about climate change?).

Cat. – Category; Subcat. – Subcategory; n.a. – not applicable, because no answers were provided for the specified question within the subcategory.

Table 4. Subcategory ‘causes’.

Level 1	Level 2	Level 3
Cat.	Subcat.	Code	Notes		Quotes’ examples
Causes of Climate Change	Causes	Deforestation		What we know	‘Deforestation and the exploitation of the world’s forests’
				What we want to know	‘Is it true that the Amazon is the second lung of the earth’
		Fires		What we know	‘Fires are causing bad causes for the climate, humans and animals’
				What we want to know	n.a.
		Greenhouse effect		What we know	‘Climate change is caused by […], greenhouse effect, […]'
				What we want to know	‘How not to cause the greenhouse effect’
		Greenhouse gases		What we know	‘Climate change comes from the excess of certain gases in the atmosphere’
				What we want to know	n.a.
		Human-induced	Explicitly mentions the role of humans (human-related causes)	What we know	‘Man is the cause of climate change’
				What we want to know	‘[…] how man’s ignorance has put himself in charge of the climate’
		Ozone layer depletion		What we know	‘The ozone layer is disappearing’
				What we want to know	n.a.
		Pollution		What we know	‘Climate change is the change in the climate that exists because of pollution’ ‘Climate change is when smoke from factories pollutes the air’.
				What we want to know	n.a.
		Recycling	It was assumed that the students believed that not recycling contributed to climate change	What we know	‘Do recycling’
				What we want to know	n.a.
		Unsustainable use of resources		What we know	‘Global warming is based on the use and abuse of nature’s resources’
				What we want to know	n.a.

Three-level hierarchical scheme encompassing categories, subcategories and codes at the end of the axial coding step and used for qualitative and quantitative analysis. This table includes the subcategory ‘causes’. Other subcategories are in Table 3 (‘phenomenom’), Table 5 (‘consequences’) and Table 6 ('strategies’, ‘context’, ‘intervening conditions’). For each code it is provided a brief description and a quote of a typical response per question (What we know about climate change? and What we want to know about climate change?).

Cat. – Category; Subcat. – Subcategory; n.a. – not applicable, because no answers were provided for the specified question within the subcategory.

Table 5. Subcategory ‘consequences’.

Level 1	Level 2	Level 3
Cat.	Subcat.	Code	Notes		Quotes’ examples
Climate Change Conse­quences	Conse­quences	Acid rain	References to acid rain as a consequence of climate change	What we know	n.a.
				What we want to know	‘Acid rain comes from climate change’
		Atmosphere changes	References to changes in the atmosphere due to climate change	What we know	‘The atmosphere suffers a lot because of climate change’
				What we want to know	‘Why it destroys the atmosphere’
		Biota	References to changes in the biota due to climate change	What we know	‘Climate change causes the death of many animals and plants’
				What we want to know	‘Will some of the animals disappear (few or many)'
		Climate changes	References to changes in climate patterns as a consequence	What we know	‘The weather is less predictable’
				What we want to know	n.a.
		Deforestation	References to deforestation due to climate change	What we know	‘Causes […] deforestation, […]'
				What we want to know	n.a.
		Fires	References to the emergence of fires due to climate change	What we know	‘Causes fires’
				What we want to know	‘Why climate change can lead to more fires’
		Freshwater depletion	References to the freshwater depletion due to climate change	What we know	‘Fresh water is becoming scarce’
				What we want to know	n.a.
		Greenhouse effect	References to the greenhouse effect as a consequence of climate change	What we know	‘Climate change causes the greenhouse effect’
				What we want to know	n.a.
		Natural hazards	References to natural hazards due to climate change	What we know	‘Natural disasters used to happen rarely. Now they happen frequently’
				What we want to know	‘Tornadoes’
		Ozone layer depletion	References to ozone layer depletion due to climate change	What we know	‘Destroys the ozone layer’
				What we want to know	‘What happens if climate change destroys the ozone layer’
		Pollution		What we know	n.a.
				What we want to know	‘Why climate change causes so much pollution (for example)'
		Sea ice changes	References to sea ice changes due to climate change	What we know	‘The poles begin to melt’
				What we Want to know	‘Why the Arctic is melting’
		Sea level rise	References to sea level rise due to climate change	What we know	‘Increases sea level’
				What we want to know	‘They lead to the melting of the ice caps and the rise of the oceans’
		Social impact	The answer explicitly mentions the effects on human populations (Social impacts). Includes references to the emergence of diseases.	What we know	‘If we don’t stop this problem as soon as possible, it could greatly affect our future’
				What we want to know	‘Could it cause diseases? If yes, which diseases’

Three-level hierarchical scheme encompassing categories, subcategories and codes at the end of the axial coding step and used for qualitative and quantitative analysis. This table includes the subcategory ‘consequences’. Other subcategories are in Table 3 (''Phenomenom’), Table 4 ('causes’) and Table 6 ('strategies’, ‘context’, ‘intervening conditions’). For each code it is provided a brief description and a quote of a typical response per question (What we know about climate change? and What we want to know about climate change?).

