Is there deep learning on Mars? STEAM education in an inquiry-based out-of-school setting

ABSTRACT

Learning intervention based on a “Mars and Space” exhibition was designed according to STEAM-education (Science, Technology, Engineering, Art and Mathematics) principles and practices in order to bridge the gap between formal and informal learning. The cognitive learning of 12-year-old students in Finland (N = 306) showed a sustained level for a six months period. The results of this study provided evidence that situational motivation was enhanced by interest in school science in the STEAM science exhibition context. This led to better cognitive learning results in the post-knowledge test. Thus, interest and situational motivation were the first steps, and the superficial situational motivation seemed to successfully change into content-based intrinsic motivation with longer-lasting, deep learning outcomes. STEAM intervention apparently produced long-term learning, and this exhibition learning setting is shown to provide an appropriate platform to reach the deeper layers to successfully retain knowledge. Boys’ scores rose even in the delayed test. Using structural equation modelling (SEM) to assess the effects on individual, motivational and situational interest in learning, situation motivation is shown to work as a catalyst and acting as a catalyst and also acted as a stepping stone for intrinsic motivation as part of relative autonomy (RAI) and a deep learning strategy.

KEYWORDS:

• STEAM
• cognitive learning
• gender
• informal learning
• situational motivation
• science centre pedagogy
• deep learning

Introduction

One of the strongest growing trends related to education in science, technology, engineering and mathematics (STEM) is the additional integration of Art as a skill (STEAM), which presumes that creativity (Miller & Dumford, 2016) and aesthetic aspects support traditional STEM education (Burnard et al., 2017). This combination is regarded as an essential aspect for further improving science education both in Europe (EU, 2015) and globally (Yakman & Lee, 2012). In general, STEAM-approaches in educational practices integrate traditional science, technology, engineering and mathematics education with art (BERA, 2017; Lähdesmäki & Fenyvesi, 2017). The “learning by doing” principle (Dewey, 1938) followed by science centre pedagogy (Oppenheimer, 1968) with its interactive hands-on methods were some of the first attempts to promote students discovering creative solutions based on experimentation and observation (Hein, 1990). Theoretically, these efforts were supported by important findings related to children, especially at the concrete operational stage (Piaget, 1970). The aesthetic elements of handicraft and art promote understanding of scientific and more abstract concepts by exposing students to concrete space and shape experiences (Dewey, 1980; Mack, 2006).

Informal education focuses on learning outside the formal education system as defined since the 1960s mainly by the UNESCO report “Learning to be” (Faure et al., 1972). This terminology concerning the separation of formal education and informal learning has been accepted in the literature for decades (Coombs, 1985; DeWitt & Archer, 2017). However, although it has often been considered only as a criticism of school systems (Gardner, 1991; Illich, 1971), since the 1990s it has become a widely accepted (Bell et al., 2009) and integrated part of school education, especially in regard to science education (Fenichel & Schweingruber, 2010). Out-of-school education is a term included in school legislation, and it refers to using informal education sources for formal education (Rennie, 2014). It forms a pedagogical link between formal education and informal learning (Braund & Reiss, 2004; Stocklmayer et al., 2010). Since the 2010s, digitalisation has been offering a wide range of opportunities for bridging the gap between formal education and several fruitful informal learning sources that surround the physical school building (Salmi, Thuneberg, & Fenyvesi, 2017). Simultaneously, during the last decade the number of research reports related to informal science learning has been growing rapidly, and beliefs about its effectiveness in enhancing learning and not only motivation is no longer based only on anecdotal evidence (Rennie, 2014).

