View PDF VersionPrevious ArticleNext Article
 
DOI: 10.1039/C9RP00018F (Paper) Chem. Educ. Res. Pract., 2019, 20, 760-771

Multiple representations in the development of students’ cognitive structures about the saponification reaction

Mónica Baptista *, Iva Martins , Teresa Conceição and Pedro Reis
Instituto de Educação da Universidade de Lisboa, Lisbon, Portugal. E-mail: mbaptista@ie.ulisboa.pt; ivamartins@campus.ul.pt; mariaconceicao@campus.ul.pt

Received 15th January 2019 , Accepted 22nd May 2019

First published on 24th May 2019

Abstract

The purpose of this study was to understand how the use of multiple representations (MR), during a sequence of lessons on the saponification reaction, can help students develop their cognitive structures. We examined (i) the effect of the teaching sequence with MR on the development of the students’ cognitive structures and (ii) how, according to the students, that sequence helped them to develop their cognitive structures. This study was conducted in three Grade 12 classes, including a total of 68 students. This study follows a pre-experimental one group pretest–posttest design, providing an intervention during the experiment. We used two data collection instruments: a Word Association Test (WAT) and a focus group interview. The results obtained from the WAT show that the students’ cognitive structures progressed from the pretest to posttest, with an increase in the number of response words and connections between words, and with a change in the nature of these connections. The results suggest that this development was influenced by the way students used MR, making use of the three MR functions proposed by Ainsworth (2006, 2008). The students recognized that the combination of the three MR functions allowed them to move across representation levels (macroscopic, symbolic and submicroscopic) and to develop a deeper and more structured understanding of the concepts related to the saponification reaction.

Introduction

Chemistry is often presented as a set of symbols, formulas and equations about objects and phenomena, invisible to the naked eye, making students see it as an abstract science and a difficult subject to understand (Taber and García-Franco, 2010). In fact, conceptual understanding in Chemistry requires observing phenomena at three levels: macroscopic (e.g., seeing and manipulating objects, experimenting with and describing the properties of materials); submicroscopic (e.g., understanding and explaining observations in terms of non-visible and abstract objects such as atoms, ions and molecules); and symbolic (e.g., translating the understanding of observations through chemical equations, analogies and model kits) (Johnstone, 1982). Studies have emphasized that combining these three levels is essential for effectively learning Chemistry (e.g., Talanquer, 2011). However, students have difficulties in moving from the macroscopic level to the symbolic and submicroscopic levels (Prain et al., 2009). This movement implies that students must develop their logical and abstract thinking skills (Blake and Nordland, 1978), as well as their cognitive structures, which involve connections between terms, concepts and processes (Derman and Eilks, 2016).

In this sense, considering the importance of students being fluent at all three levels of Chemistry (macroscopic, submicroscopic, and symbolic), it is essential to use various learning resources that help them move between these representational levels. The use of representations has been pointed out by several authors as one of the facilitators for explaining phenomena in Chemistry, enhancing students’ conceptual comprehension (e.g., Ainsworth, 1999; Ardark and Akaygun, 2004; Gilbert, 2008; Prain and Tytler, 2012; Waldrip and Prain, 2012). Representations are classified according to their real or mental domain. That is, a real representation is one that is presented to the student in the form of a drawing, videos, concrete models, text, and hands-on materials among others. The mental representation is then constructed in the student's mind (Gilbert, 2005). For the scope of this work, only representations in the real domain were considered.

The use of two or more representations when studying a concept is considered as learning with multiple representations (MR) (Tsui and Treagust, 2013; Ainsworth, 2014). MR have a fundamental role in understanding concepts and the relationships between them (Gilbert and Treagust, 2009; Tsui and Treagust, 2013), which implies the development of students’ cognitive structures, and as a field of research in education that has been gaining relevancy. For example, in the study of Al-Balushi and Al-Hajri (2014), with 28 students from the 11th grade, results show that the use of MR – animations and molecular model kits (ball-and-stick) – students to have a better scientific understanding, chirality, bond angles, stereochemistry and spatial arrangement of atoms within molecules, since these MR allowed them to manipulate and spatially visualize the molecules, transposing from 2D to 3D and vice versa. Rau (2015), in a study with 158 undergraduate students from a general Chemistry introductory course, found that the use of virtual simulations (with graphical MR) significantly improved the students’ conceptual understanding of atomic structure and chemical bonding concepts.

Although most research indicates that the use of MR favors students’ conceptual learning in Chemistry, there still aren’t many studies concerning how MR influence the development of students’ cognitive structures. This study intends to contribute to the increase of knowledge in this area. The following research questions guided our study:

(1) What is the effect of a sequence of lessons on the saponification reaction making use of MR on the development of students’ cognitive structures?

(2) From the students’ perspective, how does the use of MR, during the sequence of saponification reaction lessons, support the development of their cognitive structures?

Students’ cognitive structures

Construction of knowledge is an active process where the interaction between students and the environment is fundamental, as well as their life experiences and previous knowledge, for the understanding of concepts and their relations (Piaget, 1978; Vygotsky, 1978). This constructivist perspective of learning requires that educators and researchers understand students’ cognitive structures and how they are developed (Derman and Ebenezer, 2018). There is not only one accepted definition for cognitive structures (Derman and Eilks, 2016). In this study, we chose to use the concept of cognitive structures when we talk about relationships, established by students, between concepts, terms or processes (Liu and Ebenezer, 2002; Derman and Eilks, 2016).

In the last few decades, several methods have been proposed to study cognitive structures, such as flow maps, concept maps, word association tests and tree construction (Tsai and Huang, 2002). Word Association Tests (WAT) were developed by Johnson (1967, 1969). In WAT implementation, the researcher or educator selects relevant concepts (stimulus words) and asks students to write words associated with these concepts (response words) in a certain period of time (Nakiboglu, 2008). According to Bahar et al. (1999), WAT are considered as a “snapshot” of the students, since they do not have the time to prepare themselves and thus what is visible is the “raw state” of their cognitive structure. Through the quantity and quality of the associated words, the understanding of the concept can then be evaluated. However, according to Nakiboglu (2008), WAT present limitations because they do not clarify the nature of the relations that students establish between terms. Therefore, the analysis of students’ cognitive structures should be complemented with interviews, free writing or the construction of concept maps (Nakiboglu, 2008; Derman and Eilks, 2016).

