An investigation of learners' understanding of electrostatics illustrations in school textbooks

(1)

i

Acknowledgements

• God for granting me this opportunity to complete this study. Without Him, it would not have been possible.

• My husband, Henrico, thank you for always offering support, love and encouragement. You truly are my mainstay.

• Dr. Miriam Lemmer, thank you for being my supervisor during this study. All your effort and guidance is greatly appreciated.

• Mev. Hanli du Plooy, thank you for being my co-supervisor.

• Mev. Breytenbach, thank you for the statistical analysis of my questionnaire results. • My parents and parents-in-law. Thank you for all the support and love.

• To my friends and family, thank you for all the support and always offering a shoulder. Estien Slabbert, thank you for always being there and for all the support. • My boss and supervisor at work (Mr Pitts and Mrs Garrod), for the support offered. • The North West Education Department and the particular school for allowing me

to conduct this study.

• Thank you to all the participants and the parents of participants for allowing me to conduct the questionnaire and interviews.

(2)

ii

Abstract

The main research question to be answered in this study was: How do Grade 10

learners understand electrostatic illustrations from school textbooks?

Illustrations are used in school textbooks to communicate abstract or difficult information; this is also the case with Physical Science textbooks. Often these illustrations are simplified to try and convey only the most important aspects of what is intended. This, unfortunately, sometimes leave room for misconceptions as the illustration is then simplified to an extent where it might not be scientifically accurate.

The three types of illustrations that pertain to this study were diagrams, photographs and scientific models.

In order for learners to learn effectively using illustrations there need to be sufficient understanding of the topic at hand as well as the conventions used in the type of illustration. Prior knowledge, therefore, play an important role in the helpfulness of illustrations.

A cohort of 70 Grade 10 Physical science learners from a South African high school (in the North-West Province) took part in the pragmatic sequential mixed method study. A questionnaire was developed in order to probe the prior knowledge (including misconceptions) of the participants after which an interview was conducted with nine of the learners. Both the questionnaire results and the interviews were processed in order to answer the research questions. In the questionnaire consistency in learners’ responses were determined by using the Cronbach-Alpha method and correlations between questions were draw with the aid of Cohen’s effect sizes. The interviews were analysed by coding and mind-maps.

The results concluded that learners do struggle to understand electrostatics illustrations from school textbooks. Also when learners with a higher pre-knowledge answer questions they depend less on illustrations and more on their pre-knowledge. Learners with less pre-knowledge depend more on the different features of the illustration than the intended meaning.

Key words: electrostatics, illustrations, prior knowledge, teaching, learning,

(3)

iii

Uittreksel

Die hoof navorsingsvraag wat beantwoord moes word in hierdie studie was: Hoe

verstaan Graad 10 leerders elektrostatika illustrasies vanuit skoolhandboeke?

Illustrasies word gereeld in skoolhandoeke gebruik om abstrakte of moeilike inligting deur te gee. Dit is ook die geval in Fisiesewetenskappe handboeke. Hierdie illustrasies word vereenvoudig om slegs die mees belangrike aspekte van die onderwerp oor te dra, wat ongelukkig kan lei tot miskonsepsies omdat illustrasies soms vereenvoudig word tot op die punt waar dit nie meer wetenskaplik korrek is nie.

Die drie tipes illustrasies wat op hierdie studie van toepassing was, was diagramme, foto’s en wetenskaplike modelle.

Vir leerders om optimaal te kan leer met behulp van illustrasies het hulle die nodige kennis van die onderwerp sowel as die verskillende aspekte van die tipe illustrasies nodig. Voorkennis speel dus ook ‘n baie belangrike rol in of illustrasies bydra tot die verstaan van ‘n onderwerp.

‘n Groep van 70 Graad 10 Fisiese Wetenskappe leerders van ‘n Suid-Afrikaanse hoërskool (in die Noordwes Provinsie) het deelgeneem aan die pragmatiese opeenvolgende gemengde metode studie. ‘n Vraelys is opgestel om leerders se voorkennis (insluitend miskonsepsies) te bepaal. Onderhoude is daarna gevoer met nege van die leerders. Beide die vraelys en onderhoude is verwerk om die navorsingsvrae te beantwoord. In die vraelys is konsistensie in leerders se antwoorde bepaal met behulp van die Cronbach-Alpha metode en verwantskappe tussen sekere vrae getrek deur van Cohen se effek groottes gebruik te maak. Die onderhoude is verwerk en analiseer deur middel van kodering en breinkaarte.

Die resultate het bevestig dat leerders sukkel om elektrostatika illustrasies in handboeke te verstaan. Daar is ook bevind dat leerders met meer voorkennis minder afhanklik is van illustrasies as ander maar dat leerders met minder voorkennis meer op die verskillende eiensakppe van illustrasies fokus as op die betekenis daarvan.

Sleutelwoorde: elektrostatika, illustrasies, voorkennis, onderrig, leer,

(4)

iv

Language editor’s declaration

(5)

v

List of figures

Figure 1 The different disciplinary ways of knowing and modes of learning

Figure 2 Relationship of a representation, the referent and its meaning

Figure 3 Diagram of the semiotic representational structure

Figure 4 Interconnection between philosophical worldview, strategies of inquiry and research methods

Figure 5 Mind-map on the interview question pertaining to concept 1 – “a neutral object has no charge”

Figure 6 Mind-map on the interview question pertaining to concept 2 – “a charged objects either has electrons or protons”

Figure 7 Mind-map on the interview question pertaining to concept 3 – “friction causes charges”

(9)

ix

List of tables

Table 1 Questions that pertain to misconceptions in this study

Table 2 The different types of illustrations used in the questionnaire and where they were adapted from

Table 3 Constructs of questions for statistical analysis

Table 4 Learners’ responses to the multiple choice items (Section A) of the questionnaire

Table 5 The results obtained by the longer questions (Section B) in the questionnaire

