Students’ use of observation in geology: towards ‘scientific observation’ in rock classification

ABSTRACT

Students struggle with observing scientifically and connecting observations to scientific theory. This study investigates how students actually use observation in rock classification – a classical practical task in science education. To describe the level of students’ use of observation, data was collected by videotaping 19 small student groups (55 students aged 16–18) in Norway while they were classifying rocks. A modified version of the observation framework proposed by [Eberbach, C., & Crowley, K. (2009). From everyday to scientific observation: How children learn to observe the biologist’s world. Review of Educational Research, 79(1), 39–68] is used to analyse how students’ notice features of rocks (noticing) and interpret the geological processes forming those features (expectations) at different levels: everyday, transitional or scientific. The findings showed that none of the student groups used everyday observation. Three student groups used observation at a transitional level, whereas twelve groups performed observation that can be described as transtional/scientific level. Four student groups used scientific observation. Based on the findings, an observation framework for rock classification is proposed. The challenges encountered by the students are discussed, thus providing ideas for how teachers can support students to use scientific observation in rock classification.

KEYWORDS:

• Geology
• observation
• scientific practices
• practical work
• secondary students
• video observation
• qualitative research
• rock classification

Introduction

This study aims to provide a deeper understanding of how students use observation when classifying rocks into the major rock groups: magmatic, metamorphic or sedimentary. Classification and interpretation of rocks, including knowledge of geological processes, is a classical practical activity in science learning and is included in science and geography curricula in many countries (National Research Council [NRC], 2012; King, 2008; Norwegian Ministry of Education and Research, 2006). Despite this, review articles demonstrate that students of all ages struggle with understanding the nature of rocks, including what a rock is, how rocks can be classified into the major categories, how these categories represent the geological processes that form rocks and the mechanisms driving those processes (American Association for the Advancement of Science, 2015; Cheek, 2010; Dove, 1998; Francek, 2013; King, 2008). Students’ alternative conceptions and misunderstandings can partly be ascribed to the strategies they use when classifying rocks (Ford, 2005; Frøyland, Remmen, & Sørvik, 2016; Hawley, 2002). Typically, students sort rock types by choosing superficial features, such as colour or shape, and by comparing and identifying rocks using photographs. A possible consequence is that students classify rocks by memorising specific specimens rather than using authentic scientific practices, such as observing relevant features and interpreting those features in a geological framework. However, researchers argue that students can learn to observe more scientifically (Eberbach & Crowley, 2009; Ford, 2005; Gelman & Brenneman, 2012; Monteira & Jimenez-Aleixandre, 2015). Eberbach and Crowley (2009) elaborate on this by presenting an observation framework describing three levels of observation in biology: everyday (naïve), transitional and scientific (expert). The present study builds on this by operationalising the framework in the case of rock classification. By analysing videos of 19 small groups of students (N = 55, aged 17–18) in Norway, students’ use of observation when classifying specimens of rocks into major rock categories are described, and their level of observation is identified as everyday, transitional or scientific. Hence, the research question is: ‘What is the level of students’ observation when classifying rocks into major groups?’ This will provide a foundation for discussing possible challenges students may face regarding scientific observation. Before introducing the observation framework in detail, we review research exposing the need for more knowledge about the students’ use of observation in geology.