Several researchers have conducted studies on the use of WAT in Chemistry. For example, Nakiboglu (2008) developed a study with 40 elementary mathematics education students, with a minor in elementary science education, whose objective was to understand the development of students’ cognitive structures about the atomic structure. A WAT was used in two moments (a pretest and a posttest), with ten stimulus words. The results of this study showed that after the intervention the students expanded their cognitive structure, with respect to concepts and their relationships – for example, in the posttest on understanding the concepts related to the quantum mechanical model of the atom, such as quantum numbers, the principal quantum number and the type of orbital.

In another study, Şendur et al. (2011) used the same method (WAT) to understand if there were changes in the cognitive structures of Chemistry pre-service teachers (n = 56) on acids and bases. Results showed that after the intervention the participants created a much higher number of associations among the ten stimulus words. For example, the students were able to make meaningful associations between acid–base concepts, such as acid/pH, base/hydroxyl, titration/indicator, neutralization/acid–base, hydrolysis/water, electrolyte/conductivity and indicator/phenolphthalein.

Derman and Eilks (2016), in a study on the concept of dissolution, with 157 students from the 11th grade, using a WAT showed that the students created few associations among the eight stimulus words allusive to the concepts of solutions and the corpuscular nature of matter. Therefore, the researchers concluded that the students have a restricted understanding of the corpuscular nature of matter and the topic of solutions.

More recently, Yildirir and Demirkol (2018) investigated students’ cognitive structures on physical changes and chemical changes with a WAT, involving 153 students from the 6th grade using six stimulus words. Results showed few associations between the concepts “particle” and “chemical change”. The researchers concluded that the students do not understand a chemical change as a phenomenon occurring at the submicroscopic level. This study also led to the conclusion that there are difficulties in distinguishing physical changes from chemical changes.

Similarly, Nakiboglu (2017), in a study of 126 students from the 8th grade, also on physical changes and chemical changes, used a WAT with eight stimulus words to research their cognitive structures. Through the results obtained, it was possible to conclude that the cognitive structures of these students were not very consistent, with many concepts isolated in separate islands, and that the concept of atoms was too abstract for the students at this level.

The results also suggested the need to monitor changes in students’ cognitive structures throughout their learning, through the use of posttests, as performed by Derman and Ebenezer (2018) through the application of a WAT in a study that aimed at knowing the effect of the levels of representation (macroscopic, submicroscopic and symbolic) in the development of cognitive structures of 40 primary pre-service teachers. The authors applied a pretest, before the intervention, and a posttest, after the intervention, with two stimulus words: physical changes and chemical changes. Results showed that the participants developed their cognitive structures about the transformations at the three levels of representation.

MR functions

In Chemistry the understanding of concepts and connections across them, as well as of terms and processes, may involve the use of MR (Prain et al., 2009; Ainsworth et al., 2011). In fact, depending on the information they contain and students’ prior knowledge, MR can be used differently (Cook et al., 2008; Tsui and Treagust, 2013) and provide various contributions to the development of students’ cognitive structures (Derman and Eilks, 2016). In this sense, the use of a framework to analyze how students use MR in their learning process, i.e., to understand concepts and the relationships between them, in Chemistry can be useful (Won et al., 2014).

In this research work, we have adopted Ainsworth's framework (2006, 2008) in order to analyze how, according to students, the use of MR helped them to develop their cognitive structures during a sequence of lessons on the saponification reaction. Ainsworth (2006, 2008) developed a taxonomy on the use of MR in multimedia educational contexts. According to this author, MR have three functions: complementary roles, encouraging interpretations and constructing deeper understanding.

The complementary role of MR is to provide complementarity aspects of information about a concept to make learning easier. Another function of this MR role is to take advantage of the representation process, for example, by using two representational processes, such as manipulation of laboratory materials to explore a chemical reaction and photographs of students’ observations (Pozzer and Roth, 2003).

Encouraging interpretations offers students with more accessible representations in order to encourage and support their interpretation of more challenging representations. Representations are considered to be more accessible when the information or the representing process is more familiar to the students. For example, a video with daily life information, which is relevant in the eyes of students (Holbrook, 2008), combined with molecular kits can enable the visualization of Chemistry phenomena at various levels of comprehension.

MR can also be used for constructing deeper understanding. Doing so by relation means that students make connections across representations (Won et al., 2014). For example, students may relate molecular models built on virtual tools (Wu et al., 2001) to the chemical equation of a reaction to explain chemical transformation from the submicroscopic point of view. Constructing deeper understanding by extension means that students apply what they have learned to another representation (the two representations are essential for learning the concept). Constructing deeper understanding by abstraction means that students extend what they learnt about a given concept in a particular context and with MR to further learning.

Method

Research design

This study follows a pre-experimental one group pretest–posttest design, providing an intervention during the experiment (Creswell, 2002). This design facilitates the comparison of students’ cognitive structures, before (moment 1) and after (moment 2) a sequence of lessons on the saponification reaction, using MR. All research was approved by our Institutional Ethical Review Board.

Context

This study was conducted in three Grade 12 classes, a total of 68 students (36 = 57% female and 32 = 43% male; age range 17–19) who attended a school in the Lisbon metropolitan area. The students belonged to the upper-middle class. In Portugal, Physics and Chemistry are independent subjects starting in the 7th grade. Students’ first contact with Organic Chemistry is in the 9th grade (age 15 years old). In the 12th grade, for students who are in the Science and Technology curricular area, Chemistry and Physics are separate subjects. In the 12th grade Chemistry curriculum, students again have the subject Organic Chemistry where they encounter concepts such as acyclic and cyclic aliphatic hydrocarbons, aromatic hydrocarbons, functional groups and chemical reactions between organic compounds, among others. Prior to the development of this study, all students had already studied hydrocarbon nomenclature and functional groups. This study is focused on a chemical reaction, the saponification reaction. The three classes on the saponification reaction (total duration 360 minutes) were taught by three Chemistry teachers (with professional experience ranging from 12 to 15 years).

Each teacher conducted the lessons in their class as presented in Table 1. The students in the three classes worked in groups, with three or four students each. The authors of this article observed all three classes.