Table 6 Results obtained in Section C of the questionnaire

Table 7 Constructs of questions for statistical analysis with the average and the Cronbach-Alpha value

Table 8 Results to the interview questions about familiar scenarios and illustrations

Table 9 Codes that were given in the interviews

Table 10 Results of the interview questions about familiar scenarios and illustrations

Table 11 Results of the interviews pertaining to selected multiple choice items

Table 12 Interview results on Question 2 and 3

Table 13 Illustrations that learners found easy to understand and illustrations that learners found confusing

Table 14 Comparison between scores for pre-knowledge questions and illustration questions

Table 15 The effect that certain aspects of illustrations have on the understanding of those illustration

(10)

1

Chapter 1 Motivation, problem statement and research questions

1.1 Introduction

According to Bungum (2008) illustrations are a way of communication in textbooks. Physics textbooks use illustrations to enhance students’ conceptual understanding, especially of abstract concepts (Testa, Leccia & Puddu, 2014), such as electrostatics concepts. In this document a broad meaning to the term illustration is given that is perceived to encompass the variety of images used in textbooks to aid in description and explanation of scientific concepts. Examples are: photos, drawings and models. Illustrations are visual resources in students’ development of mental images through a process called visualisation (Dictionary.com, 2014); this implies that for a learner to use an illustration they first have to construct a mental representation of the topic at hand (Testa et al., 2014). According to Hinze, Rapp, Williamson, Shultz, Deslongchamps and Williamson (2013) people use familiar features in visualisations to make sense of the abstract information it intends to portray. This study pertains to the textbook use of illustrations and how it affects the conceptual understanding of learners.

In Section 1.2 the motivation behind the study is given with the problem statement. The aim and objectives are given in Section 1.3 and the research questions in Section 1.4. Section 1.5 gives the literature review whereas Section 1.6 describes the research design that was followed. Potential implications are described in Section 1.7 and Section 1.8 gives a summary of the chapter.

(11)

2

1.2 Motivation and problem statement

Textbook illustrations present scientific knowledge and ideas or concepts and relations according to specific principles, conventions and symbols. Students should, thus, develop an understanding of illustrations as part of the language (or discourse) of physics. Airey and Linder (2009:28) define disciplinary discourse as “complex of representations, tools and activities of a discipline”. Their representation of how meaningful learning takes place in Physics is illustrated in Figure 1. In this figure it is seen that disciplinary discourse is composed of tools, representations and activities that are made up of different modes. These modes consist of, among others, illustrations, gestures and spoken language. This study mainly emphasises the importance of illustrations as a mode of learning.

Figure 1. The different disciplinary ways of knowing and modes of learning (Airey & Linder, 2009:29).

(12)

3 If suitable illustrations (visual aids) are incorporated with text it will assist in more effective communication of what the topic means to portray (Vinisha & Ramadas; 2013). Prangsma, van Boxtel, Kanselaar and Kirschner (2009) also feel that visualisation aids learning in comparison to written text. Sometimes, however, textbook illustrations hinder learning of scientific content instead of enhancing understanding (Testa et al., 2014). According to Bungum (2008) the use of textbook illustrations is common practice, though these illustrations are “under-communicated” versions of the content and are often not specialised, therefore, they may not convey the same message as intended. If these illustrations are not used in the correct context it can also lead to misconceptions. Du Plooy (2012:3) argued that learners develop certain misconceptions because they cannot visualise specific concepts, like atoms, and therefore make use of the illustrations in textbooks as a reference point, taking them up literally.

In order to successfully read and interpret a scientific illustration, students should both have prior content knowledge underlying the illustration, as well as knowledge about its structure and conventions (Cook, Wiebe & Carter, 2008). Kohl and Finkelstein (2006:1) investigated the relationship between how students performed in problem-solving and how problem representations related to their competency in this matter and found that student performance depended much on the combination between prior knowledge and representation.

Chang (2011: 226) finds electrostatics difficult to teach because most of the key terms and concepts are not words or aspects that learners can relate to from everyday life. He identified four main difficulties when it comes to learning electrostatics: The first is that the concepts are abstract. The second is that specific concepts, that seem alike, can be mixed up by the learners (i.e. concept confusion can occur). Thirdly, students often overlook electric fields. The last difficulty is that learners struggle to apply some of the laws.

In this section the importance of illustrations as mode of learning in disciplinary discourse has been emphasized. In order to enhance meaningful learning of electrostatics, the problem of the research study reported in this dissertation is to investigate how learners interpret the abstract information conveyed in electrostatics textbook illustrations and what aspects they struggle with. The representational

(13)

4 features of the illustrations as well as the required prior knowledge and learners’ already existing concepts were taken into account.

The research questions and the aim and objectives of the study to investigate the research problem are given in the following two sections (1.3 and 1.4). For terminology used in this study, refer to Appendix 1.

1.3 Research questions

The main research question to be answered in the empirical study is:

To what extent do illustrations in Physical Science school textbooks promote understanding of electrostatics in Grade 10?

The main research question is addressed by answering the following secondary research questions:

1. What prior knowledge, including misconceptions, do learners have, related to textbook illustrations on electrostatics?

2. How does the prior knowledge of learners affect their understanding of textbook illustrations?

3. What is the effect of certain representational features, such as compositional structure and real/symbolic elements of electrostatics illustrations, on learners’ understanding?

(14)

5

1.4 Aim and objectives

The aim is to investigate how Grade 10 learners from a South African high school (in the North-West Province) understand illustrations in electrostatics in school textbooks.

The objectives of the research are the following:

• to investigate learners’ prior knowledge and misconceptions in electrostatics related to textbook illustrations;

• to analyse the effect that prior knowledge have on learners’ understanding of illustrations; and

• to determine the effect of certain representational features, such as compositional structure and real/symbolic elements of electrostatics illustrations on learners’ understanding.

The research questions and objectives are addressed in the theoretical framework discussed in the next section.