Cat. – Category; Subcat. – Subcategory; n.a. – not applicable, because no answers were provided for the specified question within the subcategory.

Table 6. Subcategories ‘strategies’, ‘Context’ and ‘Intervening Conditions’.

Level 1	Level 2	Level 3
Cat.	Subcat.	Code	Description	Question	Quote
Social responses	Strategies	Activism	References to activist actions of persons involved.	What we know	n.a.
				What we want to know	‘Who is Greta?'
		Management	References to actions for prevention, mitigation, adaptation, and how to put an end to CC	What we know	‘If no action is taken, it will get worse’
				What we want to know	'New measures to stop it’
		Monitoring	References to monitoring actions.	What we know	n.a.
				What we want to know	‘What kind of program is used to know the climate’
Severity	Context	Negative outcome	A negative outcome is mentioned, but nothing specific is listed. It includes concepts such as damaging, harmful, bad. Includes both environmental- and social-related outcomes.	What we know	‘It can harm our planet’
				What we want to know	‘What will happen to the planet if we do not change’
		Bleak Outcome	When concerns related to long-term consequences are mentioned and there is a perception on the matter’s severity.	What we know	‘It kills the planet’
				What we want to know	‘If the world will ever end because of this’
Circumstances		Pre-conditions	When students show concern about how, where, when and why climate changes occur. It was considered these issues could not be dissociated from the causes (Causes | How, Where, When, Why)	What we know	n.a.
				What we want to know	‘Why it happens’
		Post-conditions	When students show concern about how, where, when and why the consequences of climate change are/will be felt (Consequences | What, Where, When, Why)	What we know	n.a.
				What we want to know	‘What causes climate change’
Knowledge Limitations	Intervening Conditions	Knowledge perception	When the student acknowledges that he/she does not know or wants to know more.	What we know	‘I don’t know but I will study more’
				What we want to know	‘More things’
		Misconception	An idea that is wrong, probably because it has been based on a failure to understand the concept of Climate Change or based on a total lack of knowledge	What we know	‘Destroys the ozone layer’
				What we want to know	‘How not to cause the greenhouse effect’
Analysis Limitations		Unintelligibility	Includes responses for which it is not understandable the student’s point of view in the context of climate change discussion.	What we know	‘That the sun changes to the moon’
				What we want to know	‘Why high tides occur at night’

Three-level hierarchical scheme encompassing categories, subcategories and codes at the end of the axial coding step and used for qualitative and quantitative analysis. This table includes subcategories ‘strategies’, ‘context’, and ‘intervening conditions’. Other subcategories in Table 3 (''Phenomenom’), Table 4 ('causes’) and Table 5 ('consequences’). For each code it is provided a brief description and a quote of a typical response per question (What we know about climate change? and What we want to know about climate change?).

Cat. – Category; Subcat. – Subcategory; n.a. – not applicable, because no answers were provided for the specified question within the subcategory.

Once all responses were categorised, it was possible to identify the main ideas and, specifically, how students perceive their knowledge and interests about climate change (Selective coding Step). By interpreting the information given, conclusions were drawn out of the coding with support from the coding diagram (Figure 2). Based on these, the main theory regarding students’ perception of their knowledge and interests concerning climate change was formulated.

Once the database was set, a quantitative statistical analysis was undertaken (see Supplementary material for R code). First, a Non-Metric Muldimensional Scale (metaMDS function from R package Vegan) (Kruskal 1964) was performed to search for differences between the following sets of classes: (a) individual and group questionnaires, (b) 5th, 6th, 7th, 8th grades and (c) What we know and What we want to know questions. nMDS was calculated using the cosine distance function available within the designdist function, also from the R package Vegan. The cosine distance metric is suitable to represent the frequencies of words in documents and is thus widely used in text mining (B. Li and Han 2013). The statistical significance of the differences found was analysed through a Permutational Multivariate Analysis of Variance using Distance Matrices (adonis2 from R package Vegan) (Anderson 2001). It tested the null hypothesis that the dissimilarity between classes is greater than or equal to the dissimilarity within each class. The higher the R value, from ADONIS calculation, the more dissimilar the classes are.

The results showed a strong significant dissimilarity between questions What we know and What we want to know. As such, an Indicator Species Analysis (multipatt function from R package indicspecies) (De Cáceres et al. 2010) was performed to determine which codes were most responsible for the differences between the two. The strength of the association was determined using a correlation index, which is suitable to assess the preference of the codes for each question. The equalized version of the index was used so each question would have the same equal weight in the analysis (De Cáceres et al. 2010). Details of responses were also analysed through frequency graphs.