Scruggs and Mastropieri (1993) have shown that when students have the possibility to use the hands-on method, they tend to like learning more and to remember better. Further, they consider hands-on to be more effective for their learning than traditional classroom teaching methods, and especially more than learning only by seeing or hearing. Also, teachers have been shown to rate the hands-on method as the most effective method for their students (Ballantyne & Packer, 2009). This has especially been found to be true for diverse learners with learning difficulties (Brigham et al., 2011). They seem to benefit from this type of visual Augmented Reality learning with the STEAM-approach (Ibanez & Delgado-Kloos, 2018; Salmi & Thuneberg, 2017). Similarly, the integration of liberal arts into STEM is reshaping both scientific and humanities education. The voluntary, problem-solving orientation of the students can form the basis for the integration of learning in transdisciplinary educational frameworks, such as STEAM-integration (Fenyvesi et al., 2015; Thuneberg et al., 2017).

Context of learning: motivation and interest

Self-determination theory (SDT) provides a theoretically validated and practically reliable measure of motivation (see Deci & Ryan, 2002). It offers a dialectical framework for understanding how students’ inner resources and learning environment factors are interconnected. The learning environment can either thwart or enhance intrinsic motivation and integration of external motives by controlling behaviour or supporting autonomy (Reeve 2002). Thus, these factors are regarded as essential where the learning context and especially an informal environment form the basis of analysis.

The SDT theory defines motivation as a continuum (Deci & Ryan, 2002): The gradual move from amotivation (not motivated at all) to the external motivation style means that concrete incentives and avoidance of punishments act as motivators. The next stage is introjected motivation, in which those incentives or punishments are symbolic, and motivation is based on experienced pressure. In turn, in identified motivation students accept external goals because they believe that such goals are beneficial for learning. The most autonomous form of motivation, at the end of the continuum, is intrinsic motivation. Then the task is interesting as such, and students engage because they enjoy it.

When a learner is intrinsically motivated, no other person persuades the learner to learn. It leads to deeper learning, creativity, higher achievement and more volitional and greater persistence, especially in tasks which require conceptual understanding (Reeve, 2002; Ryan & Deci, 2000 ). As Görlitz (1987) has pointed out, play, exploration and curiosity are regarded to substantially enhance a child’s cognitive development. Externally motivated students, in turn, have been described as ineffective learners in informal learning settings (Holmes, 2011; Oppenheimer, 1968).

Informal science learning and out-of-school education are related to cognitive learning and motivation (Rennie, 2014). According to the literature and meta-studies (Thuneberg & Salmi, 2018), the role of situational motivation is essential in out-of-school settings. A novel setting, an impulsive environment, strong effects and new persons influence the learning behaviour. In a science exhibition, all these major elements supporting situational motivation exist. Thus, while guiding and facilitating learning in different contexts, interest has proven to be an essential affective factor in varying learning contexts (Bamberger & Tal, 2009; Renninger & Hidi, 2011). Interest is indirectly related to learning and achievement outcomes, but the mechanism of the influence is relatively complex: (1) interest is related to affective response, (2) the affective response might create persistence and (3) the persistence leads to better learning outcomes (Ainley et al., 2002). The large-scale OECD assessment report (OECD, 2013) linked attitudes and interest as important variables and confirmed earlier studies and meta-analyses (Hong, 2010). Situational interest or motivation, which is often awakened in science centre contexts (Salmi & Thuneberg, 2017) influences learning outcomes, even though this is a perfect context for instructors to “catch” and also “hold” students’ interest by manipulating the learning environment as suggested by Urdan and Turner (2005).

Situational motivation and interest are good positive predictive values when trying to understand cognitive performance and possible related gender differences, students tend to sustain an aversion towards STEM-topics (Gómez-Chacón, 2000). In several studies (e.g. Farenga & Joyce, 1999; Frenzel et al., 2010), boys have often shown stronger interest towards and feel greater enjoyment doing STEM-subjects than girls (Frenzel et al., 2007). This seems to be true even when there are no major differences in their school achievement level, performance scores or test points. Thus, there is now growing (Burnard et al., 2017) and encouraging evidence (Yakman & Lee, 2012) for utilising Art and skills to advance science education via STEAM-efforts (Thuneberg et al., 2018).