Table 1 Sequence of lessons on the saponification reaction using MR
Lesson Students’ tasks
Lesson 1 (90 minutes) Visualization of a video about soap production. The video describes macroscopically the production of soap in an existing industry in the national context. The aim of the video is to engage students with the topic (saponification reaction), using a local context as a starting point.
Discussion of the importance of the soap industry in our country.
Lesson 2 (135 minutes) Laboratory activity on soap production: selection of materials and reagents; planning the laboratory activity; and completing the laboratory activity. During the development of the activity students resorted to various representations, such as photos and laboratory materials.
Macroscopic explanation for what they observed (using photos and laboratory materials).
Lesson 3 (135 minutes) Symbolic explanation for the saponification reaction (using model kits and chemical equations).
Submicroscopic explanation for the saponification reaction.
Sharing and discussing in class the explanations for each group (considering the three levels of representation).

Data collection

In this study we used two data collection instruments: a Word Association Test (WAT) and a focus group interview. The WAT aimed to describe the effect of the sequence of lessons on the saponification reaction with MR in the development of the students’ cognitive structures. For this, the WAT was applied in two moments of the study: moment 1/pretest (M1) – three weeks before lesson 1; moment 2/posttest (M2) – three weeks after lesson 3. To stimulate the association of words, we gave the students four words, each on a separate blank sheet: ester, alcohol, soap and basic solution. The stimulus words selected are related to key concepts of a saponification reaction, known as the process for producing soap. Typically, a vegetable oil or an animal fat (mainly constituted by esters – triacylglycerols) reacts with a strong base (e.g., sodium hydroxide) in aqueous medium, to produce glycerol and soap (sodium salts of hydrolyzed free fatty acids) (Mabrouk, 2005) (Fig. 1).
image file: c9rp00018f-f1.tif
Fig. 1 Saponification reaction.

Stimulus words appeared in random order. The students were asked to write as many terms associated with the stimulus words as they could and to write a sentence including each one of the stimulus words and their response words (Nakiboglu, 2008). At both moments, M1 and M2, the students had approximately 10 minutes to perform the task (Hovardas and Korfiatis, 2006; Derman and Eilks, 2016). To ensure the content validity of the WAT, the curriculum was analyzed by the second author of this paper and two experienced school chemistry teachers provided their opinions and certified the suitability of stimulus words for this work (Nakiboglu, 2008).

The interview was conducted four weeks after lesson 3 and aimed to determine how, from the students’ perspective, the use of MR, during the sequence of saponification reaction lessons, supported the development of their cognitive structures. Sixteen students volunteered to participate in the interview, six from one class and five from each of the other two classes. The students were then divided into two groups, each with eight students. Group 1 contains students S1 to S8 and group 2 students S9 to S16. Interviews were scheduled according to the availability of each of the groups and conducted by two of the researchers. The choice of focused group interview (Patton, 1990) was intended for the students to feel more comfortable and confident in the answers to the questions posed. Each interview lasted 30 minutes.

Data analysis

With regard to the WAT, qualitative data collection was combined with quantitative and qualitative analysis procedures (Creswell, 2002). The WAT data analysis was performed based on the response frequency map method (Nakiboglu, 2008). It began by analyzing the terms associated with the stimulus words. Words that were “meaningful”, i.e., the response words of the students related to the saponification reaction, were counted and validated as response words. The frequency table was constructed by placing the stimulus words in the first row, the pretest (M1) and posttest (M2) moments in the second row, and the response words in the first column.

To establish the inter-judge reliability, the data from the WAT were analyzed independently by the second author (who holds a PhD in Chemistry) and the fourth author (who is a Science Education expert). In order to compare the analysis, we used as the criterion the counting of the total of different response words. Following Miles and Huberman (1994), the consensus between the authors was higher than 90% for the pretest and posttest (Table 2).

Table 2 Percentages of inter-judge reliability for the pretest and posttest
Stimulus word Pretest (%) Posttest (%)
Ester 96 97
Alcohol 95 97
Soap 94 95
Basic solution 96 95

In addition to the frequency table, we constructed a table illustrating the number of different responses to a given stimulus word at the two moments (Shavelson, 1974; Derman and Eilks, 2016). The number of different responses to a word is a direct indication of the “meaningfulness of the key concept” and a word without associations has no meaning (Bahar et al., 1999).

Taking into account the data presented in the frequency table, the students’ cognitive structure maps were constructed at moments M1 and M2. For this, the highest frequency interval was initially established, which corresponded to 61 ≤ f ≤ 70 for both moments. The lowest frequency level was set as 31 ≤ f ≤ 40 for M1 and 51 ≤ f ≤ 60 for M2, because all the stimulus words appeared in the maps at these frequency levels (Nakigoblu, 2016). The construction of the maps was then performed by placing the stimulus words in a frame, and the response words without a frame, and drawing lines with arrows from the stimulus words to the response words (the direction of the arrow indicates the direction of the relationship) considering the established frequency ranges, which define the cells of the map. The width of the frames and arrows is determined by the frequency value of the response word to the stimulus word and represents the strength of the associations: the thicker the arrow, the greater the frequency, and consequently the greater the association, prompting these words to be placed in the first cell. In this way, it is possible to have an idea of the strength and direction of the associations and to obtain a relation between the concepts and the students’ cognitive structures (Nakiboglu, 2008).

In order to distinguish the nature of the connections that students establish between words, and to suppress the limitation of the WAT presented by Nakiboglu (2008), a qualitative analysis of the sentences written by the students was performed, allowing us to include examples of answers for each of the frequency intervals.

Regarding the analysis of the interviews, in this research, the functions of MR presented by Ainsworth (2006, 2008) were used as categories for the analysis (each function corresponds to a category). We then used these categories to construct our analytical framework (Table 3). Each category was divided into subcategories, depending on the specific function of the MR for the development of students’ cognitive structures on the saponification reaction. For example, in the Complementary Role category, MR have different information about the saponification reaction and/or the same information in different representations. Therefore, it is further subdivided into two subcategories. We proceed similarly for the other two categories. Two researchers analyzed the interview transcripts about how the students’ use of MR during the course of lessons helped them to develop their cognitive structures, and independently coded the data taking into account the proposed framework. Disagreements or doubts were discussed in order to reach a consensus.

