Learning difficulties experienced by grade 12 South African students in the chemical representation of phenomena

Abstract

This study investigated the learning difficulties of grade 12 South African students in a national chemistry examination. A quantitative analysis of students’ performance in the examination revealed there was a significant difference between student performance in questions where students needed to execute a transformation across levels of chemical representation, than in questions that demanded representation at a discrete level, with students finding it more difficult to answer questions demanding a transformation. The qualitative probe in the second phase of the study elaborated upon this finding. When students were interviewed on their incorrect responses, there was substantial evidence to suggest that they did have an understanding of the concept at the discrete levels of representation in a transformation, but found it challenging to make connections across the levels. The findings of this study have implications for classroom practice, and do raise concerns on the current status of chemistry teaching as well as the preparedness of teachers in supporting students to achieve conceptual understanding in chemistry. What was also significant was that when students were prompted to explore connections with another level, they were able to extend their understanding to that level. A scaffolding approach in which a teacher supports students by making them focus on the connections between the levels should be considered. In this regard, teachers need to engage students on the relationship between the levels of chemical representation by being explicit about the transitions being made, and prompt students in making transitions across levels.

---

Background and introduction

The complex and abstract nature of chemistry makes the study of the subject difficult for students (Gabel, 1999; Treagust and Chittleborough, 2001). One of the reasons for the difficulties that students experience is related to the multiple levels of representation that are used in chemistry teaching to describe and explain chemical phenomena (Gabel, 1998; Nakhleh and Krajcik, 1994). These chemical representations constitute the elements of chemistry language (Hoffmann and Laszlo, 1991). Johnstone (1982, 1993) distinguished three levels of chemical representation of matter: the macroscopic level, the sub-microscopic level, and the symbolic level. Chandrasegaran et al. (2007) describe these three levels of representation relevant to the understanding of chemistry concepts as follows: (1) macroscopic representations describe properties of tangible and visible phenomena in the everyday experiences of students when observing changes in the properties of matter (e.g. colour changes, pH of aqueous solutions, and the formation of gases and precipitates in chemical reactions), (2) submicroscopic (or molecular) representations provide explanations at the particulate level in terms of atoms, molecules and ions, and (3) symbolic representations involve the use of chemical symbols, formulas and equations, as well as molecular structure drawings, diagrams, models and computer animations to symbolize matter.

In order for students to acquire conceptual understanding in chemistry, the teaching of chemistry should promote the development of “representation competence” in students (Kozma and Russell, 1997, p. 949). This competence describes the ability to represent a concept in terms of its surface features (e.g. a colour in a chemical reaction), explaining the concept at the particulate level and representing the concept symbolically. Studies have shown that students are lacking in competence in representing concepts at the sub-microscopic and symbolic levels, since these representations are invisible and abstract (Kozma and Russell, 1997; Wu et al., 2001).

Research also shows that one of the essential characteristics of chemistry is the constant interplay between the macroscopic, sub-microscopic and symbolic levels of thought, and it is this aspect of chemistry learning that presents a particularly daunting challenge to students (Bradley and Brand, 1985; Onwu and Randall, 2006; Treagust et al., 2003). In order for students to develop a deep understanding of chemical concepts they not only need to be able to translate concepts using all three levels of chemical representation, but also be able to transit across levels (Kozma, 2003). Sirhan (2007), in an overview of learning difficulties in chemistry, states that when students are faced with learning situations that demand transition across levels, their working memory space becomes overloaded. Johnstone (1991) makes the point that in the light of working memory’s limited capacity, the novice student in chemistry is unable to operate at all three levels simultaneously.

The manner in which chemistry is taught also contributes to the learning difficulty students experience in the subject. Johnstone (1991) asserts that most chemistry instruction in high school and college chemistry courses takes place at the symbolic level and students do not understand the relationship between the symbolic and the other two levels. Students struggle to interpret a chemical reaction to the microscopic level, and instead memorize what is being presented at the symbolic level in terms of chemical equations and mathematical relationships (Gabel, 1999).

Even in cases where the symbolic level is used by the teacher as a mediator between macroscopic and sub-microscopic levels, children experience difficulty. Although symbols are useful in enabling teachers to make the shifts between laboratory observations at a macroscopic level, and explanations in the form of sub-microscopic models, students may become overwhelmed by the increased cognitive load in such a presentation (Taber, 2009). Taber explains that although the ambiguity of symbols is a powerful tool in allowing teachers to make these shifts, this ambiguity may be problematic for students who have limited background knowledge. In order for students to follow the shifts, they need to be able to interpret when a symbol is being used by the teacher to represent a phenomenon, and when it is used to model the sub-microscopic world. For example, according to Taber “H₂ could be interpreted to mean a molecule or the substance” and students who fail to make the correct interpretation may not be able to readily follow these shifts in thinking (p. 100). He therefore suggests that such shifts need to be more explicitly signalled to students.