1.5 Theoretical framework

Constructivism is a theory on how a person learns from observations and experiences (Anon, 2004). It has the conjecture that people learn through experiences, which are constantly changing (McWilliams, 2016). The information that an individual process daily forms a working model that includes pre-knowledge (Paas, Renkl & Sweller, 2003) that takes up a large part of man’s cognition. Visuals are often made to be attractive as a sort of motivation (Vanisha & Ramadas, 2013) and can be used to simplify difficult information as well as make it easier to comprehend and to remember (Mandl & Levin, 1989: vii).

Bruner and Olson (cited by Molitor 1989:3) states that both text and pictures can be seen as media, because they try to visualise the reality that cannot be accessed through experiences. Selecting words and images, organising words and images and integrating the words and images are cognitive processes that are necessary for meaningful learning to occur (Mayer, Bove, Mars & Tapangco, 1996:64). Learners’

(15)

6 understanding of illustrations and its connections to words, such as used in textbooks, are therefore important to study for meaningful learning.

Semiotics provides a deeper understanding of how learners make meaning of illustrations and their affordances in terms of understanding a specific concept (Alant & Sherwood, 2017). The different types of illustrations used in this study are photographs, diagrams and scientific models. A photograph is a simple illustration that does not give any perspective of deeper meaning or insight in relationships (Uttal & O’ Doherty, 2008:57); diagrams show relationships between certain aspects of a specific phenomenon (Cromley, Weisberg, Dai, Newcombe, Schunn, Massey & Merlino, 2016) and scientific models describe a phenomenon through its important characteristics (Dictionary.com, 2016) but may not be completely connected to the real world (Hart, 2008:530).

There are thus three theoretical frames relevant to learners’ understanding of illustrations, namely constructivism, cognition and semiotics. Constructivism deals with characteristics of learners’ pre-knowledge, e.g. whether scientifically acceptable concepts or misconceptions (Driver, Asoko, Leach, Scott & Mortimer, 1994). Cognitive processes, such as attention to specific features, retrieval of existing knowledge and meaning-making of new information, affect learners’ understanding of representations such as illustrations (Sternberg & Sternberg, 2012). Semiotics provide a framework for studying learners’ interpretation of illustrations as medium of communication (Alant & Sherwood, 2017).

1.6 Research design

A research paradigm is a group of assumptions about important parts of reality that gives rise to the specific world-view that is adopted and can therefore be seen as a lens through which reality is perceived (Nieuwenhuis in Maree, 2010). In this research project a pragmatic approach is used, in which the researcher can chose the methods and procedures to follow (Creswell, 2009). A pragmatic approach is relevant when the research question is neither positivistic nor interpretive (Ihuah & Eaton, 2013). A sequential mixed method study is used in this research because one method is used to elaborate on another (Creswell, 2009).

(16)

7 In the next section the methods used in this study are discussed (Section 1.6.1) followed by a description of the participants that took part in the study (Section 1.6.2) and the data collection that took place (Section 1.6.3). Section 1.6.4 shows how the data is analysed whereas Section 1.6.5 describes the ethical aspects that pertain to this study.

1.6.1 Research methodology

A mixed methods research design is followed in this study to evaluate if a group of 70 grade 10 learners reveal the required prior knowledge or have misconceptions and to determine the effect of compositional structure and real/symbolic elements when they look at electrostatics textbook illustrations.

A mixed methods design is the most appropriate for this study, since it gives the best combination of different methods (Creswell, 2009). A questionnaire (quantitative) and interviews (qualitative) are used because the answers given by participants are subjective and there may be a wide perspective regarding the distribution of answers from the population. This provides the researcher with the necessary information to see what prior knowledge of electrostatics and of representations the learners in question possess.

The questionnaire consists of a series of questions (refer to Appendix 2). Firstly, there are 15 multiple choice questions that mainly test prior knowledge (including misconceptions) in electrostatics; these are followed by several questions showing different forms of representation and asking learners to interpret illustrations about specific concepts in electrostatics.

The three misconceptions that pertain to this study are:

1. “a neutral object has no charge” (Caillot and Xuan, cited by Baser and Geban, 2007:244);

2. “a charged body contains only either electrons or protons” (Siegel and Lee, 2001); and

3. “friction is the cause of static electricity” (Caillot and Xuan, cited by Baser and Geban, 2007:244 and Siegel and Lee, 2001)

(17)

8 The correct conceptions that the above mentioned misconceptions are based on are:

1. A neutral object consists of the same amount of positive and negative charges (De Beer, Gibbon, Jones, Kunene, Patrick, Sampson, Subramani, Visser & Whitlock, 2013)

2. A charged object contains both electrons and protons, the one is just in excess (Williams, 2012; De Beer et al., 2013)

3. Contact between electrically charged objects cause electrons to be transferred, which causes static electricity (Williams, 2012)

1.6.2 Participants

The participants were a non-random group of 70 grade 10 Physical Science learners from a school in the North-West Province, South Africa. All enrolled learners were asked to participate in a voluntary study on electrostatics.

1.6.3 Data collection

The data of this study was collected in two ways; the quantitative data was collected through a questionnaire and the qualitative through interviews. The questionnaire that was developed contained pre-knowledge as well as illustrative questions on electrostatics. Since electrostatics is not discussed in great detail in the grade 9 curriculum the grade 8 work forms the basis for grade 10 electrostatics. The questionnaire was controlled by a statistical analyst of the NWU statistical consultation services and was approved. Textbook illustrations having different compositions and real or symbolic features were included in the questionnaire. The correct ethical procedure was followed and all the requirements were met before the onset of the empirical study. After assembling participants from the grade 10 group of a secondary school (all the grade 10 Physical Sciences learners were asked to participate, participation was voluntary) in the North-West Province; learners submitted their answers to the questionnaire that was then analysed, after which an interview was conducted with nine of the learners. The learners who were interviewed were chosen purposively (three who did well in the questionnaire, three with moderate results and three who showed a lot of misconceptions).