Table 3 Analytical framework
MR role Subcategory Description
Complementary role Using different representation processes Using a video on the saponification reaction and molecular models to write and interpret the chemical equation for the saponification reaction
Using MR with different information Using laboratory materials for the production of soap and recording the observations through photographs to explain the saponification reaction
Encouraging interpretations Using familiarity to interpret a new representation Using the video and photographs as primary representations for constructing molecular models of the reagents and the products of the saponification reaction
Constructing deeper understanding Making connections across representations (by relation) Relating molecular models and the chemical equation to explain the chemical transformation of the saponification reaction
Extending the understanding of a representation to another (by extension) Using a video on soap making to describe the phenomena observed during the laboratory activity
Generalizing the chemical equation of saponification reactions (by abstraction) Examining the molecular models and the chemical equation of the reaction developed in the laboratory for generalizing the chemical equation of saponification reactions

Results

Results from the WAT

In order to examine the effect of the sequence of lessons on the saponification reaction using MR in the development of the students’ cognitive structures, a table of frequencies (Table 4) was built, which indicated the numbers of response words per stimulus word at two different moments: M1 (pretest) and M2 (posttest). We also built Table 5, which gives the numbers of different responses per stimulus word, in the pretest (M1) and in the posttest (M2). Overall, from M1 to M2, there was an increase of 27 different responses to the stimulus words. The words with a more significant increase in the response words were “ester” (three response words at M1 to twelve response words at M2) and “basic solution” (three response words at M1 and eleven response words at M2). The data presented in Table 4 were used to construct the maps of the students’ cognitive structures for each moment: M1 (Fig. 2) and M2 (Fig. 3).
Table 4 WAT frequency table for the pretest (M1) and posttest (M2)
Stimulus word
Ester Alcohol Soap Basic solution
Response word M1 M2 M1 M2 M1 M2 M1 M2
Ester 28 3 66 34
Alcohol 32 1 33 16
Soap 2 67 43 65
Basic solution 42 26 26 62
Polar 51 56
Non-polar 52 54
Solute 61 63
Solvent 57 33 62 65
Micelle 16 37
Fat 43 8 65 67 9
Carbon chain 32 24 36 22 26
Hydroxyl group 12 51 52 17 45 47 52
Carboxylic acid 24 32
Glycerol 17 38 18 12
Hydrolysis 3 21 17
Chemical transformation 14 5 8 14
Bond break 6 9
Saponification 4 13
Nucleophile 5

Table 5 Total different responses in the pretest (M1) and posttest (M2)
Stimulus word Total number of different responses
M1 M2
Ester 3 12
Alcohol 3 8
Soap 9 14
Basic solution 3 11
Total 18 45


image file: c9rp00018f-f2.tif
Fig. 2 Map of students’ cognitive structures at M1.

image file: c9rp00018f-f3.tif
Fig. 3 Map of students’ cognitive structures at M2.

In Fig. 2, which refers to the students’ cognitive structures at M1, there are four association levels and each one is characterized by the presence of two isolated islands. At the strongest association level of the students’ cognitive structures, in the 61 ≤ f ≤ 70 frequency range (Level 4), only two stimulus words – “soap” and “basic solution” – appear. “Soap” is associated with only one response word – “fat” – and “basic solution” is associated with two response words – “solute” and “solvent”. At this level, the arrows are wider, which indicates that the associations are strong. The strongest association is between the stimulus word “soap” and the response word “fat”. Some excerpts of the students’ phrases for this frequency range reveal the nature of the connections: “soap is used to remove fat” and “in a basic solution, the solvent is water and the solute can be sodium hydroxide”. These examples show that, at M1, students relate soap to fat removal (63 out of the total number of students made this connection) and that only two of the students identify fat as a reactant in soap production. Moreover, it is evident that most of the students understand the concept of solution as they identify water as the solvent and sodium hydroxide as the solute. At level 3 (51 ≤ f ≤ 60), the word “soap” is now also associated with two new response words: “polar” and “non-polar”. At this level occurs the stimulus word “alcohol”, to which students coupled two response words: “hydroxyl group” and “solvent”. Additionally, the stimulus words “alcohol” and “basic solution” are connected to each other via the response word “solvent”. Some of the sentences written by the students were: “the soap has a polar end and a non-polar end”; “the functional group of alcohols is the hydroxyl group”; and “alcohol is the solvent in an alcoholic solution”. From these examples, it can be inferred that students can make associations that show an understanding of stimulus words and response words, namely the amphipathic nature of soap and the functional group of alcohols. However, most students associated alcohol with the solvent in an alcoholic solution, but not as a product of the saponification reaction. At level 2 (41 ≤ f ≤ 50), a new association appears between the stimulus word “basic solution” and the response word “hydroxyl group”, which is represented by the thinner arrow in this cell. As stated by a student “in a basic solution there are hydroxyl groups, i.e., OH is present”. This denotes that students recognize the chemical nature of the solution, according to Arrhenius’ theory of acids and bases. In the frequency range 31 ≤ f ≤ 40 (Level 1), the stimulus word “ester” appears for the first time and it is connected to the word “carbon chain”, which in turn is also linked to the stimulus word “alcohol”. The weaker nature of these associations is illustrated by the thinnest arrows. The phrases uttered by the students provide evidence that they know that esters and alcohols have carbon chains: “esters belong to the RCOOR’ type, where R are the carbon chains” and “alcohols have a hydroxyl group linked to a carbon that belongs to a carbon chain”.

The cognitive structures of the students in the posttest (M2) are presented in Fig. 3, and it is clear that, when compared to the ones in the pretest (Fig. 2), they have more stimulus words per frequency level and all the stimulus words are connected. Accordingly, at level 2 (61 ≤ f ≤ 70), which is the strongest level of association of the students’ cognitive structures, three of the four stimulus words appear interconnected, forming a network. In this network, there are strong associations between the stimulus words “ester” and “soap”; and “soap” and “basic solution”. Other strong associations, which were also present at M1, are between the stimulus word “soap” and the response word “fat” and between the stimulus word “basic solution” and the response words “solute” and “solvent”. These features, namely the network obtained between three of the four stimulus words, suggest the occurrence of a conceptual organization as a result of the teaching sequence with MR. Furthermore, a qualitative analysis of the students’ phrases shows that the nature of the connections at M2 is related to the saponification reaction. For instance, it is stated that “in order to obtain soap we have to break a fat (olive oil), which is an ester, and add a basic solution (NaOH)”; and “soap is formed from an ester (a fat), with a basic solution, in which the solvent is water and the solute is sodium hydroxide”. In the last cell in Fig. 3, level 1 (51 ≤ f ≤ 60) is characterized by the presence of all stimulus words, whereas at M1 this only occurred in the frequency range 31 ≤ f ≤ 40. At this level the remaining stimulus word, “alcohol”, appears linked to a new response word, “hydroxyl group”, which, in turn, is also associated with the stimulus word “basic solution”, forming a unique cluster. In addition, two new associations emerge in this network from the stimulus word “soap” to the response words “polar” and “non-polar”. From the sentence part of the WAT, students wrote: “the functional group of an alcohol is the hydroxyl group, –OH”; and “soap is a complex molecule that has a polar and a non-polar end”. The nature of these associations shows that students make connections that demonstrate an understanding of stimulus words and response words.