Research on students’ use of observation when classifying rocks

Classifying specimens of rocks into major rock categories can be considered as a practical activity requiring observational geological practices. Geologists notice features and properties (e.g. patterns, grain size and the relationship between minerals) of specimens, and these observations provide clues about the geological processes describing the conditions in which the rock was formed. Thus, geologists interpret these features and make hypotheses about their formation in a geological framework. Despite being central practices in geology, research demonstrates that rock classification is challenging for students (Dove, 1996; Francek, 2013; Frøyland, Remmen, & Sørvik, 2016; Guffey, Slater, & Slater, 2017; Happs, 1982; Kusnick, 2002; Stofflett, 1993; 1994). Instead of observing features of rocks, students memorise rocks they have seen earlier, leading to a ‘namedropping’ strategy that does not align with scientific practice (Dove, 1998; Frøyland et al., 2016). Ford (2005) found that primary school students focused on intuitive, non-scientific or general features (e.g. colour, shape, weight, size and temperature) in their written descriptions of rocks and argued that students must learn to differentiate between geologically relevant features (e.g. mineral composition and patterns) and irrelevant features (e.g. size and dirt) to make a scientific classification of rocks. However, relevant features of rocks that are ambiguous can also cause challenges. Students noticing grain size were unable to use this feature to assign the specimens to a major rock category (Remmen & Frøyland, 2013; Westerback & Azer, 1991). The students surveyed by Kortz and Murray (2009) classified rocks based on colour. The students believed that black rocks are made of magma and that magmatic rocks are black. Consequently, the students failed to classify granite as a magmatic rock. These examples demonstrate that the incorrect use of observation (e.g. noticing ambiguous features or features that are irrelevant in the situation) can lead to incorrect rock classification (Dove, 1998; Ford, 2005; Frøyland et al., 2016; Guffey et al., 2017; Happs, 1982; Kortz & Murray, 2009). Connecting features to scientific theories of rock formation is another challenge for students, and a number of non-scientific and erroneous ideas about rock formation have been identified (Kusnick, 2002; Stofflett, 1993, 1994). For instance, Stofflett (1993) found that students interpreted the striped pattern in gneiss as originating from glaciation formation, underwater formation and formation in Earth’s core. Students who have a more scientific understanding of the formation of metamorphic rocks often use vague terms like ‘heat’ and ‘pressure’ (Ford, 2003; Kortz & Murray, 2009). Kortz and Murray (2009) argued that students often know that heat and pressure are involved in rock formation, but they do not know that heat and pressure can be caused by plate movements; thus they do not recognise that rocks are connected to scientific frameworks such as plate tectonics. Ford (2003) characterised this as a ‘passive response’ to understanding rock formation. Clearly, the aforementioned studies show that rock classification is a difficult task for many students, but this conclusion is predominantly based on results from paper-and-pencil tests, surveys or individual interviews. Hence, there is a need for more qualitative descriptions of students’ observational performance while assigning rocks into major categories. Frøyland et al. (2016) presents one exception by showing that even young students can classify rocks in a scientific way if they learn to notice geologically relevant features, such as patterns (e.g. dots, stripes, and layer-on-layer formations with fossils) and use those features to assign specimens into major rock categories. This suggests that students can learn to observe geologically relevant features when classifying rocks. However, the aforementioned studies did not consider the details in students’ observations, e.g. the various levels, as observation can be more or less in line with a scientific perspective. It seems that more knowledge about students’ use of observation is needed to help them develop a more scientific understanding of rocks. This requires a theoretical consideration of observation as a scientific practice and a framework for describing how observation can be carried out at different levels.

Observation as a scientific practice

Practical work and scientific practices

Practical activities are a critical aspect of science learning, as they involve ‘interacting with materials or with secondary sources of data to observe and understand the natural world’ (Lunetta, Hofstein, & Clough, 2007). According to Tiberghien’s (2000) model, this requires connecting the two domains: the domain of objects and observables, which are features and properties that can be observed directly, and the domain of ideas, which involves unobservable entities, behaviours and processes. Connecting observable phenomena to scientific ideas and theories is intricately linked in scientific practices, which are further identified as eight practices in the US Framework for Science Education (Furtak & Penuel, 2018; NRC, 2012; Osborne, 2014). In short, the practices involve planning and carrying out investigations to collect data that can be analysed, interpreted and used as evidence in arguments. However, none of these scientific practices include the term ‘observation’, despite the fact that observation is the foundation for all the aforementioned practices. Scientists use their senses, skills and knowledge to observe and produce knowledge that can lead to development of scientific theory (Gelman & Brenneman, 2012; Hodson, 1986; Osborne, 2014). Duschl and Bybee (2014), who also noted the implicit reference to observation in NRC’s practices, explained that scientific observation is embedded in other scientific practices like ‘planning and carrying out investigations’, ‘analysing and interpreting data’ and ‘engaging in arguments from evidence’. Duschl and Bybee (2014) argued that each scientific practice involves a set of ‘sub-practices’ in which observation is one essential sub-practice, and they suggested a 5D model to unpack how observation is inherent in the scientific practice of ‘planning and carrying out investigations’. In brief, the 5D model consists of deciding what and how to observe, developing or selecting procedures and tools to collect data, documenting observations, devising representations for structuring observations, determining if the data can be used as evidence, and deciding if more data or further investigations are needed.

Components of scientific observation

Deciding what and how to observe – the first ‘D’ in the aforementioned 5D model – is discipline and topic-specific. This makes scientific observation different from sensory perceptions and other kinds of observations because scientists know what to look for and how to search for it, which requires coordination of purpose, scientific theory, practice and experience (Daston & Lunbeck, 2011; Eberbach & Crowley, 2009; Hodson, 1986). Reviewing the literature, Eberbach and Crowley (2009) identified four components involved in scientific observation:

• Noticing: Noticing features of the object that are scientifically relevant;

• Expectations: Features are interpreted within a disciplinary framework;

• Observation records: Observations are carried out using the cognitive and physical tools of the discipline to record observations systematically;

• Productive dispositions: Observation involves engagement and practice in different contexts over time.