(18)

9 1.6.3.1 Quantitative data collection and analysis

The questionnaire was set up using a wide variety of questions; these questions were found in various different sources, for example current South African textbooks, exam papers and relevant internet resources. Some of the questions were self-compiled. The different types of questions that were used are multiple choice items, as well as fill in questions, which included baseline questions (prior knowledge) as well as questions to test for the misconceptions and understanding of illustrations. The different questions (and illustrations) tested for different real or symbolic features. The illustrations also included different conventions in order to see the effect of those conventions on the understanding of learners. The questionnaire content was validated by two research experts in the field of Physics education and was checked by a statistical consultant to ensure the validity and reliability of the study. A pilot study was conducted in order to see if what was meant to be tested was tested.

The results of the questionnaire was statistically processed. Frequencies of responses were obtained and Cronbach alpha values calculated for constructs of questions. Also Cohen’s effect sizes were calculated to compare Construct A (basic and prior knowledge) to the other constructs in order to see the significance of the difference between those constructs.

1.6.3.2 Qualitative data collection and analysis:

Semi-structured interviews were used to complement and probe deeper into the quantitative results. The participant for the interviews were chosen on account of their performance in the questionnaire. Three learners were selected that performed well, three average and three poor. These learners were asked to explain their reasoning to selected questions of the questionnaire. The learners were also asked about the familiarity of the illustrations in the questionnaire and whether it was understandable. The interviews were transcribed, the data coded and analysed. In the coding certain words were connected to specific concepts that pertained to the study, for example “correct concept”, “misconception” or “confusion” (see Section 5.4). The two research leaders verified the coding and analysis of the data to ensure trustworthiness.

(19)

10

1.6.4 Data analysis

The answers to the questionnaire were firstly analysed per item and frequency tables were compiled for scientifically correct, incorrect and no responses. It was also determined whether the learners possessed misconceptions of electrostatics concepts. Secondly, the questions containing content knowledge and features of textbook sketches pertaining to selected concepts and electrostatics processes were grouped into constructs and the learners’ responses analysed statistically. The statistical analysis took place in the form of Cronbach-Alpha and Cohen’s effect sizes. In the Cronbach-Alpha reliability of variables was determined according to a scale (Santos, 1999) and in Cohen’s effect sizes were used to compare information for dependent groups (Ellis & Steyn, 2003).

The qualitative part was analysed using a coding system (Section 5.4) to gather information from the interviews. It was then used to set up mind-maps in order to better interpret the effect of the results. Learners’ familiarity with the sketches used in the questionnaire as well as which they regarded easier were also determined.

1.6.5 Ethical aspects

Ethics on the disciplines of study are standards of conduct and can also be defined as a method, procedure, or perspective for deciding how to act and for analysing complex problems and issues (Resnik, 2010). Some of the general ethics are honesty, objectivity, integrity, carefulness, openness, and respect for intellectual property, confidentiality and non-discrimination.

In order to conduct the investigation the researcher needed to get permission from the headmaster of the school the participants attended and the Department of Education. In order to give a questionnaire or interview to learners younger than eighteen consent had to be obtained from their parents. Taking part in completing the questionnaire was completely voluntary and did not discriminate against or affect learners in any way. Only one person was responsible for marking the questionnaires in order for all the learners to be evaluated with consistency.

After the participants completed the questionnaire nine of them were chosen to take part in an interview (voluntary). Learners’ names as well as the school’s name were not mentioned at any point during the investigation. Learners were also free to

(20)

11 terminate their participation without discrimination at any point during the quantitative or qualitative data collection.

1.7 Potential implications of findings

The potential implications of the findings can lead to improvements in illustrations in Physical Sciences in general and specifically in school textbooks. It can eventually lead to better understanding of Physical Science concepts that are difficult to grasp due to the abstract nature of those concepts.

1.8 Chapter Summary

This chapter focuses on the reason for the study and gave an overview of the methods followed. The research question of this study is: “To what extent do illustrations in Natural Science and Physical Science school textbooks promote learning and understanding of electrostatics in Grade 10?”

Chapter two gives a discussion on knowledge and how meaningful learning takes place. It also focuses on prior knowledge and how it affects learning using illustrations. The next chapter (chapter three) gives a detailed explanation of the different types of illustrations found in Physical Science that are meant to portray certain types of information. It also discusses what the importance of those illustrations are in teaching-learning. The empirical study is based on three theories related to learning, namely constructivist theory, semiotics and cognition. Chapter four elaborates on the research design and methodology whereas chapter five gives the detailed results obtained. The last chapter gives feedback on conclusions drawn from the study and explains possible implications of the study.

(21)

12

Chapter 2 Knowledge and the learning proses

2.1 Introduction

In this chapter the focus is on knowledge and how the learning process can be affected by prior knowledge, specifically misconceptions. Section 2.2 addresses cognition and how it plays a role in meaning making in everyday life and in the classroom learning process. The next section (2.3) is about constructivism and prior knowledge. It helps readers understand why certain learners grasp specific concepts quicker than others. The characteristics of prior knowledge are discussed in section 2.4 and the next section (2.5) focuses on what a misconception is. Section 2.6 pertains to the learning of electrostatics and explains what prior knowledge provides a background for learning electrostatics. Next the effect that textbook illustrations can have on learning and how it can effect learning difficulties is discussed in section 2.7. In this section a more in depth approach is taken as to what other studies have found in this regard. The last section (section 2.8) is a conclusion of the findings of the literature study reviewed in this chapter.

2.2 Cognition and meaning making (sense making)

Learning requires certain cognitive processes (Mayer et al., 1996:64), for example to select words and images, to organise words and images and to integrate the words and images. Tversky (cited in Akaygun & Jones, 2014:783) found that both visualisations as well as written language are cognitive tools that can be used to understand how people understand and perceive a specific concept.