The traditional practice within the South African education system, at the high school level, is that instruction is predominantly at the symbolic level of representation. At grade 8 (the first year of high school) students are introduced, in the subject Natural Sciences, to symbolic representations, in the form of chemical symbols, formulae and equations. Due to their limited exposure to the sub-microscopic and macroscopic representations, students often resort to the rote learning of formulae and equations (Steenberg, 2006) without attention being given to the conceptual understanding of chemistry. This practice persists through grade 10, where Physical Sciences becomes an elective, right up to grade 12, the final year of schooling.

We expected that a new South African Physical Sciences curriculum, which subscribes to a constructivist approach to learning, would encourage teachers to explicitly engage students at the three levels of chemical representation in order to facilitate the understanding of chemistry concepts. The national curriculum document draws attention to “the difficulty for beginners in managing, all at the same time, the macroscopic, the micro-level, and the symbolic-level descriptions of chemical phenomena” and states that “continuing attention to this throughout the Further Education and Training phase is a core requirement” (Department of Education, 2005, p. 13). However, the performance of students in chemistry remains a cause for concern with more than two thirds of students attaining a mark of less than 40% in a national examination. In trying to understand the difficulties in chemical representation that students experienced in this examination, we conducted a quantitative analysis of student performance, and thereafter did a qualitative study on the trends revealed by the quantitative results.

The following research questions were formulated:

1. With regards to chemical representations, what patterns are revealed in a grade 12 national chemistry examination?

2. What are the difficulties students experience in the chemical representation of phenomena as revealed in their responses to chemistry examination questions?

In order to assess the conceptual understanding of students in terms of chemical representations, we applied a Classification Framework for Chemical Representations (CFCR) based on a review of literature. We firstly elaborate on CFCR.

Classification Framework for Chemical Representations (CFCR)

Classification can be defined as the basic cognitive task of arranging concepts into classes or categories (Stains and Talanquer, 2007). Classification plays a central role in chemistry, where it is used not only as a way to organize knowledge but also as a powerful predictive tool (Schummer, 1998). Scientists rely on classifications generally to conclude the results of their research. The diverse classification systems used in chemistry to make predictions and build explanations are based on the identification of features at different levels of representation, namely macroscopic, microscopic, and symbolic (Gabel, 1999; Johnstone, 1993). A review of the literature has pointed out that in order for students to have a deep understanding of chemical concepts they not only need to be able to translate concepts using all three levels of chemical representation, but also be able to transit across levels. Therefore in addition to the three discrete levels of representation discussed, the following four bidirectional transformations are identified: macroscopic ↔ sub-microscopic; macroscopic ↔ symbolic; sub-microscopic ↔ symbolic; macroscopic ↔ sub-microscopic ↔ symbolic (where ↔ describes a bidirectional transformation). The framework that guided our analysis of conceptual understanding in chemistry was therefore comprised of 3 discrete levels of chemical representations and four transformation categories that are discussed below.

Macroscopic ↔ sub-microscopic transformation

In this case, transformation takes place from the macroscopic to the sub-microscopic and vice versa. For example, students may be asked to burn a piece of magnesium ribbon in oxygen gas and explain the chemical changes that take place in this reaction at the particulate level. Students will observe that the magnesium ribbon burns with a bright flame forming a greyish ash. A macroscopic representation of this observation would be a combustion reaction of the metal that results in the formation of a residue. This transformation may be explained at the sub-microscopic level as follows: The magnesium atoms change into magnesium ions by donating two electrons to the oxygen atoms. The oxygen atoms change into oxide ions. The magnesium and oxide ions combine to form magnesium oxide, which appears as greyish ash.

Macroscopic ↔ symbolic level transformation

Here, transformation is required from the macroscopic to the symbolic level, and vice versa. Students need to be able to translate what they observe during a chemical reaction into symbolic language. This can also be explained using the above example. Students may be asked to burn a piece of magnesium ribbon in oxygen gas and explain this reaction using a chemical equation by specifying the phases of the reactants and products. At the macroscopic level, students will observe that the magnesium ribbon burns with a bright flame, forming greyish ash. This reaction at the macroscopic level can be represented symbolically using the equation for the reaction:

2Mg(s) + O₂(g) → 2MgO(s)

Sub-microscopic ↔ symbolic level transformation

This transformation is from the sub-microscopic level to the symbolic level, and vice versa. An example of where such a transformation is needed is in the explanation of the dissolution process of sodium chloride in water. At the sub-microscopic level, the student should be able to state that when sodium chloride dissolves in water, it ionizes to sodium ions and chloride ions. This process can be represented in symbolic form as:

NaCl(aq) → Na⁺(aq) + Cl⁻(aq).