Eberbach and Crowley (2009) also reviewed research on children’s use of observation, and they found that their observations are usually naïve and simple because they have not gained the same disciplinary tools, theoretical frameworks, and resources to practice observation in different contexts over time as scientists have. Therefore, Eberbach and Crowley (2009) developed descriptions of the differences between children and experts by distinguishing between three levels of observation: everyday, transitional and scientific observation. These three levels are merged with four components (noticing, expectations, observation records and productive dispositions) to create an observation framework adapted to ornithology (Eberbach & Crowley, 2009, pp. 55–56). This framework provides hypothetical descriptions of typical practices in each component for each of the three levels, thus reflecting different qualities of observation in ornithology. For example, the description: ‘Notice more irrelevant than relevant features that distinguish one kind from others without explicit awareness’ reflects a characteristic of everyday observation of birds, whereas an example of a scientific observation characteristic is: ‘Notice and describe relevant features and ignore irrelevant features using disciplinary structure (e.g. taxonomy)’ (Eberbach & Crowley, 2009, p. 55). Eberbach and Crowley (2009) argued that their observation framework can be a tool for thinking about disciplinary observation, and they proposed that it can be adapted to geology. The present study follows this call by applying the observation framework to analyse students’ use of observation when classifying rocks.

Methods

This section describes rock classification in the Norwegian national curriculum, along with the students and methods employed for investigating the level of students’ use of observation.

The Norwegian national curriculum for geoscience specialisation

In Norway, students can choose to study in an optional geoscience specialisation programme in upper secondary school. According to the national curriculum for geoscience specialisation, students should be able to ‘explain the formation of magmatic and metamorphic rocks by using the theory of plate tectonics’ (Norwegian Ministry of Education and Research, 2006). Regarding sedimentary rocks, students are expected to ‘explain how sediments and sedimentary rocks are formed’ (Norwegian Ministry of Education and Research, 2006). These curriculum goals provide an important background for interpreting the students’ performances of rock classification.

The students

Fifty-five students in grade 12 (aged 17–18 years) from three different schools in Norway participated in this study. The students had chosen the geoscience specialisation as part of their upper secondary education, suggesting that they had achieved the aforementioned curriculum goals regarding rock formation. The students were chosen because their teachers had emphasised observation in their teaching of rock classification in accordance with recommendations in the literature (Ford, 2005; Frøyland et al., 2016; Hawley, 2002).

Data collection

The data collection was carried out about 10 months after the students had been exposed to instruction involving practical tasks on rock classification. Specifically small groups of students (between two to four per group) were given a collection of five specimens (two metamorphic, two magmatic and one sedimentary), and asked to classify the specimens and explain why they had sorted them as they did. The specimens used were from the authors’ private collection and thus not identical to the specimens students may have seen before. This approach was also used by Frøyland et al. (2016) and aimed to encourage students to apply scientific practices versus merely memorising a specimen they had seen before (Stofflett, 1994). The student groups’ rock classification performances were video recorded using head-mounted cameras (HD GoPro). Working in small groups ensured that the students’ use of observation were more visible through their verbal and physical interactions (Hodson, 1986). After a few minutes, the student groups were asked by the researchers for their conclusions. If necessary, the students were prompted to clarify how or why they classified the specimens as they did. This provided an opportunity to justify conclusions, which is a technique used in investigations of students’ understanding of scientific concepts (Kortz & Murray, 2009; Mills, Tomas, & Lewthwaite, 2017; Osborne & Gilbert, 1979). The rock classification task carried out by 19 student groups produced 19 videos, each 5–10 min in length. Below, the video analysis is explained, beginning with the analytical framework.

Analytical framework

To describe the level of students’ use of observation when classifying rocks, Eberbach and Crowley’s (2009) observation framework was chosen. It emphasises the importance of observation as a scientific practice and acknowledges that there are various qualities of observation, depending on knowledge and experience. Two of the components, noticing and expectations, were most relevant for analysing the present data. However, the original framework was modified from its initial focus on birds, thereby making it more applicable to rock classification. For instance, the word ‘bird’ was replaced with ‘rock’ and descriptions that were irrelevant to rock classification were omitted. Table 1 shows the adapted observation framework informing the analysis.

Table 1. Observation framework adapted from Eberbach and Crowley (2009, pp. 55–56) to analyse the students’ use of observation when classifying rocks.

	Everyday observation (Novice)	Transitional	Scientific observation (Expert)
Noticing	Notice that rocks are different from other materials Notice more irrelevant than relevant features that distinguish one kind from others without explicit awareness Describe few features that may or may not conform to disciplinary structure Name kinds of rocks, but naming does not do a lot of work	Noticing more relevant features and identify patterns of features Use and describe features validated by others to identify rocks (e.g. field guides) Name more kinds of rocks Noticing stimulates related knowledge Name and organise rocks into groups	Notice and describe relevant features and ignore irrelevant features using disciplinary structure (e.g. classification) Chunk observational information and use smaller search space to notice and group rocks Name more kinds of rocks and at higher hierarchical levels Identification and naming fits rocks into complex systems and is related to complex relationships Stimulate related knowledge Infer processes from features
Expectations	Vague expectations about observations Confuse observational evidence with one’s beliefs	More explicit expectations about rocks that reflect plausible observations Explanations vary between being more scientific and more everyday	Explicit hypothesis consistent with a theoretical framework, shapes observation, search space, and documentation Skilled coordination of hypothesis and evidence