Throughout the learning process learners use cognitive processes to form perceptions and mental models, remember information and think about it (Sternberg & Sternberg, 2012; Wiley & Jee, 2010). According to Paas et al., (2003) human cognition has a small ability to process new information, but can process a huge amount of previously acquired knowledge, even if it is complicated. Only limited information that is focussed

(22)

13 on by the senses is forwarded to the working memory to be combined with existing information recalled from the long-term memory. A very large amount of long term memory is pre-knowledge associated with information that people need in everyday life.

For meaningful learning to take place, a learner must perform the required cognitive processes. Meaningful learning is “learning with understanding” (Michael, 2004:227). Other researchers confirm that understanding and meaning making bear close resemblance. Kegan (1982) describes meaning making as a way of understanding something in a way someone else experiences it. Paas et al., (2003) maintains that understanding is when a person has the ability to interpret aspects that work together in the working memory. Interesting enough, when searching the internet for a definition of “meaning making”, The Free Dictionary (2014) redirects the word to understanding and defines it as: “the ability by which one understands” or “Individual or specified judgement or outlook”. In an article about meaning making in a learning environment, Ingelzi (2000) describes meaning making in a broader way as the process in which people make sense of everything around them, for example experiences and knowledge. Meaning making is a philosophical concept that is very difficult to define and explain as a separate entity and is better explained in context, according to Pearlman (2014), who refers to meaning making as “reasoning”.

According to Blumer (cited by Enghag, Forsman, Linder, MacKinnon & Moons, 2013:627) there are three premises for meaning making. They are:

• Premise 1: Self-interactive, which uses prior experiences to give meaning to a certain situation.

• Premise 2: Social interaction, which uses communication to acquire meaning. • Premise 3: Interpretive processing, which derives meaning from previous

encounters.

Through these three premises learners learn, even difficult subjects such as Physical Science.

(23)

14

2.3 Constructivism and prior knowledge

2.3.1 Constructivist learning theory

While cognition is a study of how people acquire and use information (Sternberg & Sternberg, 2012), constructivism is a theory on how one learns from observations and experiences (Anon, 2004). According to McWilliams (2016) constructivism has the conjecture that people create knowledge through experience and that this makes it ever-changing because the context changes as experiences change.

There are two types of constructivism, social and psychological (McPhail, 2016). Social constructivism refers to how knowledge is constructed and the effect that it has on society, while psychological constructivism pertains to the essence of learning and how it takes place. One of the limitations of constructivism in education reported by McPhail (2016), is a lack of knowledge between different varieties of constructivism and the blurring of the lines between the epistemological, ontological and metaphysical as well as moral assumptions. This leads to educators incorporating their own moral views in teaching without realising it and not taking the views of the learner into consideration.

2.3.2 Prior knowledge in the learning process

Before classroom tuition, children already possess concepts and explanations for everyday life phenomena. Prior knowledge, also known as pre-knowledge, originates from previous experiences (AllWords.com, 2014). Because learners’ preconceptions are often different from scientific concepts, learning should not only be the accumulation of knowledge, but also the rearrangement of existing knowledge (Driver

et al., 1994; Vosniadou & Skopeliti, 2014). Teachers should possess the necessary

skills and knowledge to help connect learners’ pre-knowledge with the scientific knowledge in order for them to comprehend the new content based on existing content (Kikas; 2004). Greca and Moreira (1997) agrees that when information is presented to learners they frequently interpret the given information according to their previous mental models which might not be scientifically accurate. Misconceptions may consequently be formed.

(24)

15 Bakhtin (cited in Enghag et al., 2013:629) gives a substantial link between “representations” and meaning making that is used in disciplinary discourse in describing how people live their everyday lives through primary discourse and how new ideas and environments provide a secondary discourse. Accordingly, if learners have to explain a phenomenon, their different ideas coexist and the learners' minds contain a combination of the different conceptual ideas (Criado & Garcia-Carmona, 2011:772).

2.4 Characteristics of prior knowledge

2.4.1 Prior knowledge consists of a variety of knowledge structures

Prior knowledge is not a solid state as to what is known and what is unknown, and is therefore “fuzzy in nature” (Monk (cited in Criado & Garcia-Carmona, 2011:771). Mental knowledge structures consist of small fragmented elements, theory-like networks of ideas as well as misconceptions (Brown, 2014). A learner’s knowledge structures change continually and can therefore be considered dynamic.

2.4.2 Prior knowledge and misconceptions

Prior knowledge ideas that are inconsistent with scientifically correct knowledge, which happens quite often, are called misconceptions (Dega, Kriek & Mogese, 2012). Learners develop certain misconceptions, because they find it difficult to visualise specific concepts and therefore use illustrations as literal references which creates misconceptions (Du Plooy, 2012:3).

2.4.3 Prior knowledge aids learning with illustrations

Learning using multiple representations is more difficult for learners with less pre-knowledge (Cook et al., 2008). Prangsma et al. (2009) agreed that learners with less prior knowledge find MER’s more difficult to understand but they further find single representations as less confusing. Learners with higher abilities also benefit more from diagrams than learners with lower abilities (Booth & Koedinger, 2011) and learners with more knowledge rely more on diagrams than learners with less

(25)

pre-16 knowledge (Hinze et al., 2013). Cook et al. (2008) states that if a learner has a higher pre-knowledge of specific concepts, especially in chemistry, they have a better perception of the differences between models and the events that they represent.

Learners build their knowledge from prior knowledge as well as from information that is found in text as well as pictures (Molitor, Ballstaedt & Mandl, 1989:13). In order for learners to successfully read and interpret a scientific illustration, learners should both have prior content knowledge underlying the illustration as well as knowledge about its structure and conventions (Cook et al., 2008). A learner’s performance depends much on the combination between pre-knowledge and representation format (type of representation) (Kohl & Finkelstein, 2006:1; Pinto & Ametller, 2002 and Ametller & Pinto (2002)).