Macroscopic ↔ sub-microscopic ↔ symbolic level transformation

This transformation is from macroscopic to sub-microscopic to symbolic and vice versa. In this transformation, all three levels are involved. An example of a case is when students are asked to explain fully what happens when a piece of magnesium ribbon burns in oxygen. To answer this question the student should describe his/her observations, know the particulate nature of matter, and then be able to transform the above information using symbols and equations. In this case, at the macroscopic level students should be able to observe that magnesium burns in oxygen gas with a very bright flame and forms greyish ash. At the sub-microscopic level, students should be able to explain that during the burning, the magnesium atoms change to magnesium ions (positively charged) and oxygen molecules change into negatively charged oxide ions. This understanding at the sub-microscopic level would then enable students to represent the reaction symbolically as follows:

Mg(s) + O₂(g) → MgO(s)

Methodology

This study, in addressing the research questions employs a pragmatic, mixed methodology design (Johnson and Onwuegbuzie, 2004) which uses a combination of quantitative and qualitative methods and procedures. This resulted in both quantitative and qualitative data being collected.

The questions in a grade 12 national chemistry examination (Department of Education, 2009) were analysed and classified according to the CFCR. In this way the chemical representation invoked by each question became evident. Inter-rater reliability in the classification of the questions was established due to an 85% agreement in the classification with a researcher in science education. We analysed a random sample of 800 grade 12 chemistry examination scripts. These were scripts that had already been marked by external examiners appointed by the Department of Education. Quantitative data in the form of the student scores for each of the questions were analyzed statistically using the PASW version 18.0 for Windows software (SPSS). Using this software, the average performance of students in each of the 7 categories of chemical representation in CFCR was determined. A one-way analysis of variance (ANOVA) was carried out in order to test the significant difference between the means for the categories of questions.

The second phase of this study involved a qualitative analysis of student conceptual understanding in terms of the CFCR categories. Due to this being an exit examination, we did not have access to the cohort of students whose scripts were analysed in the quantitative phase. We therefore compiled a test of all the questions related to categories where student performance was particularly poor, and then administered it to a group of 30 grade 12 students at a school that was conveniently located for us. This was a later cohort of students than the one which had written the national examination. Based on a quantitative analysis of the student performance in this test, we identified 15 of these students who had performed poorly in the same questions that were located in the categories where the lowest scores were achieved in the chemistry examination. We did a qualitative analysis of test responses, trying to establish a pattern of common errors made. Thereafter, the 15 students were interviewed on their responses. We hoped here to gain an insight into the difficulty experienced by the students in answering such questions in terms of chemical representation. For example, with regards to a question in the category “sub-microscopic ↔ symbolic level transformation”, we studied the student response to this question and then asked the student to elaborate upon his/her response. We were able in this way to establish the difficulty the student encountered in this transformation.

Results

Weighting of the chemistry examination according to the CFCR categories

Table 1 shows the weighting of the categories of representation in CFCR according to the marks that were allocated to the 72 categorized questions from the chemistry examination paper that formed the focus of this study. The questions at each level were a mixture of multiple choice, short answer, explanation type and problem solving questions.

Table 1 Weighting of the chemistry examination in terms of the categories of chemical representation

Levels	Number of questions	Weighting percentage (%)
Macroscopic level	18	21.6
Sub-microscopic level	8	11.1
Symbolic level	13	18.2
Macroscopic ↔ sub-microscopic level	11	13.5
Macroscopic ↔ symbolic	8	11.3
Sub-microscopic ↔ symbolic	9	12.1
Macroscopic ↔ sub-microscopic ↔ symbolic	5	12.2
Total	72	100.0

---

It is evident from the above data that there was almost an equitable distribution of marks between the discrete categories and transformation categories of chemical representation, with the former allocated 50.9% of marks and rest of the marks allocated to the latter.

Student performance according to the CFCR categories of chemical representation

An overall average of 31.06% for all 72 questions in the examination was achieved by the 800 students with a standard deviation of 17.24%.

Table 2 shows the average percentage marks obtained by students for each category of chemical representation.