Results from the interviews

As previously mentioned, in order to know the students’ perspective on the use of MR during the sequence of lessons on the saponification reaction for the development of their cognitive structures, the data from the group interviews were analyzed and fell into three categories (Table 3).
Complementary role. According to the students, the use of MR with different information was important to help them relate concepts and understand them. This was emphasized by the two groups interviewed. As student S1 (group 1) noted: “The representations used have helped us understand the soap production reaction more deeply”. The students had been given different information, helping them to understand and relate the concepts. When asked to elaborate on their answer, the students in group 1 provided the following answers:

S2 – The video allowed us to visualize the industrial process of soap production. Here we get to know what the reagents and reaction products were.

S3 – It helped us know that to make soap we need fat and a basic solution. Soap is not the only product, we also have alcohol. Then it was also important to find about the soap industry in Portugal which I knew nothing about.

S5 – But I would say we did not know how it happened because we could not understand what we couldn't see – the structure of molecules, bonds and breakings of bonds, movement of molecules, atoms and sodium ions.

S3 – And the molecular models provided information about what we could not observe, i.e. how to write the chemical equation of the saponification reaction.

S4 – And explain the chemical transformation that occurs.

The students mentioned that the video and the molecular models they made using kit models contributed different information to writing and interpreting the chemical equation of the reaction and reaching a submicroscopic explanation for the phenomenon. According to the students, using distinct processes of information representation, for instance, using laboratory materials to synthesize soap and using photographs for recording observations helped them explain the chemical reaction of saponification. For example:

S9 – The part of the experiment in which we made the reaction in the laboratory was excellent to observe what was happening. It was actually very important. The photographs we took from the lab helped us later explain what was happening.

Interviewer – Can you elaborate further?

S9 – One thing is to observe the color of olive oil and what has changed until it became soap, which includes the color of the soap, the odor, whether or not there was a change in temperature, whether there was release or absorption of energy in the form of heat. The photographs we took helped us describe and explain in depth what was happening. This is because we can look at the photographs several times and think.

Based on the previous excerpt, we realized that the students used two representation processes: laboratory materials and photographs. This helped them make meaning of the phenomenon (submicroscopic level). In terms of providing information and enabling different representation processes, the complementary role of MR allowed the students to better grasp the three levels of representation.

Encouraging interpretations. For the students, the use of more comprehensible and familiar representations (those associated with the macroscopic explanation for the saponification reaction – video and photographs) also made it easier for them to acquire and relate scientific concepts associated to the saponification reaction. This helped them explain the phenomenon from the symbolic and submicroscopic points of view. For example, group 1 reported the following:

S7 – The video does not show the molecular structure, the chemical equation or the properties of the soap. The video was actually important for us to engage in the laboratory activity and for what we have already mentioned about having helped us in material selection and then planning.

S2 – The laboratory part of where we get hands on is very important. Taking photographs means having snapshots of what is real, of what happened, and being able to always look back at them for visual support that can help us describe what happened.

S5 – In the photographs we can observe the changes of color when we add the base to the olive oil, the change of viscosity, but of course we cannot capture what we cannot see.

S7 – Then we move from what we observed to using molecular models. This was easier and helped us explain the reaction.

From this excerpt, we can notice that the students used two MR (video and photographs) which were more familiar and comprehensible to them to build the structures of the molecules involved in the saponification reaction, using the kit models (symbolic level). In turn, as mentioned by S7, the models helped the students develop a submicroscopic explanation for the phenomenon.

Constructing deeper understanding. Using a representation for students to extend their understanding to another representation (deeper understanding by extension) can help them understand the phenomenon and make connections across concepts. From the students’ points of view, using a video enabled them to go through this cognitive process. For example:

S4 – From the video, we took information for the planning and for carrying out the experiment.

Interviewer – Can you give an example?

S8 – For example, the materials and the reagents we used. We made that decision based on what he had seen in the video.

S9 – Another aspect was the temperature. Boiling water bath or not? Do we want it or not? We discussed what happened in the video and we related that to what we were going to do in the lab and decided we would. Thus, we came to understand how we could increase the speed of the reaction by putting the reaction mixture in a boiling water bath. Since glycerol is volatile, we could only be left with the soap.

According to the students, the use of a video (animated images with text and sound) about the soap production process helped them plan an activity to synthesize soap in the school laboratory. This allowed them to extend what they learnt from the reaction they performed. In fact, what they learned from the activity, supported by the video, allowed them to understand the importance of heating the reaction mixture and how to separate the main product from the reaction (soap). Thus, the use of MR enabled a more in-depth understanding of the process involved in the saponification reaction and of the macroscopic level.

Making connections across representations (deeper understanding by relation) is another way to help students understand the saponification reaction and make new connections across concepts. This was evident in the present study, when the students manipulated molecular models and related them to the chemical equation of the saponification reaction. For example:

S12 – It is easier to go from what we observe and then explain the reaction, as opposed to going from what we do not see.

Interviewer – Can you provide concrete examples?

S13 – It was as I said, our starting point were molecular models and relating them to the chemical equation to really explain how soap is made. The models helped us understand the geometry of the olive oil molecule and how it interacts with NaOH to make glycerol and soap, the polar and non-polar ends.

S12 – In the models it is easier to count the atoms that we have in each element and to see, to break the bonds and to form others so as to create glycerol and soap. This helps us write the chemical equation.

S9 – Yes, I agree. The 3D models helped us understand chemical and ionic formulas. They have really helped us.

S15 – Using molecular models and building the molecule of fat, glycerol, NaOH and soap on our own helps us “notice” what happened in the non-visible world and explain how soap was made, not only from using molecular models, but also with the help of the equation. It helps us realize that for the chemical reaction to occur there is a rearrangement of atoms, bonds are broken and others are formed. We also come to understand and are able to explain that the function of the NaOH solution is the nucleophile.

S13 – With all this it was easier to explain why the chemical transformation occurs, i.e., how the atoms and molecules in our olive oil reorganize and move to make soap.