2.5 Misconceptions

Oxford Dictionaries (2014) defines misconceptions as “A view or opinion that is incorrect because based on faulty thinking or understanding” and is therefore a learning difficulty. The Free Dictionary (2014) defines misconceptions as “A mistaken thought, idea or notion, a misunderstanding”. A misconception (also known as an alternative conception) is a mistaken answer for a specific scenario (Baser & Geban, 2007). According to Du Plooy (2012) a possible cause for misconceptions is that learners fail to visualise basic concepts and also use models as literal forms of study and interpret textbook representations literally. Because of this, demonstrations and models must be accompanied by a previous understanding (Chang, 2011:226). Baser and Geban (2007:244) believe that the constructivist view of learning implies that new knowledge builds on old knowledge and therefore sets a barrier for new learning and a “foothold” for misconceptions.

(26)

17

2.6 Learning of electrostatics

2.6.1 Key ideas of electrostatics based on the atomic theory

Ahlers (2009) maintains that electrostatics forms the basics when it comes to understanding electricity. Together with atomic structure, it should play a central role in primary school science curricula. Static electricity entails the movement of electrons, a particle that occurs in atoms (Williams, 2012:316).

Dalton’s atomic theory contains, among others, the following key concepts underlying electrostatics phenomena: Elements are made up of small parts called atoms; atoms of a specific element is always the same; every elements’ atom looks different from the next; and atoms cannot be made or destroyed during chemical processes (Kelder, Van der Merwe & Holmes, 2011:56-63). Atoms are the basic building blocks of all matter and are made up of sub-atomic particles known as protons, neutrons and electrons (De Beer et al., 2013). Protons have a positive charge and neutrons carry no charge but these two sub-atomic particles have the same mass. Electrons have a very small mass and have a negative charge. The central part of the atom is called the nucleus and contains the protons and neutrons, whereas the electrons are found in energy levels surrounding the nucleus known as orbitals. Electrons can move between orbitals by absorbing or loosing energy and have great importance in all chemistry (Kelder et al., 2011:63). Electron transfer is the movement of electrons from one atom to the next (Kennepohl & Solomon, 2003).

When two objects touch and are then separated, it causes a charge known as contact or static charge (Williams, 2012:316). It results from a transfer of electrons between the objects in contact. An ion is a positive or a negative charged particle. Positive charged particles are called cations and have more protons than electrons, hence the positive charge. Negative charged particles are called anions and contain more electrons than protons.

According to Lewis terms, the formation of ionic bonds (discussed in chemistry at school level) occurs due to the transfer of electrons from one atom to the next (Atkins, 2014). When the transfer occurs, the valence electrons from the more electropositive element are removed from the atom and thus leave a positive ion. The electrons that

(27)

18 were removed from the atom now move into the empty orbitals of the valence shell of the more electronegative atom. Through this the outer orbital is filled.

Polarisation is when a charged object is brought close to a neutral object (Grayson, Harris, McKenzie & Schreuder, 2011). The charges inside the neutral object that is the same as that of the charged object, attempts to move further away and the opposite charges inside the neutral object attempts to move towards the charged object. This causes a shift of the electrons inside the object to create a more positive and more negative side.

2.6.2 Learning problems or misconceptions pertaining to electrostatics

Allen and Coole (2012) said that some learners will already have certain concepts about topics even before it is taught. According to Chinn and Malhotra (cited in Kikas (2004) a misconception often occurs when newly learned information, that is abstract, is integrated with previous information. This leads to a new interpretation of the knowledge to fit in with everyday life.

Chang (2011: 226) finds electrostatics difficult to teach because most of the key terms and concepts are not words or aspects that learners can relate to from everyday life. Also Chang states that there are four main difficulties when it comes to learning electrostatics. The first is that the concepts are more abstract, the second is that specific concepts that seem alike can be mixed up by the learners, thirdly learners often overlook electric fields and the last is that learners struggle to apply some of the laws. Greca and Moreira (1997) also maintain that concepts on electric fields are abstract and therefore many learners struggle to understand the meaning. Furthermore, resources, such as devices (Van der Graaff generator, plastic rods with cloths, compasses etc.) to enhance learners’ understanding by demonstrate electrostatics phenomena, can sometimes be untrustworthy and consequently it can affect learners’ learning of the subject (Ahlers, 2009).

(28)

19 Examples of misconceptions in electrostatics (that will also be the focus of this study) are:

• “a neutral object has no charge” according to Beaty (1998) and Caillot & Xuan (cited in Baser & Geban, 2007:244);

• “a charged body contains only either electrons or protons” (Siegel & Lee, 2001); • “friction is the cause of static electricity” according to Beaty (1998), Caillot &

Xuan (cited in Baser & Geban, 2007:244) and Siegel & Lee (2001).

The first of the three misconceptions addressed in this study is about the charge of a neutral object. Beaty (1998), Ahlers (2009) and de Beer et al. (2013) maintain that a neutral object is still matter and therefore made up of atoms. An atom is made up of protons, neutrons and electrons, the sub-atomic particles. Any atom is made up of charges because of the protons and the electrons.

The next misconception is about the subatomic particles found in a charged object. Once again an object is made out of atoms, which are made up of protons, neutrons and electrons (Beaty, 1998; Ahlers, 2009; De Beer et al., 2013). The difference is that the outer energy level of the atom loses or gains electrons and therefore becomes more positive or negative respectively; nothing happens to the protons and the neutrons in the center of the atoms and therefore the object still consists of protons, neutrons and electrons (Williams, 2012:316).

The last of the three misconceptions mentioned above states that static electricity is only caused by friction. Beaty (1998) maintains that all that is required for static electricity is touching and therefore friction is not necessary. He continues by offering an example that if hair is rubbed with a balloon it becomes charged. This is because the rubbing increases the contact area and however it is needed, it does not cause the unbalanced charges and therefore does not cause the “electrification”.

These misconceptions are common among learners and unfortunately enforced by some textbooks (Beaty, 1998).