Data analysis

The 19 videos were analysed alongside transcripts of students’ performances of rock classification. By focusing on the use of observation emerging in the interaction between students and the rock specimens, the students’ use of observation was compared with the descriptions in the observation framework in Table 1 – an approach that can be characterised as deductive (Derry et al., 2010). More specifically, the analysis of the noticing component focused on the features the students mentioned when classifying the rocks and whether these features were relevant in a geological framework. According to the observation framework, everyday observation was evident if students prioritised irrelevant features, whereas predominance of relevant features reflected transitional observation. Scientific observation was recognised when students only referred to relevant feature. The analysis of the expectation component involved a consideration of how the students interpreted the features within a geological framework. If students explained features of rocks using erroneous or naïve understandings of geological processes, these were assessed as everyday observations. Students explaining relevant features by referring to parts of relevant geological processes were interpreted as using observation at a transitional level. Scientific observation was recognised if students connected features of rocks and geological processes to scientific frameworks such as plate tectonics. Plate tectonics was chosen because it aligned with the goals in the geoscience curriculum mentioned above.

The researchers (authors) viewed the videos individually and determined the level of observation identified in the students’ performances, resulting in an agreement between the researchers in 17 of 19 group performances of rock classification. Disagreements between the researchers in the analysis and interpretations of the students’ use of observations were solved by reviewing and discussing video data and the analytical framework repeatedly until agreement was reached, thus enhancing the reliability of the findings (Derry et al., 2010).

Findings

All 19 student groups concluded the task by classifying the specimens into the three major rock categories. However, considering how the students reached their conclusions, differences in their use of observation emerged. While none of the student groups used everyday observation, fifteen student groups used observation that indicated a transitional level and four groups used scientific observation. There were great variations among the fifteen student groups using transitional observation. Twelve of the student groups used observation that was more advanced than the other three groups, but less advanced than the level of scientific observation. Therefore, the transitional level of observation was divided in two sub-levels, denoted as transitional (less advanced) and scientific/transitional (less advanced than scientific observation, but more advanced than the transitional level). These findings are further explained below by presenting an illustrative case of each of the three levels: transitional, scientific/transitional and scientific observation. All names presented in the findings are pseudonyms.

Transitional observation in the noticing and expectation components

Three student groups used observation reflecting a transitional level in both the noticing and expectations components. In the noticing component, these students began their rock classification by mixing irrelevant and relevant features, but they ended up using relevant features in their conclusion. When prompted by the researcher, they provided fragments of rock formation theories aligning with a transitional level in the expectation component. Details of the transitional level are illustrated by group members Mattias, Tom and Knut.

Naming and noticing irrelevant and relevant features

When the boys encountered the specimens, they tried to recall names of rocks rather than noticing features.

Tom [picks up a magmatic specimen]:

‘This is not basalt, but what was it called? I can’t remember … ’

When Tom could not remember the specific name, he suggested they classify the specimens according to features such as dark and light colours. The peers did not listen; rather, they began to notice specific features of the specimens like plane sides, weight, shininess and granite’s ‘bad smell’, along with relevant features, such as patterns (dots, stripes and layers), grain size, colour and crystals. At this point, the mixture of irrelevant and relevant features indicated observation on an everyday level. However, the students did not pursue the features any further. Instead, they began focusing on identification of specific specimens and the major category to which each belonged, as illustrated in the following quotation and Figure 1.

Mattias [picks up a magmatic sample]:

‘This is rombeporphyr, right? A magmatic rock.’

Mattias’ remembrance of the specific name – rombeporphyr – and the major group to which it belonged indicated that he applied his prior knowledge and experience. However, the data does not reveal whether he was guessing or whether the classification was based on noticing features, but it may correspond to ‘name and organise objects into groups’ at the transitional level in the observation framework. Regarding the sedimentary and metamorphic specimens, the students were more uncertain and returned to noticing features. For example, when the researcher asked about the metamorphic specimen, the students stated the following.

Mattias:

‘You can see that it [the metamorphic specimen] does not have clear layers.’

Tom:

‘We think it has banding, so it’s gneiss.’

Researcher:

‘So you think it’s metamorphic?’

Tom:

‘It’s hard to say.’

Here, the students identified gneiss by noticing ‘lack of clear layers’ and ‘banding’ as relevant features of metamorphic rocks. Although Tom identified gneiss by noticing features, he was uncertain about which major rock category gneiss belongs to. This indicated a certainty in naming the rock, but an uncertainty in organising the specimen into a major category.

Figure 1. The students recalled the specific names of the specimens and the major categories to which they belonged, and they noticed mostly relevant features of rocks.

Fragments of rock formation theory

The students at the transitional level did not interpret the features on their own. When prompted by the researcher, the students provided a hesitant answer, as exemplified here.

Mattias [referring to metamorphic specimens]:

‘Ehhh, these are formed by high pressure and temperature.’

Knut continued explaining that sedimentary rocks are formed by ‘layer-on-layer’, and magmatic rocks are ‘igneous rocks, formed by lava deep down’ without really referring to the features in the present specimens or including any reference to a theoretical framework explaining layer-on-layer or lava formation. Their interpretations of the specimens consisted of fragments of theories of rock formation, which was analysed as transitional in the expectation component.