3. Results

The first subsection presents the general trends for the responses provided. Then are presented the results that allow us to understand whether there are significant differences between the type of questionnaire (individual vs group), between the grades (5^th, 6^th, 7^th, 8^th) and between the two types of questions (What we know and What we want to know). Afterwards, are presented the codes that are most responsible for the significant differences found between the types of questions. The two following subsections provide a more detailed analysis of the categories, subcategories and codes found for each question – What we know about climate change? and What we want to know about climate change? Examples of student responses, reflecting their range and diversity, are provided to support the categorization. The previous analysis allowed to build the main theory about the students’ perception about climate change, which is then presented in the last subsection.

3.1. General trends of responses

Students tend to provide a higher number of responses (students could provide up to 4 responses per question) in group sessions than in individual sessions. Also, students tend to provide a higher number of responses for the What we know than for the What we want to know question. In addition, students from the 5^th grade delivered more responses, considering the total expected, than students from the remaining grades (Figure 3).

Figure 3. Percentage of responses for the What we know about climate change and the What we want to know about climate change questions, considering the total expected per grade and question.

The results show that: (a) there are no significant dissimilarities between the responses provided by the four grades sampled (R = 0.0552, p value = 0.18). (b) there is a small dissimilarity between individual and group questionnaires (∼4% of the variation explained by), that can most probably be disregarded as suggested by the p value > 0.05 (p value = 0.0566), and (c) there is a strong significant dissimilarity between responses given to the What we know and the What we want to know questions. Results indicate that 83% (R = 0.8302, p value = 0.0003) of the variation in distances is explained by these two questions. The nMDS plot (Figure 4) clearly shows the dissimilarity between them. The codes that are explaining the differences are explicitly indicated in Table 7. Out of the nine codes significantly associated with the What we know question, Sea ice changes as a consequence, Climate Change as concept, Pollution as a cause and Fire also as a cause, are the top codes more strongly associated with this question. Out of the three codes significantly associated with the What we want to know question, Pre-conditions as context is the code more strongly associated.

Figure 4. nMDS plot. Legend: ind – individual questionnaires; gp – group questionnaires; gr5 – 5^th grade; gr6 – 6^th grade; gr7 – 7^th grade; gr8 – 8^th grade; know – What do we know about climate change. want – What do we want to know about climate change. Codes are shown in grey. Intervening Conditions subcategory codes: Intv_unint—unintelligibility; Intv_kperc – knowledge perception; Intv_mis – misconception. Concept subcategory codes: Conc_acid – acid rain; Conc_CC – climate change; Conc_date – date framing; Conc_ext – extreme weather events; Conc_gwarm – global warming; Conc_greenh – greenhouse effect; Conc_rad – radiation; Conc_forec – weather forecast; Conc_defor – deforestation. Causes subcategory codes: Caus_fire – fires; Caus_greenh – greenhouse effect; Caus_greengas – greenhouse gases; Caus_human – human–induced; Caus_ozo – ozone layer depletion; Caus_poll – pollution; Caus_recyc – recycling; Caus_unsus – unsustainable use of resources. Consequences subcategory codes: Consq_acid – acid rain; Consq_atmos – atmosphere changes; Consq_bio – biota; Consq_CC – climate changes; Consq_defor – deforestation; Consq_fire – fires; Consq_fresh – freshwater depletion; Consq_poll – pollution; Consq_greenh – greenhouse effect; Consq_haz – natural hazards; Consq_ozo – ozone layer depletion; Consq_icech – sea ice changes; Consq_sealev – sea level rise; Consq_social – social impact. Context subcategory codes: Contx_neg – negative outcome; Contx_bleak – bleak outcome; Contx_pre – pre–conditions; Contx_pos – post–conditions. Strategies subcategory codes: Strat_act – activism; Strat_manag – management; Strat_Mon – monitoring.

Table 7. Indicator Species analysis results.

What we know about climate change
	stat	p.value
Consq_icech	0.839	0.0008	***
Conc_CC	0.835	0.0008	***
Caus_poll	0.795	0.0017	**
Caus_fire	0.745	0.0048	**
Intv_mis	0.696	0.0096	**
Conc_gwarm	0.664	0.0205	*
Consq_bio	0.618	0.0109	*
Consq_fire	0.585	0.0378	*
Consq_social	0.544	0.0038	**
What we want to know about climate change
	stat	p.value
Contx_pre	0.841	0.0008	***
Contx_pos	0.784	0.0058	**
Strat_manag	0.776	0.0008	***
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01′*’ 0.05′.’ 0.1

Shows the codes that are significantly associated with questions What we know about climate change and What we want to know about climate change, and thus explain the significant dissimilarity between the two.

stat – statistics; Consq_icech – Sea ice Changes (consequences subcategory); Conc_CC – climate change (concept subcategory); Caus_poll – Pollution (causes subcategory); Caus_fire – Fire (causes subcategory); Intv_mis – Misconceptions (intervening conditions subcategory); Conc_gwarm – Global warming (concept subcategory); Consq_bio – Biota (consequences subcategory); Consq_fire – Fire (consequences subcategory); Consq_social – social responses (consequences subcategory).