(29)

20

2.7 Effect of textbook illustrations on learners’ learning difficulties

According to a study conducted by Testa et al. (2014) learners often come to wrong conclusions, and therefore develop misconceptions, due to models of certain phenomena that are inadequate. In their study, secondary school learners were given six illustrations from a textbook used in Italian secondary schools about certain astronomical phenomena. Interviews were also conducted. After interpretation of all the responses it was found that inadequate geometric models of the phenomena in question caused misconceptions and difficulties in understanding certain aspects. They also maintained that representations did not aid in understanding the concepts at hand.

A study by Berends and Van Lieshout (2008) determined the effect of illustrations on arithmetic problem-solving. They took 24 word problems that were illustrated using four types of illustrations with an increasing amount of information in the illustrations. Their results revealed that illustrations did not aid in solving word problems and is redundant in the specific context.

Contrary to Berends and Van Lieshout (2008), Booth and Koedinger (2011) found that diagrams did aid the learners who were older and had a higher ability for problem-solving, but that younger and less academically inclined learners did not benefit. This followed from a study of high school and college learners (373 learners) by taking algebra problems in a story format and adding diagrams in order to find out if diagrams were beneficial to the learners.

In a study on illustrations in Norwegian physics textbooks conducted by Bungum (2008) nine different Norwegian secondary school physics textbooks were used. The dimensions of the illustrations were analysed and then constructed into five different modes of illustrations. It was mainly a qualitative study. It was found that more recent textbooks had illustrations that were abstract and did not necessarily communicate the intended meaning.

Evagrou, Erduran and Mantyla (2015) had a more theoretical approach and their study about the role that visual representations play in scientific practices included mainly case studies. The conclusions drawn from the study suggested that there should be a shift from merely using illustrations in order to understand the content to “engaging

(30)

21 in the process of science”. This statement means that science should not be taught merely to acquire knowledge or to understand the acquired knowledge but also to engage in science and practicing science.

Vanisha and Ramadas (2013) concluded (in a study on visuals of the water cycle in science textbooks) that visuals are often made to be attractive (and may not be purely scientific). This can play a motivational role and if the text is complicated an illustration may aid learners in comprehending the information in question.

“Images caused students to misunderstand” was one of the findings from a study conducted by Colin, Chauvet & Viennot (2002). This particular study was on how learners read images in optics.

2.8 Conclusion

The information that is processed daily forms a working model that includes pre-knowledge (Paas et al., 2003). This pre-pre-knowledge takes up a large part of people’s cognition. When a person learns with understanding it is called meaning making and this process leads to the acquisition of knowledge that can confirm (or contradict) pre-knowledge.

This study focuses on misconceptions of certain elements in electrostatics and how illustrations can influence the learning process and aid (or hinder) the correcting of those conceptions. Also some of the tools for learning different types of illustrations and how learners interpret it are discussed and a relation is drawn as to the intended meaning versus what is understood.

Illustrations can be very helpful in the teaching-learning environment, but can also lead to misconceptions. It was found throughout this literature study that a learner with a higher pre-knowledge seems to learn more efficiently using illustrations, than a learner with a lower pre-knowledge. The authors of textbooks usually need to try and explain the work and therefore sometimes deviate from the “purely scientific” explanation in an attempt to make it easier to understand. Also because of that reason, textbook illustrations and models may deviate from formal science in many instances.

(31)

22 In the next chapter the theory on illustrations and models are given. This pertains to the types of illustrations and the use of those illustration in teaching and learning.

(32)

23

Chapter 3 Theory on illustrations and models

3.1 Introduction

Disciplinary discourse was defined in chapter one as: “complex of representations, tools and activities of a discipline” (Airey & Linder, 2009:28). Figure 1 in Chapter one describes the “tools” that disciplinary discourse is composed of. Among the modes of learning were images, which is the focus of this study. Learners’ interpretation of textbook illustrations is investigated in a semiotic framework, as proposed by Testa et

al. (2014).

In this chapter information is gathered from literature on different types of representations and their specific uses in the context of teaching-learning. Section 3.2 gives a description of the framework of this study (semiotics) whereas section 3.3 discusses the use of different illustrations. The next section (section 3.4) indicates the type of illustrations applicable to this study. Section 3.5 focuses on how illustrations impact model-forming in learners. An explanation on multiple representations is given as it pertains to this study in section 3.6 and section 3.7 focuses on the different aspects of illustrations. The last part of chapter three (section 3.8) explains the importance of illustrations and models in teaching-learning and textbooks.

For the purpose of this study, images, models and illustrations will be referred to not only as tools of instruction of science but also as ways of passing on knowledge and ideas in science education. In the literature study reported here, illustrations and images are perceived to convey the same meaning and are thus used alternatively.

(33)

24

3.2 Semiotics framework

The reading of images is a semiotic process (Lopez & Pinto, 2017). Semiotics as a study of meaning making of representations (also called signs) contains three inter-related aspects, as shown in Figure 2 (Jamani, 2011). A representation such as a word or illustration depicts an object, concept or idea as referent of which the meaning is interpreted.

Figure 2. Relationship of a representation, the referent and its meaning (Adjusted from Jamani, 2011).

Testa et al. (2014) proposed a semiotic framework according to which a visual representation can be classified as narrative or conceptual (Figure 3). Narrative tells a specific story of a scenario taking place. This scenario can either be naturalistic or abstract. If the representation is naturalistic it needs no further explanation, whereas an abstract narrative illustration tells a story of a process taking place of which the reason might be difficult to see with the naked eye. A conceptual representation can also be naturalistic or abstract. Naturalistic in this instance is more realistic whereas abstract includes diagrams that explain a phenomenon. Conceptual representations can further be classified as classificatory, analytical or symbolic. Classificatory gives the relationships between certain elements of a phenomenon, analytical gives the relationship of a system as a whole and symbolic gives relationships of different aspects with meanings and elaborations on the deeper impact that it portrays. According to Stylianidou (2002) these relationships (of conceptual representations) are more fixed.