Deep learning strategies were first introduced in the late 1970s as the converse of superficial learning strategies in the context of activity learning (Marton & Säljö, 1976), content motivation (Jidesjö, 2012; Marton, 1983) and informal learning (Salmi & Thuneberg, 2017). During its recent renaissance, deep learning is usually associated with the concept of achieving excellence at school through an appropriate educational system (EU, 2015). Generally, it is described as a process which successfully allows a transfer of newly acquired knowledge into existing individual deep networks of knowledge management (NRC, 2012). Learners acquire proficiency beyond just memorising facts or concepts, techniques or procedures. This includes appropriate applications of existing knowledge bases to new real situations. Thus, it combines a deeper understanding of core academic contents with the ability to apply that understanding to novel problems by requiring a range of competencies (AIR, 2015). In fact, deep learning prepares learners to be or become successful citizen in their adult lives. However, linking the long-term educational outcomes sufficiently with influential factors is still an issue of frequent discussion. What makes the approaches for advancing deep learning strategies in education even more topical (Marton, 1983) is the growing number of superficial learning strategies related especially to digital learning (Salmi, Thuneberg, & Vainikainen, 2017).

The objectives of our study were as follows: (1) to assess the cognitive learning outcome after participation in our inquiry-based out-of-school setting and to monitor the sustainability of the achieved learning effect over a half-year duration and (2) to identify influential variables which promote cognitive learning.

Method

There were 306 participants, of which 55 per cent were boys. The students were 11–13-years-old (mean 12.2 years). They were from eight schools in the Helsinki metropolitan area in Finland and were randomly selected from the more than 2 000 students representing this age group that attended the exhibition as part of routine out-of-school education according the official school curriculum, which in Finland (FNCC, 2014) gives a lot of freedom to teachers to utilise informal learning settings, such as science centres.

Description of the Mars-exhibitions of the present study

The context of this study was a mobile interactive mathematics exhibition called “Mars & Space” that was supported by a planetarium programme with the same topic. Entire school classes were taken to it as part of their STEAM education. The students visited it in order to acquire knowledge and skills that would support the curricular STEAM learning goals of 12–13-year-old sixth graders. The exhibition was originally designed by the Ontario Science Centre (Canada) and later-on modified for Heureka, the Finnish Science Centre. The exhibition consisted of 30 interactive, “hands-on”, concrete and digital exhibition objects with topics relating to basic physics, astronomy, biology and psychology. The key idea was to test and challenge visitors’ own capacities, knowledge, attitudes and even willingness to participate in a journey to Mars – and back! Students were allowed to use, test, explore and learn in their own way during a 90-minute timeframe. After that, they attended a 25-minute planetarium presentation about the basics of the Solar System and the Universe. The exhibition guide only spoke in an introductory and tutorial role. The classroom teacher was only responsible for practical arrangements.

Measures

There has been much discussion (Bitgood, 1988; Falk & Dierking, 2002) about what methods to use in the research of science education and especially in informal learning settings. For example, in the Science 323 thematic number “Making the science of Education” (Alberts, 2009), one of the key findings was that researchers should use standardised measurements and methods especially in an informal learning context (Greenfield, 2009). Following those recommendations, this study applies test methods and measurement tools that have already been successfully utilised in several formal and informal learning contexts in meta-studies (Thuneberg & Salmi, 2018).

Deci and Ryan (2002) autonomous motivation: Testing autonomous motivation was based on Self-determination theory (SDT). It was only administrated as a pre-test because, after a short visit, there should not be big changes in the overall motivation, which is related to the whole personality. The Deci-Ryan Motivation (SRQ-A: Self-Regulation Quality – Academic) test includes 32 standardised items on a Likert scale (1–4). It has been used for several decades internationally (Kaplan, 2008) and nowadays has also been used several times in informal learning contexts. An example of a typical item is as follows: “Why do I try to answer hard questions in class? – Because I want the other students to think I'm smart.” (1 = not true at all, 2 = not quite true, 3 = somewhat true, 4 = totally true.). The summative variables are located on the self-determination continuum in order from external to intrinsic: External, Introjected, Identified and Intrinsic.