Table 2 Descriptive statistics for percentage achieved per category of chemical representation

Levels	Mean (%)	Std. Deviation (%)
Macroscopic level	45.56	21.62
Sub-microscopic level	40.67	18.73
Symbolic level	53.25	19.25
Macroscopic ↔ sub-microscopic level	23.38	18.45
Macroscopic ↔ symbolic	25.36	15.61
Sub-microscopic ↔ symbolic	17.36	10.10
Macroscopic ↔ sub-microscopic ↔ symbolic	16.12	12.36

---

The descriptive statistic shows that there are observable differences in the performances across the categories of representation. The highest performances by students were in the discrete categories (macroscopic, sub-microscopic and symbolic), while students performed poorly in the transformation categories (macroscopic ↔ sub-microscopic; macroscopic ↔ symbolic; sub-microscopic ↔ symbolic and macroscopic ↔ sub-microscopic ↔ symbolic). A one-way analysis of variance (ANOVA) was used to compare the performances of students across the 7 categories of chemical representation. In order to perform an ANOVA, which is a parametric test requiring strict assumptions to be met, the test for normality for all 7 groups as well as for equality of variances (or spread) was carried out. Since all group sizes (categories) were below 50, the appropriate test for normality was the Shapiro–Wilk test. The test showed exceedence probabilities (Sig.) all greater than 0.05, and this is evidence that the groups (categories) could all be considered to have come from normal distributions. To test the second assumption on the equality of variances, the Levene’s test for Homogeneity of Variances was done. A p-value of 0.347 was obtained and therefore this assumption too was met.

A one-way ANOVA showed that there was statistically significant difference at the p < 0.05 level in mean percentages between the categories of chemical representation: F (6,65) = 2.84, p = 0.04. Post-hoc comparisons using the Tukey HSD test indicated significant differences in the mean percentages for the following pair of categories:

• Macroscopic (M = 45.56; SD = 21.62) and sub-microscopic ↔ symbolic (M = 17.36; SD = 10.10)

• Macroscopic (M = 45.56; SD = 21.62) and macroscopic ↔ sub-microscopic ↔ symbolic (M = 16.12; 12.36)

• Symbolic (M = 53.25; SD = 19.25) and sub-microscopic ↔ symbolic (M = 17.36; SD = 10.10)

• Symbolic (M = 53.25; SD = 19.25) and macroscopic ↔ sub-microscopic (M = 23.38; SD = 18.45)

The above result did suggest to us a possible difference in performance between the discrete categories and the bidirectional transformational categories. We therefore decided to merge the categories in creating two larger categories, namely a combined discrete category comprising the macroscopic, sub-microscopic and symbolic categories and a combined transformation category consisting of the transformation categories. Table 3 shows the mean percentages for the combined categories.

Table 3 Mean percentages for the combined discrete and transformation categories

Combined discrete category		Combined transformation category
Mean (%)	SD	Mean (%)	SD
46.49	19.62	20.56	14.21

---

A one-way ANOVA showed that there was a statistically significant difference in mean percentages between these two groups: F (1,65) = 3.34, p = 0.03, with the students performing more poorly in the transformation category questions than in the discrete category questions.

Due to the significantly poor performance of students in questions situated in the transformation categories, we decided to direct our attention to such questions in the subsequent qualitative phase of this study. Here we sought to gain some insight into the difficulty students encountered in engaging in these transformations. In this regard, a test comprising of transformation questions was administered to a later cohort of 30 students. We now present findings on the qualitative analysis of responses by these students to questions in each of the 4 transformation categories.

Qualitative analysis of student responses in transformation categories

For the purposes of this article we report only the questions in a particular transformation category where students encountered the most difficulty.

Macroscopic ↔ sub-microscopic ↔ symbolic transformation. In this category students needed to translate the problem using all three forms of chemical representation. We discuss test responses to questions 9.6 and 10.2.1.

Question 9.6 referred to the following reversible reaction that has reached equilibrium at 470 °C in a closed container.

N₂(g) + 3H₂(g) ⇌ 2NH₃(g) ΔH < 0

Students were required to state and explain how the value of the equilibrium constant K_c would be affected if the temperature was increased to 800 °C.

In terms of the levels of representation, students needed at the macroscopic level to understand that when the temperature was increased this would result in a lower concentration of ammonia molecules and an increase in the concentration of hydrogen and nitrogen molecules (an effect at the microscopic level). Thereafter, they needed to engage at the symbolic level in explaining the effect on the K_c. This entailed writing the Law of Chemical Equilibrium equation, (at the symbolic level) and then deducing that K_c would decrease when the denominator increases and the numerator increases.

An analysis of student responses showed difficulty in shifting across the three levels. Firstly, students were unable to relate the macroscopic change (increase in temperature) to the microscopic change (increase and decrease in concentration of molecules). The following test responses demonstrate that students did not correctly explain the effect of temperature increase on the concentration of molecules.

With the higher temperature, we can have much more of the ammonia gas to be formed faster and the K_c must now become bigger.

Everything will take place faster and the reaction will be quicker with the higher temperature.

In cases where students did correctly connect the macroscopic and sub-microscopic levels, they had difficulty extending this understanding to the symbolic levels in terms of the Law of Chemical Equilibrium equation. The following response reflects this gap between the sub-microscopic and symbolic levels:

The number of ammonia molecules is going to be used up more quickly and we are going to get more and more of the N₂ and H₂ particles to be formed. K_c must become bigger because of temperature increase.