Molecular models (reagents and products of the chemical saponification reaction) allowed the students to break bonds and create others and write the chemical equation of the reaction from that point. The connection across these two MR helped them to understand the chemical reaction of saponification in light of the corpuscular theory of matter (submicroscopic level). In fact, from the previous excerpt it is clear that the students explained the chemical reaction of saponification based on the corpuscular nature of the matter as they resorted to terminology such as: “there are rearrangements of the atoms, bonds are broken and others are formed, a reaction occurs” and “the molecules of olive oil reorganize and move to make soap.” The following example reinforces this idea:

S11 – Molecular models are static. We do not see the movement of the atoms or when chemical bonds break. Nevertheless, they have many potentialities because they allow us to see the bonds between atoms, the molecular structure and their geometry, i.e., what the molecular structures of olive oil, glycerol and soap are. We were able to “see” the polar and the non-polar ends.

S14 – And this is very important in order to understand the chemical equation. The chemical equation is the simplest schematic form of “seeing” the saponification reaction and identifying the reagent: the olive oil that is an ester with a long carbon chain. From the fat in basic medium, you get the soap and glycerol which is an alcohol. Chemical equations are also static, i.e., we cannot “see” particle motion or bond breaks. Still, the chemical equations and the molecular models help us think about this “act of seeing” that is invisible.

S12 – After working on our chemical equation, it was easier to come up with a general equation for soap synthesis because we already knew the general formula of the esters.

As reported by the students, relating MR – molecular models and the chemical equation – helped them visualize the spatial structures of the molecules and their rearrangement during the chemical reaction of saponification (symbolic level) and explain what happens at the submicroscopic level. Likewise, in their explanation, the students relied on the corpuscular theory of matter, showing a deeper understanding of the phenomenon under study. In addition, S12's excerpt also showed that there was an understanding of the phenomenon for most esters. For this student at least, what he learned by relating MR enabled him to develop a deeper understanding, i.e., an abstract understanding of the chemical saponification reaction.

Discussion

With the present study, we aimed to understand how the use of MR, during a sequence of lessons on the saponification reaction, can help students develop their cognitive structures. In order to accomplish this, we examined (i) the effect of the teaching sequence with MR on the development of the students’ cognitive structures and (ii) how, according to the students, that sequence helped them to develop their cognitive structures. In order to answer the first research question, a WAT was applied at two different moments: pretest (M1) and posttest (M2). Data were collected from M1 and M2 and were analyzed and compared. From Table 5, which refers to the number of different response words for each stimulus word, it is evident that this is much higher at M2 (18 different words at M1 and 45 different words at M2). These results reveal that in the posttest, after the teaching sequence using MR, the students had a deeper understanding of the concepts and were able to make more connections using new words. This means that, by increasing the number and complexity of connections, these words become more meaningful to the students (Bahar et al., 1999). The analysis of the maps of the students’ cognitive structures in the pretest (Fig. 2) and in the posttest (Fig. 3) shows that there were changes in the students’ cognitive structures as a result of the instruction sequence.

Before the sequence of lessons (moment M1, Fig. 2), the students’ cognitive structures were characterized by the presence of isolated islands at all frequency levels, which is indicative of a static organization of the cognitive structures (Derman and Eilks, 2016), and there are no direct associations between stimulus words. At M1, students displayed more disconnected ideas and, although they increased the words associated with the stimulus word “basic solution”, including two stimulus words, “alcohol” and “ester”, they could not establish a network between all the stimulus words. Furthermore, the qualitative analysis of sentences written by the students in the pretest, which was meant to evaluate the reasoning behind the connections they had made, showed that the students did not have a comprehensive and a solid understanding of the concept related to the saponification reaction.

The map of the students’ cognitive structures at M2 (Fig. 3) shows that the teaching sequence using MR made a significant difference in the students’ cognitive structures. As a result, it can be seen that the map displayed in Fig. 3 resembles a more structured arrangement, with all the stimulus words interconnected. At level 2, the strongest level in terms of the association of the students’ cognitive structures, the appearance of three out of the four stimulus words, strongly linked, can be seen: the strongest association among two stimulus words is between the words “ester” and “soap”, followed by the association between “soap” and “basic solution”. At level 1 (51 ≤ f ≤ 60), the remaining stimulus word, “alcohol”, appears and it is connected, by means of the response word “hydroxyl group”, to the preexistent cognitive structure. On the other hand, at M1, the four stimulus words only appeared at frequencies ranging from 31 to 40 and they were not all connected. This suggests that conceptual learning had occurred, as a result of the teaching sequence (Nakiboglu, 2008). Also, qualitative results gathered from the students’ sentences confirmed that the nature of the connections reflects the understanding of the concepts related to the saponification reaction.

Although there are several differences between the pretest and posttest, some features are similar, and they are important for the interpretation of the students’ existing cognitive structure. For instance, at M1, the association between “soap” and “fat” was the strongest in terms of frequency in the students’ cognitive structures and the same association also appears as one of the strongest at M2. Another identical association can also be found at level 4 at M1 and level 2 at M2 between the stimulus word “basic solution” and the response words “solute” and “solvent”. According to Nakiboglu (2008), these similarities can be used to infer the influence of the students’ pretest cognitive structures in the construction of their new cognitive structures, as a result of the learning process. Nevertheless, the same author acknowledges the fact that one of the limitations of his study is that “the WAT by itself could not give any insight into which type of link is being made by the student” (Nakiboglu, 2008, p. 321). In other words, without any qualitative analysis, we cannot draw conclusions on the reasoning behind the connections made by students. In fact, in the present study, the qualitative analysis of the students’ phrases shows that the nature of the connections is different at M1 and M2. Whereas at M1 students associate “soap” with “fat” removal in the washing process, in the posttest this connection is related to the saponification reaction, and they were able to identify “fat” as a reactant in this reaction that produces “soap”, the main product of the saponification reaction. As to the other common association, although at M1 the students identified sodium hydroxide as a possible solute in a basic solution, only at M2 did they indicate that this basic solution, in which sodium hydroxide was the solute, is a reactant in the saponification reaction. Accordingly, attention should be paid when interpreting the maps of the students’ cognitive structures: even though they may share some identical links, the nature of those links is different at M1 and M2.