The SRQ-A test includes a formula (Ryan & Connell, 1989) by which the individual RAI (Relative Autonomy Index) is calculated. We used this score as an indicator of the level of autonomous motivation. A positive RAI score indicates a rather autonomous experience, and a negative score an externally controlled person who relies more on adults or others than on oneself. The reliability of the test was α = 91 (32 items).

Situational motivation was measured using a questionnaire consisting of 15 Likert scale items (scale 1–5, totally agree – totally disagree). The questionnaire was administered as a post-test only. This test provided information about how attractive the students found the exhibition. The reliability of the test was α = .86 (15 items) (item MJ6 was left out because including it would have decreased the reliability).

Cognitive Visual Reasoning was monitored by responding to the Raven test. Subjects’ general cognitive competences were measured by the visual reasoning test Raven Standard Progressive Matrices, which addresses the capacity to learn and the capacity to embrace and remember knowledge once learned (Raven et al., 2003). This test has been successfully utilised earlier in several formal and informal learning contexts and in meta-studies (Thuneberg & Salmi, 2018). In each item, the subject is asked to identify the missing element which completes a pattern. The test consisted of five sets (A, B, C, D, E) of 12 items each. To avoid cognitive over-loading, the time limit to complete the test was 12 min. All items were coded dichotomously as correct or incorrect (including the items not completed). The final test score was calculated as a sum of correctly answered items within the time limit. The reliability of the test was good (α = .80; 60 items).

The knowledge test consisted of 45 ad-hoc items related to the content areas of the school curriculum (FNCC, 2014) combined with the Mars and Space exhibition. All knowledge tests contained the same items (in the pre-, post- and delayed post-test phases). However, the participants were never aware of the testing cycles. The first questionnaire (T-0) was completed two months before participation. A subsequent pilot test forced the dropping of five items (due to being too easy, too difficult or irrelevant). An example of a typical item: “Gravity is stronger on Mars than on Earth.” The answering options to the statements in the test were 1 = true, 2 = untrue, 3 = I don’t know. According the research literature and meta-analysis (see Thuneberg & Salmi, 2018), this type of test related to “uncertainty in knowing” has turned out to be successful especially in an out-of-school context. Subsequently, only correct answers were analysed. The post-knowledge test (T-1) was monitored two weeks after participation, and the delayed post-test (T-2) six months later. The reliabilities were calculated α = .66 for T-0, α = .70 for T-1 and α = .57 for T-2 (33 items).

The pupils’ task was to judge whether the statements were correct or incorrect. The characteristic feature of the approach of present study was that the students also had the option to say that they do not know the answer. The final test scores for pre- and post-test were calculated by summarising the remaining items. The knowledge measures showed to be reliable enough. The pedagogical idea of a science centre exhibition is to give the visitor opportunities to learn according their own motivation, interest and “free-will”. It differs from the school curriculum approach according which everyone should learn all from the restricted topic.

The school achievement variable was the summary of four school grades (Physics, Chemistry, Mathematics and native language). All students were classified into three categories according to their school achievement level: A+ = Above Average School Achievement (A+; 25% of the students in each class), A = Average School Achievement (A; 50% of the students in each class) and A- = Below Average School Achievement (A –; 25% of the students in each class)

Monitoring the learning context: school vs. science centre

The Semantic Differential method (Osgood, 1964) was used for measuring students’ interest towards learning science in a school settings and in the science centre. They had to evaluate 14 pairs of adjectives on a five-point scale (e.g. studying science in school is interesting … boring). The reliability of the school science interest scale (pre-test) was α = .85 (14 items), and for the science centre science (post-test) it was α = .87 (14 items). The questions were the same, but the students had to relate their attitudes in the pre-test with the “school science learning” context and in the post-test with the “science centre learning” context.