From the interviews, the disconnection between the levels became more apparent. We firstly established that students understood the meaning of ΔH < 0, and thereafter focused on their ability to shift across the levels in answering the question. The following excerpt from the interview shows how a student having successfully moved from the macroscopic to the sub-microscopic level was unable to explain the effect on K_c at the symbolic level.

Researcher: Okay so tell me what the change on equilibrium will be when the temperature is increased?

Student: There must be a shift in the equilibrium reaction because of this.

Researcher: Explain further.

Student: I can say that more of the ammonia is going to be decreased and the nitrogen and hydrogen must become more.

Researcher: Why do you say this?

Student: Because heat is like a product with the ammonia molecules. Once you make the temperature bigger the heat must break up the ammonia to form more of the nitrogen and hydrogen.

Researcher: So now how is the K_c affected?

Student: It must become bigger because K_c is always higher with increase temperature.

Researcher: But how do you know this?

Student: It must be like this.

It was interesting to note that when this student was then prompted to write the Law of Chemical Equilibrium equation for the reaction, and asked to mathematically consider the effect of the change in the concentrations on the fraction, he was able to correctly state that the K_c would decrease.

Question 10.2.1 was also classified as demanding a macroscopic ↔ sub-microscopic ↔ symbolic transformation. This question was formulated as follows: “Magnesium is used to protect underground pipes against rusting. The diagram below shows an iron pipe connected to a magnesium bar. Use the Table of Standard Reduction Potentials to explain why magnesium can be used to protect an iron pipe against rusting”.

In answering this question students need to understand the concept of rusting at the macroscopic, sub-microscopic and symbolic levels of representation, and also be able to transform the concept from one level to another. Rusting is a common phenomenon that students encounter at the macroscopic level and are aware of the need to protect certain metals that are more susceptible to rusting than others. The question demanded that students understand at the sub-microscopic level that a substance undergoes rusting when it loses electrons. In order to explain that magnesium can be used for protection, students needed to establish that it loses electrons more readily than iron. This required an engagement at the symbolic level to be able to compare the standard electrode potentials (E⁰) of these metals. An analysis of responses to this question revealed that students were unable to shift across the three levels of representation. In particular, they had difficulty linking the sub-microscopic and symbolic levels. Students appeared to have a grasp of rusting as being a process whereby a substance loses electrons. Despite the question explicitly stating that students should use the Table of Standard Reduction Potentials in their explanation, students could not apply their knowledge of E⁰ values at the symbolic level to answer the question. The following test responses testify to this:

The one is going to be giving out the electrons and must be rusting. Something about the electrons going away and it becoming rusted.

Rusting shows it can lose some electrons. Iron will lose and magnesium will also lose.

In the interviews we sought further clarification on this difficulty. When students were questioned, it became evident that an impasse was reached in transiting between these levels. The following excerpts from the interview indicate this:

I know that the magnesium must do the rusting better and so the magnesium is saved. It is all about electron being lost. I could not understand this thing about the table.

I can see that the magnesium must lose it (electron) better than the iron. But I am not sure why this must happen.

When these students were questioned specifically on the meaning of the standard electrode potentials they appeared to have a correct understanding of what this meant. The following responses were obtained with regard to this:

It means how easily a substance can get an electron.

It is about getting reduced. What I mean is for it to get accept electrons.

The above example again illustrates that students who did have an understanding at the discrete levels of chemical representation were unable to answer the question due their inability in making connections between the levels.

Submicroscopic ↔ symbolic transformation. In this category students needed to engage with the question at both the sub-microscopic and symbolic levels of representation.

Question 5.6 that was classified as this type of transformation referred to the following flow diagram:

Students were required to write the formula of the reactant in reaction F that would react with ethene to produce bromoethane. The question demanded that student firstly have a representation of ethene and bromoethane at the sub-microscopic level. This entailed knowing the structural formulae of these molecules, and then identifying a molecule that would react with ethene to form bromoethane. Once this molecule was identified, they needed to represent the molecule symbolically by writing its molecular formula.

The responses showed that students had a poor conception of the reaction at both the sub-microscopic and symbolic levels. A common answer given was for the other reactant was “Br₂”, instead of “HBr”. At the interviews students who answered similarly in the test were probed on this response. It was clear that they did not have a correct representation of the reaction at the sub-microscopic level, and this prevented them from making the correct connection at the symbolic level. We probed students on their understanding at the sub-microscopic level by asking them to draw diagrams to show the mechanism of the reaction between ethene and the other reactant. It was evident that students held a limited structural conception of the ethene and bromoethane molecules in terms of the atoms present in each molecule, and the bonding between these molecules.