3.2. What children know and want to know about climate change

3.2.1. What we know about climate change

In general, students seem to know more about the consequences of climate change than about the causes or the concept itself (Figure 5a).

Figure 5. Percentage of responses for the What we know about climate change question considering the total number of responses in each school grade, per school grade. Results do not discriminate between individual and group session’ responses. (a) Percentage of responses by category. (b) Percentage of responses by code for the category Concept. c) Percentage of responses by code for the category Causes. d) Percentage of responses by code for the category Consequences. e) Percentage of responses by code for the categories Severity and Circumstances. f) Percentage of responses by code for the category Social Responses.

With regard to consequences, students seem to know more about the relationship with sea ice changes and biota (Figure 5d). With regard to causes, students clearly show that they are aware that pollution is a main cause of climate change (Figure 5c). Examples of responses that refer to pollution are ‘Climate change is the change in the climate that exists because of pollution’ and ‘Climate change is when smoke from factories pollutes the air’.

Although not widely mentioned, it is worthwhile to remark that some students reveal to be aware of the relationship between climate change and human population, as some responses clearly establish a link between both. This link appears from the point of view of causes (‘Man is the cause of climate change’), consequences (‘More and more climate change is hurting us’) and also social responses (1 item only) (‘If no action is taken, they will get worse’).

With regard to the concept of CC, students’ responses are more focused on references to long-term shifts in temperatures and weather patterns (‘climate changes item’) (Figure 4b) (For instance ‘The earth is getting warmer and warmer’ and ‘The seasons get changed’) and not so much on its relationship with the greenhouse effect. Some students also define it by mentioning the concept of global warming (‘Causes global warming’).

It is also relevant to mention that students, especially the 8^th graders, declare to know, and seem to be concerned about, the negative outcomes of CC – as they use words such as ‘damaging’, ‘harmful’, ‘bad’ – and, at some extent, with the bleak outcome of it, as they mention the long-term consequences and expressions indicating a perception on the matter of severity (Context). An example would be an answer indicating that ‘the world is ending’ and ‘our world is being destroyed’, ‘we are committing suicide’.

It is relevant to notice that CC circumstances (how, where, when and why) hardly emerged as topics about which the students declared to know anything about.

3.2.2. What we want to know about climate change

The variety of codes found that detail climate change causes, consequences and concept themes is greater for the What we know question, than for the What we want to know, which may justify the boom of responses that fall into the CC circumstances (how, where, when and why) category for the What we want to know question. This may be an indication that the students are interested in understanding the CC topic as a whole and not just particular topics. In particular, students seem to be curious about CC circumstances related to pre-conditions for climate change, i.e. they acknowledge they would like to know more about how it all started, how climate change actually ‘works’, without specifying any aspect in particular (Figure 6a-e). Examples are: ‘How climate change occurs’ and ‘I wonder why this happens’. This pattern is more pronounced in the individual questionnaires than in responses from group sessions. It may also be an indication that their knowledge about climate change is not enough to list specific issues other than those they already seem to have some knowledge about. Notice that students declared they would like to know more about pollution as a cause (‘Which company started to pollute the environment’), about the consequences of CC to biota (‘If animals die’) and about long-term shifts in temperatures and weather patterns as a definition of the CC concept (‘Why the seasons are getting mixed up’). These three topics emerge, at the same time, as topics that the students declare to know more about. The 5^th grade is, however, an exception, as they do not seem to want to know more about the specific pollution-problem, but seem, nevertheless, interested in knowing more about the relationship between humans and climate change causes (human-induced code).

Figure 6. Percentage of responses for the What we want to know about climate change question considering the total number of responses in each school grade, per school grade Results do not discriminate between individual and group sessions’ responses. (a) Percentage of responses by category. (b) Percentage of responses by code for the category Concept. (c) Percentage of responses by code for the category Causes. (d) Percentage of responses by code for the category Consequences. (e) Percentage of responses by code for the categories Severity and Circumstances. (f) Percentage of responses by code for the category Social Responses.

With regard to social responses, understood as social strategies to cope with climate change, students acknowledged they want to know more about management issues, i.e. those related to measures for prevention, mitigation and adaptation, as well as solutions to put an end to CC (Figure 6f). These issues are more pronounced in responses from 8^th grade students. Examples are the following responses ‘What can you do to improve’ and ‘What we can do to stop them’. Notice also, that the two references included in the ‘activism’ code, found within 7^th and 8^th grade responses, are related to Greta Thunberg: ‘Who is Greta’.