We used path modelling (AMOS 22) to analyse the observed data regarding the theoretical assumptions. The variables for the model were chosen, but based on the theory, as well as on bivariate Pearson’s correlations. Gender, RAVEN, RAI and attitude towards science learning in the school learning context were used as covariates to control the effects on the measured pre- and post-knowledge test variables, situational motivation (only as a post-test) and attitude towards science learning in the science centre learning context. The goodness of fit evaluation of the models was based on a χ²-test and NFI, CFI (good fit > .90, or better >.95) and RMSEA (reasonable fit < .08, good fit < .05) (see Byrne, 2010).

In order to enhance the interpretation of the path-analysis results, it was necessary to obtain knowledge of the pre- and post-test means and differences between the groups, as well as the significance of the change. One-way and multivariate analysis of variance and GLM repeated measures were applied for these purposes (Bakeman, 2005). The effect size measure was partial η² (interpretation: > .01) small, > .06 middle, > .14 large (Cohen et al., 2001, p. 272). For analysis of the categorical variables (gender/school achievement A-, A, A+/asking voluntary question) we used Cross-tabulation and the Chi square test.

Missing values: On average there were 8.8 per cent missing values in the variables studied. The missing values most in the School achievement variable, in which this information was not available for 22 per cent of students.

Results

The GLM Repeated measures analysis of the three time-point tests showed that there was a significant learning effect in knowledge (partial η² = .792); the differences: pre/post knowledge test (p < . 001), pre/delayed test (p < .001); post/delayed test non-significant (p = .203). (Figure 1).

Figure 1. Pre, post and delayed knowledge test results of boys and girls.

Boys did differ in the pre-test compared to the girls, F(df1, 294) = 6.869 (p < .01), but this was not the case in the pots-test (p = .775) or in delayed post-test (p = .185). However, in the boys’ group all the changes were significant (pre/post-test, p < .001; pre/delayed, p < .001; post/delayed, p < .035), whereas in the girls’ group the difference between the post- and delayed knowledge test was non-significant (p = 1.00).

The school achievement groups did not differ in any of the knowledge-tests: Pre-test (p = .449) and post-test (p = .382), delayed post-test (p = .669). (Figure 2).

Figure 2. Motivation and curiosity of boys and girls based on the voluntary questions.

The questionnaire also contained one voluntary item where the students could ask the guide person at the Mars exhibition one or more questions. This item described the curiosity and the strength of motivation for the topics of the exhibition.

It turned out that the girls who asked questions did not differ from the girls who did not ask a question in the pre-test, p = .540 and in the delayed test, p = 987, but they received higher post-knowledge scores, p < .01, partial eta sq = .076. Based on the Chi-square test the girls that asked questions did not differ from the other girls regarding school achievement level (A+; Average; A-). However, the one-way analysis of variance test showed that they were more autonomous and liked the science education at school more than those girls who did not ask a question.

The boys who asked their own questions did not differ in their knowledge scores from those who did not ask a question (pre, p = .334; post, p = .602; delayed, p = .391).

The bivariate relationships of the variables were analysed by obtaining the correlations (Table 1). To elaborate the analysis further, a structural equation path model was created. Using SEM, it was possible to discover the complexity of the relationships and their relative effects as knowledge learning predictors.

Table 1. Correlations of the variables studied.

	Know T0	Know T1	Know T2	RAI	Raven	School Context	Science Centre Context
Know T1	.53***
Know T2	.57***	.51***
RAI	.10	.11	.01
Raven	.07	.00	.11	−.08
School Context	.26***	.23***	.07	.16**	−.03
Science Centre	.16**	.08	.15*	.03	.09	.27***
Situation Motivation	.21**	.26***	.21**	.19**	−.23***	.33***	.31***

The predictors of the knowledge change when the cognitive and motivational effects were taken into account are presented in Figure 3. The final model with only significant paths fitted the data well (χ² = 17.9, df = 17, p = .397; NFI = .955, IFI = .998; TLI= .994; CFI = .998, RMSEA = .013).