To summarise, the students began their classification by noticing both irrelevant and relevant features and naming specimens. In their final conclusion, the students referred to relevant features. When prompted, the students provided fragments of scientific theories that seemed loosely connected to the specimens’ features. Therefore, this student group was characterised as using observation on a transitional level in both noticing and expectation components.

Scientific in the noticing component and transitional in the expectation component

Twelve of the student groups used scientific observation in the noticing component and transitional observation in the expectation component; therefore, this level was labelled ‘scientific/transitional’. These student groups typically classified the specimens into the major rock categories with no or minor errors, they noticed relevant features and they interpreted features by referring to fragments of scientific theories about rock formation. Below, the group of Peter, Georg and Tore provides an illustrative case of the scientific/transitional level.

Noticing relevant features

The group recognised the magmatic specimens instantly without verbalising their noticing, but they spent a long time (about seven minutes) discussing whether a specimen was sedimentary or metamorphic by noticing and examining its features. Figure 2 and its accompanying excerpt illustrate a portion of their discussion.

Tore [holding a metamorphic specimen]:

‘It looks sedimentary at first glance.’

Peter:

‘But it doesn’t have clear stripes.’

Tore [puts two metamorphic samples together]:

‘If you compare these two with stripes … ’

Peter [referring to the first metamorphic sample]:

‘Actually, I think this is metamorphic because it doesn’t have layers like this [pointing at a slate/sedimentary rock]. It should have clear layers if it is a sedimentary rock.’

The excerpt reveals that the students examined relevant features – stripes and layers – to distinguish between the sedimentary and metamorphic specimens. This corresponded to descriptions of scientific observation such as ‘notice and describe relevant features and ignore irrelevant features’, ‘chunk observational information and use smaller search space to group objects’ and recognising the feature as ‘being connected to a complex system’ – the major rock categories. The discussion in the prior excerpt stimulated Peter to recall a previous experience.

Peter:

‘There is a rock nearby that is polished. And that’s metamorphic, and then you can see stripes.’

Tore:

‘But they are not straight and horizontal layers, they bend.’

Peter [pointing at the current metamorphic sample]:

‘These are rough, and they cannot divide in two.’

The excerpt shows a use of observation that can be associated with descriptions at both the transitional level of ‘noticing stimulates related knowledge’ and the scientific level of ‘stimulate related knowledge’. At the end of the discussion, Peter justified their conclusion about the specimens.

Peter [holding a metamorphic specimen]:

‘Here you can see clear layers, and there is a difference between the layers. And this one [holding a metamorphic specimen] – we are not sure, but we believe it’s a metamorphic or sedimentary. It does not have clear layers, which we think is a feature of sedimentary rocks. But the layers are discontinuous, so we believe it’s metamorphic.’

Peter’s answer revealed that they had classified one metamorphic specimen as sedimentary and one metamorphic sample as metamorphic. However, because their conclusion was supported by noticing and discussing relevant features (e.g. whether the layers were clear and straight/horizontal), their noticing complied with a scientific level.

Figure 2. The students discussed whether the specimen (to the left) was striped or had a layer-on-layer pattern.

Connecting features to fragments of rock formation theory

The students verbalised their understanding of how the metamorphic and sedimentary specimens described above were formed, exemplified in Peter’s quote below.

Peter:

‘If I remember right, metamorphic rocks have been one type of rock, and then it has been under high pressure, and then became another rock.’

Peter’s explanation showed an understanding of the cyclic nature of rocks, as rocks can change as a result of processes caused by high pressure. This reflected an advanced understanding of rock formation – more so than just transitional. However, like the students using transitional observation, Peter did not include causes for high pressure in his explanation. For example, there was no connection to plate tectonics. High pressure and temperature was also used by his peer, Georg, when the group was prompted by the researchers to justify their conclusion of magmatic specimens.

Georg:

‘These are magmatic because they are dotted and have been under high pressure and temperature. And that influences the consistency of the rock [pointing at specific features of the specimen].’

The comment reveals that Georg associated dotted patterns with temperature and pressure, but he did not mention the causes of temperature and pressure. It could be that he was thinking of metamorphic rocks, which is often the case in many students’ explanations of metamorphic rocks, but this cannot be answered with our data. Nonetheless, none of the two foregoing examples included a reference to a scientific framework, such as plate tectonics, in their interpretations of the formation of the rock categories. Hence, their performances were more consistent with the transitional level in the expectation component.

To summarise, scientific observation in the noticing component was evident by students’ noticing of relevant features during their classification of the specimens into major rock groups, their use of relevant prior experiences, and their attempts to interpret some of the features by applying associated scientific theories. In the expectation component, the group’s performances were consistent with a transitional level, evidenced by their simple explanation of rock formation that lacked reference to plate tectonics. These findings showed that the students used observation at a scientific/transitional level.