A student who displayed such an inadequacy at the structural level was asked to describe the mechanism of the reaction. The following excerpt shows how through probing questioning the correct response was elicited.

Researcher: How does the bromine react with ethene? Explain this in terms of atoms.

Student: The one bromine atom goes onto the carbon

Researcher: Show me how this happens

The student draws the following diagram and points out how the bromine bonds with the carbon.
Researcher: But how many bonds can we expect to find on carbon?

Students: Four

Researcher: How many do we find in your diagram on carbon 2?

Student: Five

Researcher: So how can we get four?

Student: The double bond must break to become single

Researcher: Good. So what will we now get? Can you draw it.

The student draws the following diagram..
Researcher: But now the other carbon (carbon 1) has only three bonds.

Student: There must be something else. I think another hydrogen must go onto it.

Researcher: Where is that going to come from?

Student: It must be a hydrogen with the bromine reacting with ethene.

Researcher: So write down the formula of the reactant with ethene

The student writes HBr.

Macroscopic ↔ symbolic transformation. Question 11 was formulated as follows: “Aluminium is one of the most abundant metals on earth, yet it is expensive largely because of the amount of electricity needed to extract it. Aluminium ore is called bauxite. The bauxite is purified to yield a white powder, aluminium oxide, from which aluminium can be extracted. The diagram below shows an electrolyte cell used for the extraction of aluminium at temperatures as high as 1000 °C”.

In question 11.4 (a sub-question) students were asked to “Use a half-reaction to explain why carbon dioxide gas is formed at one of the electrodes”. This required a transformation from the macroscopic to the symbolic level. At the macroscopic level students needed to understand that due to the high temperature, carbon will burn with oxygen to form carbon dioxide. Based on this macroscopic representation, students were required to represent this reaction symbolically as follows:

C(s) + 2O²⁻(g) → CO₂(g) + 4e⁻

The analysis of the test responses showed that students appear to have a sound understanding of the reaction at the macroscopic level, but encountered difficulty in translating this to the symbolic level. A common response was that students wrote the following molecular equation:

C(s) + O₂(g) → CO₂(g)

In the interviews, students who answered in this way in the test were asked to explain the reaction. The responses showed that students did have an understanding of the half-reaction taking place. The following interview excerpt shows evidence of this:

Researcher: Tell me about this reaction.

Student: When the oxygen reacts we know it is going to be reduced and the carbon must be oxidised.

Researcher: How does this happen?

Student: I know the oxide must be oxidising the carbon and the carbon becomes oxidised.

It was evident that students correctly identified carbon as undergoing oxidation but were unable to extend this understanding to the symbolic level by writing the oxidation half-reaction.

Macroscopic ↔ submicroscopic transformation. Question 11.5, also a sub-question in question 11, referred to this type of transformation. Students were required to explain: “Why should the carbon electrodes be replaced regularly?” At the macroscopic level students needed to firstly establish that the carbon electrode would become thinner. Thereafter, they needed to connect with the sub-microscopic level by explaining that the carbon atoms were leaving the electrode as they were being oxidised due to a loss of electrons. A common response given in the test was that the electrode was “being used up”. The following exchange with a student who answered similarly at the macroscopic level shows how through prompting he was able make the transition to the sub-microscopic level and formulate an explanation in terms of the carbon atoms and electrons.

Researcher: Describe what happens to the carbon electrode.

Student: I think it will get lighter.

Researcher: Why?

Student: The carbon must be getting used up.

Researcher: What do you mean by this?

Student: They are going away.

Researcher: What?

Student: The carbon atoms will be going into the liquid.

Researcher: What causes this to happen?

Student: They must be losing electrons so the aluminium can form on the other side.

It becomes clearly evident from this exchange and others that have been discussed that students did have a conception of the chemistry phenomenon referred to in the question, but this was limited to a particular level of representation. When students were prompted by the interviewer to make connections with other levels of representation, they were able to do this, leading to them successfully answering the question.

Discussion of findings

This study investigated the conceptual understanding of chemical representation of grade 12 South African students in a national chemistry examination. We were guided in this analysis by the Classification Framework for Chemical Representation (CFCR) that was derived from a review of literature. This framework enabled us to categorise questions in the examination in terms of chemical representation, and thereafter served as a framework to guide our analysis of students’ test responses in the subsequent qualitative phase of the study. A quantitative analysis of the examination marks revealed a poor performance in all 7 categories of chemical representation. However, there was a significant difference between student performance in questions where students needed to execute a transformation across levels of chemical representation, and in questions that demanded representation at a discrete level, with students encountering more difficulty in answering transformation questions. This finding correlates well with those of other studies that point to students having trouble integrating representations of reactions and developing coherent understanding (Bradley and Brand, 1985; Krajcik, 1991). The qualitative probe in the second phase of the study did elaborate upon this. When students were questioned on their incorrect responses, there was substantial evidence to suggest that they did have an understanding of the concept at discrete level of representation, but found it challenging to make connections across the levels. This finding is in agreement with a study conducted by Bunce et al. (1991) who interviewed students while they solved chemistry problems out loud. Their study indicated that students rarely thought about the phenomenon holistically in terms of the multiple meanings but merely searched for a singular representation that fitted the problem. We maintain that the qualitative and quantitative research methods used did complement each other in generating more complete knowledge on the performance of students in a chemistry examination, and on how these results can be explained in terms of their representational competency. The findings highlight the need for students to develop representational competency in chemistry learning.