In light of these findings, we can state that, as a result of an instruction sequence with MR about the saponification reaction, there was a development in the students’ cognitive structures. Moreover, the information about the students’ cognitive structures should be addressed by complementary methods to WAT since it is important to determine not only the connections made by students but, more importantly, the nature of those connections. In our study, we were able to establish that the ones present in the pre-existing cognitive structure, although important, didn’t reflect any relation with the saponification reaction. Overall, our findings are in accordance with previous works that describe the use of MR as a tool to increase students’ outcomes (Al-Balushi and Al-Hajri, 2014; Rau, 2015; Al-Balushi et al., 2016), and the changes that cognitive structures undergo as a result of the learning process (Nakiboglu, 2008; Derman and Ebenezer, 2018).

Using Ainsworth's taxonomy (2008) allowed us to examine how, according to the students, the sequence of lessons on the saponification reaction using MR helped them develop their cognitive structures. The results showed that, from the students’ perspective, MR made it easier for them to understand the concepts and connections across them related to the saponification reaction. In fact, in order to develop their cognitive structures about the chemical phenomena under study, it was important to access the three MR functions, namely, complementary role, encouraging interpretations and constructing deeper understanding.

As to the first function, the complementary role of MR with different information, the students used the video (macroscopic level) and molecular models (symbolic level) to write down the chemical equation (symbolic level) and explain the chemical reaction of saponification at the submicroscopic level. Moreover, they also used two processes of representation, laboratory materials and photographs (both at the macroscopic level) to help them understand the phenomena at the submicroscopic level. In fact, using MR with different information and different representation processes facilitated the students’ transition in terms of their understanding across representation levels (macroscopic – symbolic – submicroscopic). It is noticeable that the complementarity role of MR makes students develop their cognitive structures as it enables them to make use of multiple types of knowledge in Chemistry (Derman and Ebenezer, 2018). The findings from this research study are consistent with those from previous studies on the role of this function on students’ understanding of scientific concepts (Won et al., 2014).

Regarding the second function of MR, encouraging interpretations, according to the students, the use of the video and photos (macroscopic level), since they are more accessible and familiar, helped them in the construction of the structures of the molecules involved in the saponification reaction (symbolic level). In addition, it helped them to build an effective explanation for the phenomena at a submicroscopic level. In fact, research has shown that using MR to encourage students to explain more challenging representations is one way to enable the comprehension of concepts (Won et al., 2014). Just like the complementarity role of MR, this function also enables students to, by familiarity, understand the phenomena at the three representation levels (macroscopic – symbolic – submicroscopic).

As to the third function, the students recognized that the use of MR enabled them to more deeply understand the saponification reaction, namely: (i) to extend the understanding of a video about the soap production process to an activity plan (i.e., using laboratory materials); (ii) to make connections between molecular models and the chemical equation so as to understand the saponification reaction and (iii) to extend the understanding of the phenomena to esters in general (by abstraction). With this MR function, students developed a deeper understanding of the concept and its connections. In doing so, they were able to mobilize multiple types of knowledge in Chemistry (macroscopic – symbolic – submicroscopic) to explain the saponification reaction. This transition between the three representational levels is referred to in the literature as being fundamental to the comprehension of Chemistry (Taber, 2001; Gilbert and Treagust, 2009).

Conclusions

Effective student understanding of the saponification reaction entails their ability to explain it, through the use of multiple knowledge representations at the three levels: macroscopic, symbolic and submicroscopic (Derman and Ebenezer, 2018). However, to enable students to move across these three levels, it is crucial that the teaching strategies used facilitate the development of their cognitive structures. In the present study, the strategies applied during the teaching using MR (video, laboratory materials, photographs, kit models and chemical equations) assisted that development. In fact, data from a WAT show that students’ cognitive structures progressed from M1 to M2, with an increase in the number of response words and connections between words, and with a change in the nature of these connections. The results suggest that this development was influenced by the way students used MR, making use of the three MR representations proposed by Ainsworth (2006, 2008). The students recognized that the combination of the three MR functions allowed them to move across representation levels and to develop a deeper and more structured understanding of the concepts related to the saponification reaction. Thus, this research study has attempted to answer questions from previous research studies (e.g., Ebenezer, 2001; Derman and Eilks, 2016) by showing that a sequence of lessons using MR that incorporates the functions of complementary role, encouraging interpretations and constructing deeper understanding can support students’ learning and the building of well-structured cognitive networks. Future studies could attempt to better understand how MR functions can influence the development of the students’ cognitive structures in other content topics of chemistry.

Conflicts of interest

There are no conflicts to declare.