Scientific level in both noticing and expectation components

Four student groups classified rocks via scientific observation in both noticing and expectation components. Specifically, they noticed and described relevant features, which were then used to classify specimens in major rock categories and then at higher hierarchical levels (e.g. three types of magmatic rocks). Regarding expectations, the students included references to plate tectonics in their interpretations of the features of the specimens, sometimes without prompts from the researchers. For instance, the students referred to plate collision when discussing the formation of stripes in the metamorphic specimens, which aligns with ‘making an explicit hypothesis consistent with a theoretical framework’ in the expectation component. Although three groups classified the specimens correctly, there was one pair, Aud and Lise, who made an error in the final conclusion. Accordingly, this pair is used as an illustrative example to support the findings above while discussing how they reached an incorrect conclusion despite the fact that they used observation consistent with a scientific level.

Noticing features and naming them at higher hierarchical levels

Lise and Aud began by sorting the specimens into the three main rock categories by noticing and describing relevant features, such as crystals, dots, stripes and layer-on-layers. After this, the students noticed more distinguishing features between different types within each major rock category (subgroups), which enabled them classify the specimens at a more specific level. For instance, regarding the sedimentary specimen, they used an additional feature to determine whether it was clay slate or alum slate, as depicted in Figure 3.

Figure 3. Aud tried to ‘draw’ with the slate to see if it was clay or alum slate. Since it did not leave a mark on the paper, Aud and Lise concluded that it was clay slate.

The above example shows the students’ ability to classify and name specimens in a manner that aligns with ‘name more kinds of rocks and at higher hierarchical level’ in the observation framework, as they used patterns for an initial classification of the specimens before proceeding with more specific features to distinguish between two types of sedimentary rocks.

However, Aud and Lise spent a long time discussing the features of one of the specimens.

Aud:

‘It is stripy on this side, like layer-on-layer.’

Lise:

‘Perhaps it’s not changed?’

Aud:

‘It could be magmatic, and it gained a unique pattern.’

They ended up classifying it as between magmatic and metamorphic, although a geologist would have classified it as metamorphic. When the researcher asked about the conclusion, Lise provided the below reasoning.

Lise:

‘We decided to sort it between magmatic and metamorphic. We thought that it was a bit changed, but we were unsure.’

Lise’s quote revealed a realisation that rocks are changeable material, which corresponds to ‘identification fits into complex systems and relationships’ in the framework. Although their conclusion was erroneous, the students classified an unfamiliar specimen by making a hypothesis based on its feature, which could reflect ‘infer processes from features’ in the observation framework.

Connecting features to plate tectonics

Without prompts from the researchers, the girls noticed the size of the dots and the dark or light colour and used these features to interpret whether the magmatic specimens were extraneous or intrusive and mafic or felsic. This could match several descriptions in the observation framework, including ‘infer processes from features’, ‘skilled coordination of hypothesis and evidence’ and ‘explicit hypotheses consistent with a theoretical framework’. Regarding the metamorphic specimens, Aud and Lise discussed whether the stripy specimen was formed by regional or contact metamorphosis, which showed that they had detailed knowledge about the formation of metamorphic rocks. However, they did not pursue their hypothesis of type of metamorphism any further. This is not surprising, as it is difficult to infer metamorphic processes from features in a small piece of rock. Because the students did not realise this limitation, it may be that their ability to coordinate hypotheses and evidence regarding metamorphic rocks was not fully developed in this case. When the researcher asked about their classification of the metamorphic specimen, Lise responded that it had been formed by high pressure and temperature during collision between plates, which reflected an ability to connect rock formation to the theoretical framework of plate tectonics.

To summarise, the students using scientific observation in both components were recognised by the following performances: noticing the patterns (dots, stripes and layer-on-layer) as relevant features to classify the specimens into the main rock categories, classifying the specimens at a higher hierarchical level, inferring formation processes from specimen features, and connecting the formation processes to plate tectonics. The students also attempted to coordinate noticing and expectations, although this did not always result in a correct conclusion.

Discussion

The purpose of this study was to investigate the level of observation in students’ classification of specimens of rocks into major groups. This was based on our literature review indicating that observation is a challenging scientific practice requiring disciplinary-specific knowledge of what and how to observe, thus revealing a need for a better system for mapping and assessing students’ use of observation in rock classification. Accordingly, the findings from this study can be concluded in an empirically based observation framework adapted for rock classification, presented in Table 2. The framework in Table 2 is a synthesis of the original framework from Eberbach and Crowley (2009), the literature review on students’ use of observation in rock classification and the findings, resulting in four levels of observation in the noticing and expectation components.

Table 2. Observation framework for rock classification based on Eberbach and Crowley (2009, pp. 55–56) and the present study.