With regard to severity, although again not much mentioned, it is worthwhile to notice that although a few students declared to know something about the severity of the problem (negative and bleak outcomes), others mention they want to know more (Figure 6e).

3.2.3. Intervening Conditions

Intervening conditions describe attributes that may influence participants’ perception about the phenomenon (knowledge limitations), but also the incapability of the analyst to understand the relevance of the students’ response to the overall analysis (analysis limitations). In this study three hindering conditions were identified: knowledge perception, misconception and unintelligibility (Figure 7). Expressions falling into these categories constitute approximately 5% (from group sessions) to 7% (from individual questionnaires) of the total.

Figure 7. Percentage of responses by subcategory for the Knowledge and Analysis Limitations Categories (Intervening Conditions) considering the total number of responses in each school grade, per school grade. Results do not discriminate between individual and group sessions’ responses. a) Percentage of responses for the What we know about climate change question. b) Percentage of responses for the What we want to know about climate change question.

Knowledge perception includes expressions of students straightforwardly acknowledging that they do not know or want to know more about the subject. For example ‘I don’t know but I will study more ‘. Only three responses fall into this category (2% of total number of responses) and these only emerge within responses from individual questionnaires.

The misconception category includes ideas that are wrong, probably because they have been based on a failure to understand the concept of climate change or based on a total lack of knowledge. Misconceptions are more related to the What we know question and reveal misconceptions associated with CC consequences, causes and concept. Specifically, ozone layer depletion (‘This phenomenon is occurring due to the destruction of the ozone layer’) and the greenhouse effect (‘How not to cause the greenhouse effect?’) were both mentioned as causes and consequences of CC; and acid rain (‘One of the characteristics of these changes is acid rain’) and greenhouse effect (‘Is a type of greenhouse’) were somewhat used as a definition for climate change, and climate was mentioned as a synonym for temperature (‘Climate change is the change in temperatures’). When analysing individual and group responses all together, the percentage of misconceptions increases from the 5^th to the 6^th grade and then gradually decreases with the developmental stage up to the 8^th grade.

Unintelligible responses are those that include aspects for which it is not understandable the student’s point of view in the context of climate change discussion. The majority were given within the responses for the What we want to know question and by students answering individual questionnaires. An example would be ‘Why high tides occur at night’ or ‘Where is the plastic island’.

3.2.4. Main theory

Students are aware of the climate change phenomenon and they recognize its anthropogenic causes, as well as its consequences to humans and the environment. In fact, students realizing that they know nothing, or not enough to provide responses, about CC constituted a small percentage (knowledge limitations).

On the other hand, students struggled to identify social responses (strategies), but showed a particular interest in learning more about these. Likewise, students showed a strong interest in understanding how severe CC may be and in learning about how, where, when and why CC occurs (context). The few times students clearly specified CC causes and consequences they would like to know more about, these were, in general, the same issues that had already been stated as already learnt: pollution, biota and sea ice changes.

Students’ knowledge about CC is potentially hampered by: knowledge gaps about which the students are aware of; misconceptions with regard to climate change, its causes and consequences without the students’ acknowledgement, and difficulties in expressing arguments (written and oral communication).

4. Discussion

4.1. General trends

Overall, the CC-related issues presented are in line with the conceptual model presented by Shepardson et al. (2012), in which at least the following aspects of knowledge should be studied in climate change education: (a) natural causes and changes in the climate system (e.g. ‘We know that climate change is the increase in the earth’s temperature’), (b) atmosphere and pollution (e.g. ‘Climate change comes from the excess of certain gases in the atmosphere’), (c) amounts of snow and ice (e.g. ‘Glaciers are melting’), (d) oceans (sea level, temperature and life) (e.g. ‘Causing the sea level to rise’; ‘It is taking away the habitat of many animals (e.g. polar bears’), (e) soil and vegetation (e.g. ‘there is plant drought due to high temperature’) and (f) impact on humans (e.g. ‘It is a great danger for humanity’). Nevertheless, although students seem to be aware of the terminology of climate change, some responses indicate they are still struggling to explain and link issues in practice. In this study, the misunderstandings found, for instance those related to the emission of greenhouse gases (GHG) and the ozone layer, are the most obvious evidence. However, other responses, might already indicate a notion of feedback. For instance, when students indicate deforestation and fires as causes of CC: ‘Deforestation and the exploitation of the world’s forests’ and ‘The drivers of climate change are motor vehicles, industry and forest fires’. Deforestation is not a cause by itself, but it may potentiate CC due to the reduction in carbon dioxide absorption (Y. Li et al. 2022). Fires in turn, release greenhouse gases, which alone might not be a problem, but, in reality, their intensity and frequency might be increasing due to climate change (a consequence) and therefore might be potentiating the emission of greenhouse gases (Singh 2022).