The findings of this study have implications for classroom practice. Previous research has shown that many high school teachers do not highlight the inter-connectedness of the levels of representation in their teaching, and as result students do not see the linkages between these levels (Gabel, 1999; Sirhan, 2007). The heavy emphasis on the symbolic level of representation, has contributed to students having a skewed perspective of the subject. The present study provides evidence on the need for teachers to engage students on the relationship between the levels by being explicit about the perspective being used, and the transitions being made. This notion of explicitness in enabling shifts across the levels was alluded to earlier (Taber, 2009). What was also significant in our study was that when students were prompted to explore connections with another level, they were able to extend their understanding to that level. This finding can be explained by students now being forced to make their thinking visible (Chiu and Linn, 2012). According to Rozenblit and Keil (2002), prompts can spur students to question their comprehension, and realize inconsistencies and gaps in their ideas. By asking students to explain their ideas we evoked metacognition in students by creating an awareness in students of their own cognitive processes (Flavell, 1981). This finding does correlate well with other studies on the effectiveness of prompting in classroom situations (Davis, 2003; Davis and Linn, 2000).

The findings also raise the question: Are science teachers sufficiently prepared to support students in making the connections between the levels of representation in chemistry? Although this study did not focus on the teaching of chemistry, it can be inferred by the inability of students to make transitions across the levels of representations that teachers pay scant attention to this in their teaching. Studies have shown that teachers have difficulty in interrelating macroscopic, sub-microscopic and symbolic conceptions in a proper way (Del Pozo, 2001; Lee, 1999; Valanides, 2000). De Jong and Taber (2007) identified the preparation of chemistry teachers as a key issue in the teaching of multiple meanings in chemistry, and also remark that studies on the professional development in-service of teachers on this issue are scarce. Studies that have been conducted with prospective chemistry teachers on this aspect have yielded results indicating an improvement in the conceptual understanding of these teachers and a greater awareness of conceptual difficulties that may arise in the classroom (De Jong and Van Driel, 2004; Kokkotas et al., 1998). It is therefore recommended that more extensive research be undertaken on the implementation of in-service development programmes where teachers acquire strategies on how to engage students on the relations between the multiple meanings in chemistry and transitions between them.