Components of observation	Everyday observation based on literature review	Transitional	Scientific/transitional	Scientific observation
Noticing (observing relevant features of specimens of rocks)	Notice more irrelevant than relevant features of rocks Name specimens at a specific level without justification or relation to major rock categories (‘namedropping’ based on memorisation)	Notice more relevant features and use them to classify specimens into the major rock categories Name specimens at a specific level and classify the specimen into a major group	Notice relevant features and use them to classify specimens into the major rock categories Notice features, stimulating related knowledge and experience with rocks that are used to classify specimens Infer rock formation processes based on features	Notice and describe relevant features and use them to classify specimens into the major rock categories Classification and naming fits specimens on a more specific level than the major rock categories (e.g. three types of magmatic rocks) Infer rock formation processes based on features
Expectations (connecting features to rock formation theories)	Erroneous or naive explanations of rock formation	Provide fragments of rock formation theory (e.g. ‘pressure and temperature’, ‘layer-on-layer’)	Connect features to fragments of rock formation theory (e.g. ‘pressure and temperature’)	Make hypothesis about the rock’s formation based on features Connect features and rock formations to plate tectonic processes

Notably, none of the student groups analysed in this study used observation consistent with an everyday (naïve) level in their classification of rocks, which is an improvement compared to previous findings mentioned in the literature review showing that students ignore critical features of rocks and demonstrate misconceptions about rock formations. The main explanation for the findings presented here may be that the participating students had been exposed to instruction ten months earlier. However, the analysis revealed various qualities in the students’ use of observation. Four of the 19 groups used scientific observation in both noticing and expectation components, whereas the other 15 student groups used observation on a transitional level in the expectation component. Thus, the expectation component, interpreting features of rocks within a geological framework, appeared to be the greatest weakness in the students’ use of observation in the present study. This may not be surprising as linking observations and scientific ideas may be the greatest challenge for students during practical work (Lunetta et al., 2007). Therefore, to help students develop scientific observation in rock classification, the challenges that emerged in the findings will be discussed.

Justifying naming by noticing features

The student groups exemplifying transitional and scientific observation used naming in their classification of the specimens. Naming is a characteristic of everyday, transitional and scientific levels in the original observation framework (Eberbach & Crowley, 2009). However, our examples revealed differences in how the students used naming. The students representing the transitional level named specimens without further justifications. This could be interpreted as a ‘namedropping’ approach based on memorisation of specific specimens (Frøyland et al., 2016). However, in the present example, the student identified the rombeporphyr specimen as a magmatic rock, which is certainly more advanced than at an everyday level. This could possibly be interpreted as at a scientific level, since an experienced geologist would probably not justify their classification of specimens. Experts’ observations become tacit after prolonged experience with phenomena (Eberbach & Crowley, 2009). For comparison, the students using scientific observation were more explicit about their noticing. They verbalised their noticing of patterns of rocks in order to classify them into major rock groups before proceeding to name specimens based on noticing specific features. Hence, it appears to be challenging to interpret the level of the students’ use of observation in naming rocks. One way of dealing with this potential problem can be to ask students to justify their naming by noticing features.

Noticing features that are relevant in the situation

In contrast to studies cited earlier, all student groups in the present study noticed predominantly relevant features of rocks, which reflected that the students had developed knowledge of what and how to observe (Duschl & Bybee, 2014). However, ‘relevant features’ appeared to differ across the three levels of observation identified in the present study. Students at the transitional level noticed relevant features like colour, patterns, banding, shininess, and hardness. Although these are relevant features in geology, they may not have been the most relevant features in the current situationof classifying specimens into major rock categories. For instance, colour is not considered a helpful feature when classifying rock specimens into major categories (Westerback & Azer, 1991), whereas shininess and hardness are features that are more relevant for identification of minerals. Therefore, it can be a challenge for students to notice those features that are relevant in the situation. By contrast, the students using scientific observation in the noticing component referred to dots, stripes and layer-on-layers as relevant features of major rock categories. This is consistent with Frøyland et al. (2016), who argue that noticing patterns helps students to distinguish between the three major rock categories. However, the students representing the scientific observation level went further and proceeded to notice features that are relevant at a higher level in the rock classification system, such as three groups of magmatic rocks. Therefore, scientific observation in rock classification may require students to notice features that are relevant at different hierarchical levels.

Using prior experiences with rock classification

The boys representing the scientific/transitional level drew on prior experiences to make sense of the stripes and layer-on-layer patterns seen in the specimen at hand. This suggests that they had multiple experiences noticing these patterns. This aligns with the fact that developing scientific observation requires extensive training and experience over time (Eberbach & Crowley, 2009). It should be noted that the students in our example were able to productively use their prior experience to perform a close examination of the specimen to decide if it was sedimentary or metamorphic. This contrasts with students investigated in other studies who had been exposed to rock classification in different contexts over time but lacked experience with using observation to perform rock classification tasks (e.g. Remmen & Frøyland, 2013). Thus, scientific observation may not only require persistence in training and engagement in rock classification; it may also require the ability to combine prior experiences with observation of rocks to know what and how to notice.