These findings are not a surprise, due to the complexity of the theme. To deal with it, Jurek et al. (2022) propose that any approach to increase knowledge about climate change should first raise the factors underpinning the foundations of CC, and then promote the understanding of interactions to enhance the capacity to correctly interpret causes and estimate consequences of climate change. Cantell et al. (2019) even propose the ‘bicycle model’ to deal with the broad scope of CC. The ‘bicycle model’ is a holistic education model that emphasizes the importance of the following aspects: knowledge, thinking skills, values, identity, worldview, action, motivation, participation, future orientation, hope and other emotions, and operational barriers. The most interesting characteristic of this model is that, because it aims to foster critical engagement anchored in knowledge, it might actually provide a hands-on link between science education and science communication.

Long-standing research suggests that one of the issues that traditionally distinguishes education and communication is knowledge. Science education makes an optimistic assumption that knowledge content suffices for people to reach an informed, scientifically-based decision (Baram-Tsabari and Osborne 2015; Kohen and Dori 2019). Science communication, in turn, prioritises engagement because it recognises that people’s attitudes towards controversial scientific issues are deeply culturally-related and better explained by values, emotions, ideology, social identity, and trust in scientific and other institutions (Bromme and Goldman 2014; Feinstein 2012; Laslo, Baram-Tsabari, and Lewenstein 2011). Is it then possible to increase CC knowledge while evoking ‘the thrill of discovery’ and ‘telling the world about the significance of research results’ (Strauss 2005), so students may feel empowered and capable of making informed decisions? The answer may be two-fold. On the one hand, climate education should be broadened to fields of knowledge other than content knowledge, including knowledge of methodologies, the role of uncertainty and the role of communication (Ryder 2001). On the other hand, climate education ought to include pedagogical methods that foster thinking skills and allow students to resonate on their own experiences and reality (Strauss 2005; Trott 2020).

Using the results from this work, an example could be selecting pollution as a cause – an item in which students are interested in – and follow a critical thinking approach to discuss the scientific findings, and related uncertainty, that allow us to perceive it as a major cause of climate change. As pollution is driven by human actions, the behavioural habits of students and their families could also be explored allowing students to understand their role, as well as the political and economic connections.

4.2. Unanticipated and anticipated results

Evidence suggests that the accumulation of knowledge is a key factor promoting the development of memory across childhood (Brod and Shing 2022). It was thus expected that, in this study, older students would mention a larger spectrum of CC-related issues than younger students. However, the results indicate there is no dissimilarity between grades despite the increase in the proportion of responses per question from lower to higher grades from the 6^th up to the 8^th grade. The unexpectedly high proportion of responses provided by the 5^th grade was attributed to the lack of individual questionnaires. No individual questionnaires were available for the 5^th grade, and, as shown by the questionnaires for the 6^th, 7^thand 8^th grades, students tended to provide a lesser number of responses when answering individually, which would decrease the proportion of responses from the 5^th grade. This may have also contributed to the statistical results found.

It was also expected that the percentage of misconceptions would decrease with increased development (Keane and Griffin 2018), but in our case study it only occurs from the 6^th grade onwards and only due to questionnaires from group sessions. These results were in part attributed to the fact that the number of questionnaires is lower for the 5^th grade and no individual questionnaires exist for this cohort. However, if the results are not related to the number of questionnaires, and the 6^th grade is indeed more prone to misconceptions than the 5^th grade, then it might be important to follow Kumar et al. (2023) suggestion and, in the future, develop tailored activities for different age cohorts to improve learning outcomes regarding CC.

Despite not being significantly associated with the What we want to know question, the codes sea ice changes, pollution and biota are among the causes and consequences that the students refer more frequently they would like to know more about. On the contrary, these issues are significantly associated with the What we know question. This was unanticipated but not a complete surprise, especially because our target audience are children. Due to their young age, it is plausible to assume that many of the participants in the questionnaires have not yet encountered several of the complex issues around CC. In such a case, it is plausible to assume that they are unaware of their ignorance and, to some extent, incapable of recognizing that they do not know something (İnan 2020) and thus are expressing more frequently issues they are familiar with. Supporting this idea is also the fact that the number of responses for the What we know question is higher than for the What we want to know question. In some ways, this contradicts the rule of thumb that researchers should not rely on their judgment to select the themes that should be transmitted to children (Bruine De Bruin and Bostrom 2013). If researchers rely only on children’s knowledge and interests, which many times come from previous interactions with family and social media, some important concepts and interconnections might never be presented to children, creating ‘information corridors’ (Lopakov, Prokopovych, Solodkyi 2022).