Figure 3. The SEM path model of the Mars and Space learning project.

The analysis in the pre-test situation showed a significant moderate size relationship between the variables of school learning context and pre-knowledge. School learning context also had a small correlation with relative autonomy (RAI). Being a boy was connected to higher pre-knowledge test results and being a girl with the higher relative autonomy test (NB. The level of significance of the last mentioned was p = .06, but the effect was included due to improvement of the model). Cognitive reasoning (Raven) had no statistically significant connections to the other pre-test variables.

Because these pre-test relationships were controlled, it was possible to discover their “purified” role as predictors on the post- and delayed post-test variables:

Knowledge test: Pre-knowledge predicted both post- and delayed post-knowledge. The pre-test knowledge also predicted situational motivation. Post-knowledge predicted delayed post-knowledge. However, the learning was also predicted directly or indirectly by the other studied variables in addition to the previous knowledge.

Cognitive reasoning had a negative effect on situational motivation but a positive smaller one on the science centre learning context and on the delayed post-knowledge test.

Relative Autonomy (RAI) predicted delayed post-knowledge scores. It also predicted situational motivation (NB. The significance level of the effect was p = .058).

Science learning contexts: The school learning context predicted positively situational motivation and science centre learning context, but it predicted negatively delayed knowledge-test.

Science centre learning context effects did not reach significance on any of the variables.

Situational motivation had a moderate effect on Science centre learning context but smaller effects on post- and delayed post-knowledge.

Gender: Being a girl had a positive effect on situational motivation.

In addition to the above analyses, the moderation effect of gender was analysed in order to find out whether the same model was appropriate for both girls and boys. The moderation effect was non-significant and, thus, the models were invariant (Chi-square difference between the unconstrained and the fully constrained model = 6.00, p = .97).

The variables of the model explained 29 per cent of the post-knowledge test results and 43 per cent of the delayed knowledge results. One fourth of the variance of situational motivation was explained by the model and 22 per cent of the science centre learning context.

Discussion

The results indicate several optional routes for this process of long-persisting outcomes in learning. The main dilemma is that so many things have happened to the pupils in the school – and especially outside the school – during the half-year period. We cannot confirm that the results in the delayed tests were an effect of the exhibition. However, it is even more difficult to deny that the intervention would not have had an impact, inspiration and motivation to further study the topics related to Mars, Space, technology and science also in their leisure time.

The detailed results presented in the previous chapter seem to meet the objectives of the study “to assess the cognitive learning outcome after participation in an inquiry-based out-of-school setting and to monitor the sustainability of the achieved learning effect over a half-year duration”, as the study’s main outcomes shows that the exhibition visit did enhance students’ cognitive learning both immediately after the exhibition visit and also in the delayed post-test phase six months later.

According to the objectives of the study we also tried “to identify influential variables which promote cognitive learning”. This happened by studying correlations and analysing predictors of the applied Structural Equation Path model. Based on the data the key findings of the indicators supporting the objectives are presented here (1–3):

1. Prior knowledge plays a key role in this process. Hence, the more the students knew before the exhibition, the more they learned and could retain what they learned, even in the delayed test situation six months later. However, although the biggest predictor of learning is previous knowledge, there are also cognitive, motivational and interest factors. In addition, gender also plays a small role in the interplay created by the STEAM-experience involving visual and concrete hands-on elements and collaboration with peers.

2. It is perhaps somewhat unexpected that the higher the reasoning ability, the less the students were motivated by the moment, the science centre and the Mars and Space learning situation. However, higher reasoning had a positive connection with the general attitude towards learning science in the Science centre context. We could not detect a direct indication that higher cognitive reasoning would connect to other knowledge test phase results, but it was found to have a small effect on the long-lasting retainment of knowledge.