Recognising specimens in between major rock categories

Students using scientific observation encountered challenges when noticing features and decided to categorise a metamorphic specimen in between magmatic and metamorphic rocks. This indicates that these students realised that some rocks are intermediate between major rock groups, which contrasts with earlier findings showing that students do not understand the changing and cyclic nature of rocks (e.g. Kortz & Murray, 2009). This was slightly more advanced than Peter’s knowledge; he recalled that metamorphic rocks have evolved from another type of rock through geological processes on a general basis. Although the students using scientific observation based their decision on noticing the specimen’s dots and stripes, their conclusion in the particular situation was erroneous. Their performance reflected an understanding of the complex nature of rocks, but determining the relationship between dots and stripes appeared to be a challenge at the level of scientific observation. Since scientific observation also involves classification ‘into complex systems related to complex relationships’ (see observation framework), it may not be sufficient to simply notice features of the major rock categories. Rather, how the features are related to each other through geological processes should also be considered. The rock cycle can serve as a ‘complex system’ to address this challenge, as it can enhance students’ understanding of the nature of rocks as part of the Earth’s systems (Kali, Orion, & Eylon, 2003). Therefore, it can be argued that scientific observation in rock classification involves an understanding of how features gradually change through processes that can be explained by the rock cycle.

Connecting rock classification to plate tectonics

The students using transitional observation in the expectation component explained the formation of metamorphic and magmatic rocks by using phrases like ‘high pressure and temperature’. This is consistent with what Ford (2003) called ‘passive response’ – reflecting a simple understanding of rock formation. Apparently, the ‘pressure and temperature’ response was loosely connected to the features that the students had noticed in classifying the specimens and in line with Kortz and Murray’s (2009) finding, they did not describe what causes the pressure and temperature. For comparison, the students using scientific observation included plate collisions in their explanation of metamorphic rocks, which indicates that these students achieved the competence goal in the Norwegian national curriculum for geoscience: ‘metamorphic rocks should be explained by applying the theory of plate tectonics’ (Norwegian Ministry of Education and Research, 2006). Therefore, achieving a level of scientific observation in the expectation component regarding metamorphic rocks would involve application of plate tectonic processes in the present study context. Although very brief, this finding also provides some interesting thoughts about the scientific observation level in rock classification. Connecting features of major rock categories to rock formation theories and plate tectonics may be necessary for developing students’ use of observation in the expectation component beyond passive responses, such as ‘high pressure’ and ‘temperature’. However, the two concepts; rock classification and plate tectonics; tend to be investigated more separately (e.g. Ford, 2005; Mills et al., 2017). Therefore, it can be proposed that more research is needed to explore how students connect rock classification to plate tectonics and vice versa.

Limitations of interpreting features in a small piece of rock

The findings indicate that the students had more or less scientific knowledge of rock formation, but the ability to use it to infer processes from features varied across student groups. Although observation also involves determining if the data can be used as evidence (Duschl & Bybee, 2014), the students using scientific observation did not always understand the limitations of using features as evidences for some geological processes. For example, Aud and Lise’s observation of metamorphic rocks showed that they had knowledge about contact and regional metamorphosis, but they really weren’t able to find features that could be used as evidence in their interpretation of the metamorphic specimens. On one hand, it showed that they knew more about metamorphic rocks than what was evident in the small pieces of rocks at hand. On the other hand, it showed that the students did not really understand that features of specimens cannot always be used as evidence for large, scale geological processes. In the case of metamorphic rocks, one needs to observe the rock in its natural surroundings and know the geological history to make hypotheses about types of metamorphic processes. This finding may represent a misconception of metamorphic rocks in rock classification not reported by the aforementioned studies. Nonetheless, knowing what theories are possible and relevant in the current situation may be a challenge that needs to be overcome in order to achieve scientific observation when classifying hand-size rock samples.

Conclusion and implications

To investigate the research question, ‘what is the level of students’ observation when classifying rocks into major groups?’, a modified version of Eberbach and Crowley (2009) theory-driven observation framework was applied. The findings illustrated how small groups of students, who had been exposed to instruction, used observation in line with a transitional or scientific level. Hence, none of the student groups used observation in line with an everyday level, which challenges previous research showing that students struggle to observe geologically relevant features when classifying rocks. However, the analysis revealed the need for nuancing the observation framework. While most student groups used noticing on a scientific level, their use of expectations aligned with a transitional level. Hence, an additional level of observation in the case of rock classification was identified as scientific/transitional (see Table 2). The summary of our findings in Table 2 proposes an adapted and expanded framework that teachers can use as a starting point when assessing the level of students’ observations during practical work with rock classification. Furthermore, the discussion identifies some possible challenges that teachers can be aware of in the process of supporting students to use observation in rock classification; justifying naming by noticing, noticing features that are relevant in the situation, using prior experiences with rock classification, recognising specimens in between major rock categories, and connecting rock classification to plate tectonics. The observation framework for rock classification in Table 2 together with the discussion provide a basis for designing innovative curriculums that support students’ development of scientific observation in geology.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Kari Beate Remmen http://orcid.org/0000-0002-6286-9313