References

1. Bradley J. D. and Brand M., (1985), Stamping Out Misconceptions. J. Chem. Educ., 62(4), p. 318.
2. Bunce D. M., Gabel D. L. and Samuel J. V., (1991), Enhancing Chemistry Problem-Solving Achievement Using Problem Categorisation, J. Res. Sci. Teach., 28(6), 505–521.
3. Chandrasegaran A. L., Treagust D. F. and Mocerino M., (2007), The development of a two-tier multiple-choice diagnostic instrument for evaluating secondary school students’ ability to describe and explain chemical reactions using multiple levels of representation. Chem. Educ. Res. Prac., 8(3), 293–307.
4. Chiu J. L. and Linn M. C., (2012), The role of self-monitoring in learning chemistry with dynamic visualizations. In A. Zohar and Y. J. Dori (ed.), Metacognition in Science Education: Trends in Current Research (pp. 133–163). Dordrecht: Springer.
5. Davis E. A., (2003), Prompting middle school science students for productive reflection: Generic and directed prompts. J. Learn. Sci., 12, 91–142.
6. Davis E. A. and Linn M. C., (2000), Scaffolding students’ knowledge integration: Prompts for reflection in KIE. Int. J. Sci. Educ., 22(8), 819–837.
7. De Jong O. and Taber K. S., (2007), Teaching and learning the many faces of chemistry. In S. K. Abell and N. G. Lederman (ed.), Handbook of Research on Science Education (pp. 631–652). New Jersey: Lawrence Erlbaum Associates Inc.
8. De Jong O. and Van Driel J., (2004), Exploring the development of student teachers’ PCK of multiple meanings of chemistry topics. Int. J. Sci. Math. Educ., 2, 477–491.
9. Del Pozo R. M., (2001), Prospective teachers’ ideas about the relationships between concepts describing the composition of matter. Int. J. Sci. Educ., 23, 353–371.
10. Department of Education, (2005), Subject assessment guidelines: Physical Sciences. Pretoria: Government Printer.
11. Department of Education, (2009), Retrieved April 2, 2009, from http://www.education.gov.za/LinkClick.aspx?fileticket=91j%2B7Xs3Qlc%3D&tabid=609&mid=1700.
12. Flavell J. H., (1981), Cognitive monitoring. In W. P. Dickson (Ed.), Children’s oral communication skills. New York: Academic.
13. Gabel D. L., (1998), The complexity of chemistry and implications for teaching. In B. J. Fraser and K. J. Tobin (ed.), International handbook of science education (pp. 233–248). Great Britain: Kluwer Academic Publishers.
14. Gabel D. L., (1999), Improving teaching and learning through chemistry education research: A look to the future. J. Chem. Educ., 76, 548–554.
15. Hoffmann R. and Laszlo R., (1991), Representation in chemistry. Angew. Chem., 30, 1–16.
16. Johnstone A. H., (1982), Macro- and micro-chemistry. Sch. Sci. Rev., 64(227), 377–379.
17. Johnstone A. H., (1991), Why is science difficult to learn? Things are seldom what they seem. J. Comput. Assist. Learn., 1, 75–83.
18. Johnstone A. H., (1993), The development of chemistry teaching: A changing response to changing demand. J. Chem. Educ., 70(9), 701–705.
19. Johnson R. B. and Onwuegbuzie A. J., (2004), Mixed methods research: A research paradigm whose time has come. Educational Res., 33(7), 14–26.
20. Kokkotas P., Vlachos L., Kouladis V., (1998), Teaching the topic of the particulate nature of matter in prospective teachers’ training courses. Int. J. Sci. Educ., 20, 291–303.
21. Kozma R. B., (2003), The material features of multiple representations and their cognitive and social affordances for science understanding. Learn. Instr., 13(2), 205–226.
22. Kozma R. B. and Russell J., (1997), Multimedia and understanding: Expert and novice responses to different representations of chemical phenomena. J. Res. Sci. Teach., 34, 949–968.
23. Krajcik J., (1991), Developing students’ understanding of chemical concepts. In S. Glynn, R. Yeany and B. Britton (ed.), The psychology of learning science (pp. 117–147). Hillsdale: Erlbaum.
24. Lee K. W. L., (1999), A comparison of university lecturers’ and pre-service science teachers’ understanding of a chemical reaction at a particulate level. J. Chem. Educ., 76(7), 1008–1012.
25. Nakhleh M. B. and Krajcik J. S., (1994), Influence of levels of information as presented by different technologies on students’ understanding of acid, base and pH concepts. J. Res. Sci. Teach., 31, 1077–1096.
26. Onwu G. and Randall E., (2006), Some aspects of students’ understanding of a representational model of the particulate nature of matter in chemistry in three different countries. Chem. Educ. Res. Prac., 7(4), 226–239.
27. Rozenblit L. R. and Keil F. C., (2002), The misunderstood limits of folk science: An illusion of explanatory depth. Cognitive Sci., 26, 521–562.
28. Schummer J., (1998), The chemical core of chemistry l: A conceptual approach. Hyle, 4, 129–162.
29. Sirhan G., (2007), Learning difficulties in chemistry: an overview. J. Turk. Sci. Educ., 4(2), 2–20.
30. Stains M. and Talanquer V., (2007), Classification of chemical substances using particulate representations of matter: An analysis of student thinking. Int. J. Sci. Educ., 29, 643–661.
31. Steenberg E., (2006), Investigation of the development of bridging students’ cognitive skills relevant to the use and understanding of chemical formulae and equations. Unpublished MSc Thesis, University of South Africa, Pretoria.
32. Taber K. S., (2009), Learning at the symbolic level. In J. K. Gilbert and D. F. Treagust (ed.), Multiple Representations in Chemical Education (pp. 75–108). Dordrecht: Springer.
33. Treagust D. F. and Chittleborough G., (2001), Chemistry: A matter of understanding representations. In J. Brophy (Ed.), Subject–specific instructional methods and activities. (pp. 327–346). Oxford: Elsevier Science.
34. Treagust D. F., Chittleborough G. and Mamiala T. L., (2003), The role of sub-microscopic and symbolic representations in chemical explanations. Int. J. Sci. Educ., 25(11), 1353–1368.
35. Valanides N., (2000), Primary student teachers understanding of the particulate nature of matter and its transformations during dissolving. Chem. Educ.: Res. Prac. Eur., 1, 249–262.
36. Wu H. K., Krajcik J. S. and Soloway E., (2001), Promoting understanding of chemical representations: students use of a visualisation tool in the classroom. J. Res. Teach., 38, 821–842.

---