However, not accounting for children’s previous knowledge and interests might lead to failure, as engagement is more easily compromised (Dong, Jong, and King 2020). Nevertheless, one should not disregard the idea that students although already aware of the relationships between CC and pollution, biota and sea-ice changes, might be willing to understand them better. Further studies could explore what exactly students know, or do not know, about these responses, and also why they seem to be so appealing.

Considering the timeline, the fact that Greta Thunberg was mentioned only twice – one directly questioning who she was – was also unexpected. This work was implemented in early 2020 and Greta Thunberg appeared very frequently in the media in the two years before. Greta Thunberg is a climate activist who became well-known after protesting outside the Swedish parliament in 2018, inspiring thousands of young people across the world who joined her by skipping school to protest (Buettner 2020; Pickard 2021; Tattersall, Hinchliffe, and Yajman 2022). In 2019, she sailed across the Atlantic in a journey that also crossed Portugal, to attend a UN climate conference in New York to deliver her most famous speech. Also in 2019, she received the first of three Nobel Peace Prize nominations for climate activism. During this highly active period of Greta Thunberg the eldest students of this work were about 12–13 years old, and the youngest around 9–10 years old. Although there is evidence that news consumption by children is low (Edgerly et al. 2018), they are exposed to ‘second-hand’ news from those around them (Cantor and Nathanson 2006; Davies 2008) and learn a lot more about the public sphere, than previously believed (Carter 2013).

Conversely, CC misconceptions related to the greenhouse effect, the ozone layer and acid rain were anticipated. Chang and Pascua (2016), Liarakou, Athanasiadis, and Gavrilakis (2011) and Shepardson et al. (2011) had already found misunderstandings concerning the greenhouse effect process; whereas others had previously shown findings in which participants consider the greenhouse effect is equal to global warming (Andersson and Wallin 2000; Boylan 2008) or to the ozone layer (Acikalin 2013; Kilinc, Stanisstreet, and Boyes 2008; Yazdanparast et al. 2013). These are essential concepts to understand and assimilate the high complexity of climate change, but their relationship with CC remains misunderstood.

Ozone depletion, Greenhouse effect and Acid Rain are linked in several ways to climate change, but are not considered major causes, or even consequences, of climate change. Atmospheric ozone changes may potentiate or reduce the effect and impact of climate change because Ozone losses in the lower stratosphere may have a cooling effect on the Earth’s surface (Hufnagl et al. 2023; Neale et al. 2021) and ozone increases in the troposphere contribute to the ‘greenhouse’, and thus heating, effect (Liu et al. 2022).

The Greenhouse effect is a phenomenon necessary for human survival, as it allows maintaining a mild temperature on the Earth’s surface. The problem is that the extraordinary emission of GHG to the atmosphere, such as carbon dioxide (CO₂), methane (CH₄) and nitrous oxide (N₂O), is increasing the greenhouse gases layer, increasing the greenhouse effect and consequently the Earth’s temperature, which leads to changes in Earth’s climate. So, the issue is not the existence of the greenhouse effect – which is not a synonym of climate change – but its intensification.

Acid Rain is a regional phenomenon caused by toxic industrial gases, including sulfur dioxide (SO₂), nitrogen oxides (NO_X) and chloride (Cl^-), which break down in the atmosphere into their respective acids. Whereas climate change is a global phenomenon caused by the increase in GHG concentrations since the Industrial Revolution (Tavassoli and Kamran-Pirzaman 2023). Although different, the two are linked. Reducing the use of fossil fuels to reduce the emission of GHG Carbon Dioxide (CO₂) will have a positive effect on the reduction of acid rain-related gases also produced from the burning of fossil fuels, such as sulfur dioxide (Reddy and Venkataraman 2002). Also, the recovery of surface waters from acidification over decades following reductions in atmospheric sulfur dioxide and nitrogen oxide emissions, may be compromised due to CC because it could limit the improvements in acid-base status (Shao et al. 2021).