3. Our third result contributes to the gender discussion. Before participation, boys had slightly higher science knowledge scores than girls. This difference vanished in the post-tests, which implies that girls learned more effectively. The Mars and Space science exhibition attracted girls somewhat more than boys. There was a slight indication that girls were more autonomous than boys. Because the higher RAI was connected to higher situational motivation, it is possible that the motivational effects were partly involved with the fact that there was no longer a difference between the girls and the boys in the knowledge test results after the Mars and Space visit. Consequently, our setting provided the appropriate environment to lower the traditional gap.

Our results show that liking science learning at school enhanced science learning and showed up in the higher pre knowledge-test results. Liking science learning at school also predicted aroused enthusiasm for the Mars and Space exhibition, and through that indirectly better post- and delayed post-test results. Although both liking science at school and situational motivation further predicted an attitude of liking learning at the science centre, this had no additional statistically significant role on the knowledge learning results.

Our major result, which points to long-term cognitive learning, contributes to a scarce body of literature. Using the STEAM-approach (Sochacka et al., 2016), students learned both the facts and scientific inquiry in the Mars-exhibition. This is both an obvious and a nearly naïve notion: immediately after of the educational activities students remembered at least some new facts. In fact, the short-term learning in the post-test after two weeks can be attributed to remembering and not deep learning. There is little research related to the long-term effects of informal science learning (Falk & Dierking, 2016). Tenacity of memory is regarded as a substantial attribute for reaching deep learning in out-of-school settings (Schmid & Bogner, 2015).

Another successful option that is thought to sustain learning effects lasting for six months is to build upon some kind of repeating learning content in order to help consolidate newly constructed knowledge (Salmi, Thuneberg, & Fenyvesi, 2017). An essential precondition is to avoid cognitive overload (Meissner & Bogner, 2012), which due to novel surroundings especially in informal settings might occur (Rennie, 2014). Nevertheless, our STEAM intervention apparently produced long-term learning, and this is why our exhibition learning setting is shown to provide an appropriate platform to reach the deeper layers to successfully retain knowledge. Of course, our setting is just one of many others. However, generating one successful disseminationmay help to support others.

As second major result, the importance of situational motivation is apparent for reaching long lasting cognitive learning (Bamberger & Tal, 2009). As novelty has already been shown to be one of the principal factors in encouraging learning (Braund & Reiss, 2004; Rennie, 2014), also in our case new environment contributes to that variable (Braund & Reiss, 2004; Falk & Dierking, 2002; Salmi & Thuneberg, 2017; Zoldosova & Prokop, 2006). It happens through curiosity and involves both external and intrinsic factors. Situational motivation is short-lasting, attention tends to be orientated to irrelevant subjects, and learning can easily lead to superficial results (McClelland, 1951). However, it also enhances active observation behaviour and the use of the five senses. Moreover, situational motivation is connected with attractiveness (Bitgood, 1988; Falk & Dierking, 2002), and it was found to be one of the keys for explaining visitor behaviour and learning in an exhibition context. It is the first step in deeper learning, described as holding power (Screven, 1992). One concrete indicator of holding power is how much time students intensively spend in the hands-on demonstration in an interactive exhibition. However, in order to achieve the goal of transforming external regulations into internal engagement and further into self-endorsed engagement, the crucial factor is experience of autonomy, as the SDT-theory posits (Reeve, 2002).

Individual motivation has repeatedly shown its ability to enhance learning both at school and in informal learning settings (Rennie, 2014). However, our knowledge about the relative role of a more stable motivation, situational motivation and interest in learning in different contexts on learning outcomes is still scarce. In this study we compared science learning interest in formal school environments with that at the science centre Mars and Space exhibition. We related these results with the autonomous motivation and with the situational motivation that was aroused in the exhibition. The exhibition experience was found to be beneficial to groups that were previously found to be less interested in science, namely girls and students achieving less well academically at school. While increasing interest in science is a goal as such, the aim was also to analyse whether science knowledge would increase after the exhibition compared with the pre-exhibition situation. A further goal was to see if this information could be recalled five months after experiencing the exhibition. Previous meta-research (Thuneberg & Salmi, 2018) showed that the positive effects of the informal environment on learning gains were not self-evident, but that they varied according to the students’ prior interest in science and their readiness to take responsibility for setting their own goals (Renninger, 2000).