References

  1. Ainsworth S., (1999), The functions of multiple representations, Comput. Educ., 33, 131–152.
  2. Ainsworth S., (2006), DeFT: a conceptual framework for considering learning with multiple representations, Learn. Instruct., 16, 183–198.
  3. Ainsworth S., (2008), How do animations influence learning? in Robinson D. and Schraw G. (ed), Current Perspectives on Cognition, Learning, and Instruction: Recent Innovations in Educational Technology that Facilitate Student Learning, Charlotte, NC: Information Age Publishing.
  4. Ainsworth S., Prain V. and Tytler R., (2011), Drawing to learn in science, Science, 333(6046), 1096–1097.
  5. Ainsworth S., (2014), The multiple representation principle in multimedia learning, in Mayer R. E. (ed), Cambridge handbooks in psychology. The Cambridge handbook of multimedia learning, New York, NY: Cambridge University Press.
  6. Al-Balushi S. M. and Al-Hajri S. H., (2014), Associating animations with concrete models to enhance students’ comprehension of different visual representations in organic chemistry, Chem. Educ. Res. Pract., 15, 47–58.
  7. Al-Balushi S. M., Al-Musawi A. S., Ambusaidi A. K. and Al-Hajri F. H., (2016), The effectiveness of interacting with scientific animations in chemistry using mobile devices on grade 12 students’ spatial ability and scientific reasoning skills, J. Sci. Educ. Technol., 26, 70–81.
  8. Ardark D. and Akaygun S., (2004), Effectiveness of multimedia-based instruction that emphasizes molecular representations on students’ understanding of chemical change, J. Res. Sci. Teach., 41, 317–337.
  9. Bahar M., Johnstone A. H. and Sutcliffe R. G., (1999), Investigation of students’ cognitive structure in elementary genetics through word association tests, J. Biol. Educ., 33, 134–141.
  10. Blake A. J. D. and Nordland F. H., (1978), Science instruction and cognitive growth in college students, J. Res. Sci. Teach., 15, 413–419.
  11. Cook M., Wiebe E. N. and Carter G., (2008), The influence of prior knowledge on viewing and interpreting graphics with macroscopic and molecular representations, Sci. Educ., 92, 848–867.
  12. Creswell J., (2002), Educational research: planning, conducting, and evaluating quantitative and qualitative research, Upper Saddle River, NJ: Merrill Prentice Hall.
  13. Derman A. and Ebenezer J., (2018), The Effect of multiple representations of physical and chemical changes on the development of primary pre-service teachers cognitive structures, Res. Sci. Educ.,  DOI:10.1007/s11165-018-9744-5.
  14. Derman A. and Eilks I., (2016), Using a word association test for the assessment of high school students’ cognitive structures on dissolution, Chem. Educ. Res. Pract., 17, 902–913.
  15. Ebenezer J. V., (2001), A hypermedia environment to explore and negotiate students’ conceptions: animation of the solution process of table salt, J. Sci. Educ. Technol., 10, 73–91.
  16. Gilbert J. K., (2005), Visualization: a metacognitive skill in science and science education, in Gibert J. K. (ed), Visualization in science education, Dordrecht: Springer.
  17. Gilbert J. K., (2008), Visualization: an emergent field of practice and enquiry in science education, in Gilbert J. K., Reiner M. and Nakhleh M. (ed), Visualization: theory and practice in science education, Dordrecht: Springer.
  18. Gilbert J. K. and Treagust D. F., (2009), Chemical education: linking the representational levels of chemistry, Dordrecht: Springer.
  19. Holbrook J., (2008), Introduction to the special issue of science education international devoted to PARSEL, Sci. Educ. Int., 19, 257–266.
  20. Hovardas T. and Korfiatis K. J., (2006), Word associations as a tool for assessing conceptual change in science education, Learn. Instruct., 16, 416–432.
  21. Johnson P. E., (1967), Some psychological aspects of subject matter structure, J. Educ. Psychol., 58, 75–83.
  22. Johnson P. E., (1969), On the communication of concepts in science, J. Educ. Psychol., 60, 32–40.
  23. Johnstone A. H., (1982), Macro- and microchemistry, School Science Review, 64, 377–379.
  24. Liu X. and Ebenezer J., (2002), Descriptive Categories and Structural Characteristics of Students’ Conceptions: An Exploration of the Relationship, Res. Sci. Technol. Educ., 20, 111–132.
  25. Mabrouk S. T., (2005), Making Usable Quality Opaque or Transparent Soap, J. Chem. Educ., 82, 1534–1537.
  26. Miles M. B. and Huberman A. M., (1994), Qualitative Data Analysis: An Expanded Sourcebook, Thousand Oaks, CA: Sage Publications.
  27. Nakiboglu C., (2008), Using word associations for assessing non major science students’ knowledge structure before and after general chemistry instruction: the case of atomic structure, Chem. Educ. Res. Pract., 9, 309–322.
  28. Nakigoblu C., (2016), Present in part at International Conference in Mathematics, Science & Technology (ICEMST), Probing High School Students’ Cognitive Structure about Chemical Changes Through Word Association Test, Bodrum, Turkey, May, 2016.
  29. Nakiboglu C., (2017), Examination 8th grade students’ cognitive structures about physical and chemical changes through word association test, Eurasia Proc. Educ. Soc. Sci., 7, 49–51.
  30. Patton M., (1990), Qualitative Evaluation and Research Methods, Thousand Oaks, CA: Sage Publications.
  31. Piaget J., (1978), Success and understanding, Cambridge: Harvard University Press.
  32. Pozzer L. L. and Roth W., (2003), Prevalence, function and structure of photographs in high school biology textbooks. J. Res. Sci. Teach., 40, 1089–1114.
  33. Prain V. and Tytler R., (2012), Learning through constructing representations in science: a framework of representational construction affordances, Int. J. Sci. Educ., 34, 2751–2773.
  34. Prain V., Tytler R. and Peterson S., (2009), Multiple representation in learning about evaporation, Int. J. Sci. Educ., 31, 787–808.
  35. Rau M. A., (2015), Enhancing undergraduate chemistry learning by helping students make connections among multiple graphical representations, Chem. Educ. Res. Pract., 16, 654–669.
  36. Şendur G., Özbayrak Ö. and Uyulgan M. A., (2011), A study of determination of pre-service chemistry teachers’ understanding about acids and bases, Proc. Comput. Sci., 3, 52–56.
  37. Shavelson R. J., (1974), Methods for examining representations of a subject-matter structure in a student's memory, J. Res. Sci. Teach., 11, 231–249.
  38. Taber K. S., (2001), Building the structural concepts of chemistry: some considerations from educational research, Chem. Educ.: Res. Pract. Eur., 2, 123–158.
  39. Taber K. S. and García-Franco A., (2010), Learning Processes in Chemistry: Drawing Upon Cognitive Resources to Learn About the Particulate Structure of Matter, J. Learn. Sci, 19, 99–142.
  40. Talanquer V., (2011), Macro, Submicro, and Symbolic: The many faces of the chemistry “triplet”, Int. J. Sci. Educ., 33, 179–195.
  41. Tsai C.-C. and Huang C. M., (2002), Exploring students’ cognitive structures in learning science: a review of relevant methods, J. Biol. Educ., 36, 163–169.
  42. Tsui C.-Y. and Treagust D. F., (2013), Introduction to multiple representations: their importance in Biology and Biological Education, in Treagust D. and Tsui C.-Y. (ed.), Multiple representations in biological education, Dordrecht, Springer, pp. 3–18.
  43. Vygotsky L. S., (1978), Mind in society: the development of higher psychological processes, Cambridge, MA: Harvard University Press.
  44. Waldrip B. and Prain V., (2012), Learning From and Through Representations in Science, in Fraser B. J., Tobin K. G. and McRobbie C. J (ed), Second international handbook of science education, Dordrecht, The Netherlands: Springer Science + Business Media BV.
  45. Won M., Yoon H. and Treagust D. F., (2014), Students’ Learning Strategies With Multiple Representations: Explanations of the Human Breathing Mechanism, Sci. Educ., 98, 840–866.
  46. Wu H.-K., Krajcik J. S. and Soloway E., (2001), Promoting understanding of chemical representations: students’ use of a visualization tool in the classroom, J. Res. Sci. Teach., 38, 821–842.
  47. Yildirir H. E. and Demirkol H., (2018), Revealing students’ cognitive structure about physical and chemical change: use of a word association test, Eur. J. Educ. Stud., 4, 134–154.
This journal is © The Royal Society of Chemistry 2019