4.3. Climate change awareness

The results demonstrate that students have some consciousness toward the CC problem, including human interactions and effects. Although only a few explicitly mentioned humans as a cause of climate change, most of the responses related CC to pollution, and thus one can infer that the students are aware that human action drives climate change. Nevertheless, the role and engagement of humans towards prevention, mitigation and adaptation measures still seem very unclear. Students declare they want to know more about what type of social responses exist, but seldom (1 answer) declare to know anything about social responses to climate change. This is aligned with the results obtained by Ratinen and Uusiautti (2020) who found ‘that students [in Finland] understood incompletely the issues around climate change and its mitigation and adaptation’. At the same time, the students seem to be aware of the severity of the problem and are interested in the social impacts. This raises questions as to what directions education and communication initiatives should take since CC can be seen as a psychological threat in the sense that it causes anxiety and may lead to declining mental well-being among young people (Léger-Goodes et al. 2022; Martin et al. 2022; Vergunst and Berry 2022). And this is a consequence not only of the CC information received at school, but also of the information received indirectly through media and discussions with friends and parents (Sanson et al. 2018). In these circumstances, there is a need to support children’s knowledge, while at the same time avoiding contributing to exacerbate their uncertainties and anxieties. Trott (2020) suggest the way forward is to increase knowledge using strategies that potentiate engagement and drive children’s empowerment. When designing such strategies one should take into consideration that pro-environmental engagement seems to be explained by factors such as age and gender (Keith et al. 2021; Liefländer and Bogner 2014), knowledge (Ienna et al. 2022; Meinhold and Malkus 2005; Zeng, Zhong, and Naz 2023; Zhu et al. 2022), social influence (Lazaric et al. 2020; Simiyu et al. 2022), habits (Gkargkavouzi, Halkos, and Matsiori 2019; Sarmento and Loureiro 2021; Sarpong and Amankwaa 2022) and constructive hope (Ojala 2012 2015; Ratinen 2021). Constructive hope, in particular, is positively related to environmental engagement and seems to play an important motivational role (Ojala 2012, 2015). Constructive hope is defined as hope ‘based on positive re-appraisal, trust in different societal actors, and trust in the efficacy of individual action’. It has been defined as opposed to the concept of ‘denial hope’ in which ‘one does not think that climate change is as big a problem’ (Ojala 2012). For the particular case of climate education, instilling constructive hope seems to play a vital role not only in fostering pro-environmental engagement, but also to counteract the negative emotions found in this study and previous ones (Léger-Goodes et al. 2022; Martin et al. 2022; Vergunst and Berry 2022) which are at the basis of children anxieties. Climate education should thus include a strong hope dimension by implementing strategies that allow students to switch perspective and put trust in more powerful societal actors (Meaning-focused coping strategy (Ojala 2016)), as well as strategies that increase awareness, the feeling of shared responsibility, and the belief that CC can be mitigated by changing our behaviour (Ratinen 2021; Ratinen and Uusiautti 2020). Although Ratinen (2021) found constructive hope to be weakly correlated with students’ confidence in their own climate change knowledge, knowledge is still a predictor of pro-environmental engagement and thus, as suggested in the previous section, combined methods that foster knowledge growth built upon students’ cultural reality while instilling constructive hope could be a particularly useful overall strategy. This approach might be rather feasible to implement by science communicators and educators after the definition of a suitable pedagogical strategy. But for researchers, which only engage in climate change communication initiatives from time to time, might not be so obvious and is definitely more challenging. Capacitating researchers to engage with non-academic audiences, in particular children, is thus a crucial step towards a combined approach that celebrates science and encourages the critical evaluation of CC-related scientific endeavours (Baker 2019; Cormick et al. 2015; Meenakshi 2021), while at the same time motivates for active citizenship.

5. Conclusions

This study relies on a prior-knowledge self-assessment followed by a qualitative analysis under the assumptions of the Grounded Theory to understand whether self-perception of knowledge and interests regarding climate change, of children between 10 and 13 years old, can support the design of climate change communication initiatives, especially in educational contexts.

There are significant dissimilarities between what children know about climate change and what children want to know about climate change. Children are quite specific in topics they declare to have some knowledge about, which are related to consequences, causes and the concept of climate change. On the contrary, tend to declare they would like to know more about how, what, when and why climate change occurs. And when specifying topics they would like to know more, these are the same for both questions (pollution, biota, sea ice changes). This may be an indication that their knowledge about climate change is not enough to request learning about specific issues other than those they already seem to have some knowledge about.

As such, the main conclusion, is that although previous knowledge and interests should be accounted for when designing science communication initiatives in educational contexts – as suggested by the literature – the initiatives should avoid ‘information corridors’ potentially caused by children’s conditioned interests. Nevertheless, it is suggested that the solution should not be relying on researchers’ own scientific alignments. Rather, the solution should rely on initiatives that broaden fields of knowledge introducing innovative approaches that go beyond researchers’ pre-assumptions – such as presenting methodologies and uncertainty in climate change science -, while at the same time resonating with children’s experiences and reality and instilling constructive hope.

Disclosure statement

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

Additional information

Funding

This work was supported by FCT – Fundação para a Ciência e Tecnologia under grants UIDB/04292/2020 awarded to MARE, LA/P/0069/2020 awarded to the Associate Laboratory ARNET, UIDB/04035/2020 awarded to GeoBioTec, 10.54499/CEECINST/00152/2018/CP1570/CT0007 awarded to Zara Teixeira, UI/BD/150952/2021 awarded to Cátia Marques and SFRH/BD/147777/2019 awarded to Carlos Gonçalves. It was also supported by Fundo Azul [FA_06_2017_026]; and EEAGrants/Norway Grants [EEAGrants/Call1/2020/11].