Science exhibitions face the dilemma of whether or not they are capable of orientating and enhancing momentary, strong situational motivation into long-lasting intrinsic motivation (Salmi, 2003; Rennie, 2014). This is also one of the biggest challenges for STEAM-education in open learning environments, such as science centres (Salmi, Thuneberg, & Vainikainen, 2017). The results of this study provided evidence that situational motivation was enhanced by interest in school science in the STEAM science exhibition context. This led to better cognitive learning results in the post-knowledge test.

During the last decade the number of research reports related to informal science learning has been growing rapidly, and beliefs about its effectiveness in enhancing learning and not only motivation is no longer based only on anecdotal evidence (Falk & Dierking, 2002; Rennie, 2014; Thuneberg & Salmi, 2018). However, there are only few results related to long term learning (Salmi & Thuneberg, 2017). Often the results are effective immediately after the visit. Short term effects are rather remembering than learning, while the results of this study are encouraging: The learning results remained and get even stronger after half a year, which is actually a long period in the life-cycle of twelve year old students!

Main implication of the results for further studies and practical pedagogy in schools and informal learning settings is to have pre- and post-materials both for teachers and pupils for better understanding of the context and interactive learning environment. The exhibition and planetarium programme also gave the students an opportunity to see things in a new light: reframing is a way to find new, creative solutions for enhancing learning (Mattila, 2000; Olier et al., 2017). Thus, interest and situational motivation were the first steps, and the superficial situational motivation seemed to successfully change into content-based intrinsic motivation with longer-lasting, deep learning outcomes.

Disclosure statement

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

Additional information

Funding

This work was supported by Seventh Framework Programme. The study was funded by the European HORIZON-2020.

Notes on contributors

Hannu S. Salmi

Hannu S. Salmi has been working since 2007 as Research Director in the University of Helsinki, Faculty of Educational Sciences, and earlier as Professor of Science Communication in Sweden. He has been working as Director of Research and Development in Heureka, the Finnish Science Centre with wide experience of more than 30 EU-funded projects. He has also been the member of the EU-DG Research Advisory Board of the Science Education. In 2017 professor Salmi has been the expert member in The Global Network of Science Academies Committee for Climate Change and Science Education. His main research with several articles are presenting areas of informal science learning, scientific literacy, motivation, and Augmented Reality.

Helena Thuneberg

Helena Thuneberg works as senior researcher and lecturer in The University of Helsinki, Faculty of Educational Sciences. Her main research is focused in motivation both in formal education and informal learning setting with several publications related to museum visitors and pedagogy in several countries.

Franz X. Bogner

Franz X. Bogner First doctoral degree in Zoology [1987: Dr.rer.nat. University of Regensburg]; second degree: Habilitation in Biology Didactics [1996; Dr.rer.nat.habil. University of Munich]; post-doctoral fellowship at the Cornell University [1989–1991, Ithaca, New York]. After a first full professorship at the University of Education in Ludwigsburg/Stuttgart (1997–2004), he was appointed at the University of Bayreuth, again as a full professor and head of the Institute of Biology Education. In team with a colleague of Mathematics, he is the director of the Z-MNU (Centre of Math & Science Education) at the University of Bayreuth. He and his research group are mainly involved in pre-service teacher education and in-service teacher enhancement. Prof. Bogner’s research projects consistently included cognitive (and emotional and attitudinal) assessment. The impact factor of Prof. Bogner is one of highest of German Science Educators (h = 24), his best-cited paper (Bogner 1998) is 288 cited.