Referent/Signified (e.g. object, concept, idea)

or illustration)

Meaning/Signification (e.g. interpretation of the

meaning)) Representation/Sign

(34)

25 Figure 3. Diagram of the semiotic representational structure (adapted from Testa et

al., 2014)

Semiotics provides a deeper understanding of how learners make meaning of illustrations and their affordances in terms of understanding a specific concept (Alant & Sherwood; 2017). It focuses more on what learners can do and less on what is illusive to them and is therefore constructive to learning and the improvement of learning. According to Ametller and Pintó (2002) this framework explains a language that is an important tool for conveying an educational message in teaching science. Waldrip, Prain and Carolan (2010) maintains that learning in science is a process by which learners have to understand and learn to understand how to construct correct scientific meaning from “representational conventions” in the subject.

Visual representations Narrative Conceptual Naturalistic Abstract Classificatory Analytical A Symbolic Naturalistic Abstract

(35)

26

3.3 Use of illustrations as representations

A representation is any ensemble of images, words or characters that is used to portray a certain meaning (Elia, Gagatsis & Demetriou, 2007:659). Schnotz (cited by Elia et al., 2007:659) states that there are two forms of representation. The first is descriptive and the other depictive. Descriptive gives the basic structure and use of what is meant to be shown whereas depictive gives in more detail aspects of the abstract message it tries to convey.

Illustrations are communication media and are important in the communication of physics, especially in textbooks (Bungum, 2008). Bruner and Olson (cited by Molitor 1989:3) argued that both text and pictures can be seen as media, because they try to visualise the reality that cannot be accessed through experiences. Illustrations can thus be used to make abstract concepts more understandable by linking the familiar with the unfamiliar (Hinze et al., 2013).

Illustrations can be used to simplify difficult information and they are also supposed to make it easier to comprehend and to remember that information (Mandl & Levin, 1989: vii). Evagorou et al. (2015) agrees that the use of visual representations can make the interaction with abstract information and scenarios easier.

“Different aspects of understanding can be revealed through words and through drawings” (Akaygun & Jones, 2014:783). Akaygun and Jones also found in a study that they conducted about a comparison between written and pictorial explanations of physical and chemical equilibria, that people conveyed the same type of information using different visual tools, regardless of their level of knowledge of the subject.

3.4 Types of illustrations and their affordances

According to the Free dictionary (date of access: 12 Nov. 2017) an illustration can be seen as any image that represents, decorates or explains written text. As seen in the different sub-sections of this section it was found that the different types of illustrations have affordances that are unique to the type of illustration. Also the different types of images (in particular photographs diagrams and models) can be used for different

(36)

27 purposes, depending on the type of information it means to portray, the level of understanding it means to assist and whether it includes deeper perspectives.

3.4.1 Photographs

A photograph can be seen as a very simple form of visualisation and is a copy of whatever it means to portray, but it does not elaborate on any perspective as to the deeper meaning and does not give any insight as to relationships to other objects or scenarios (Uttal & O’ Doherty, 2008:57). When photographs are used as a means of representation, the background information, captions and different perspectives should not be neglected in order for photographs to have the desired effect (Gilbert, Reiner & Nakhleh, 2008). Gilbert et al (2008) also maintains that photographs are seen as a more “real” form of representation and is therefore not usually discussed among other forms of representation. It can be used as decorative as well as illustratively but mostly in subjects such as Biology. Uttal and O’Doherty (2008:58) also maintains that even though people can distinguish differences in photographs, it does not make it a good form of representation since it does not always convey the message that it means to emphasise. It can therefore be concluded that photographs are sometimes used for the purpose of conveying information but is mostly for descriptive purposes and does not give deeper insight into the deeper meaning.

3.4.2 Diagrams

Diagrams are tools that show relationships between certain aspects of a specific phenomenon and are made up of certain “conventions”, such as arrows, labels and colour (Cromley et al., 2016). Dictionary.com (2017) also outlines diagrams as something that describes the process of a specific phenomenon.

Diagrams are abundantly used in school textbooks but diagrams vary greatly in design and whether it assists in learning had not yet been determined (Gilbert et al., 2008). Butcher (2006) argues that diagrams that are designed to help the process of learning are more likely to support deeper comprehension of the represented information,

Diagrams that are designed to help the process of learning are more likely to support deeper comprehension of the represented information (Butcher, 2006). Booth and Koedinger (2011) state that learners with higher abilities do benefit from diagrams, but

(37)

28 learners with lower ability do not, and it may even hinder their performance. Therefore diagrams should be used cautiously for learners at a young age. In a study done by Hinze et al. (2013) learners were given a familiar diagram and were expected to solve chemistry problems by using this diagram. It was found that learners with more pre-knowledge relied more on the diagrams than learners will less prior-pre-knowledge.

Hembree (1992) concluded that across 16 studies learners had more correct answers to word problems if it was accompanied by diagrams, but according to De Bock, Verchaffel, Janssens, Van Dooren and Claes (2003) diagrams can be harmful in learning high school geometry as well as primary school arithmetic word problems (Berends & Van Lieshout, 2009).

According to some studies, diagrams thus aid in understanding and therefore learning while other studies show impairment of learning.

3.4.3 Scientific models

According to Dictionary.com (2016) a model in science is “A systematic description of an object or phenomenon that shares important characteristics with the object or phenomenon”. Reference.com (2016) defines a scientific model as: “a conceptual, mathematical or physical representation of a real-world phenomenon” and continues that a model is made for a phenomenon that is “partially understood” but not easy to observe. The Collins English Dictionary (2009) describes a model as a way of constructing or showing how something appears and therefore puts the illustration into context. A model is never completely connected to real-life and the relationship between the model and the application is often left out of the model (Hart, 2008:530). From these definitions and descriptions it is seen that scientific models represent selected characteristics or descriptions of real world objects or phenomena for enhanced understanding.

(38)

29 Models can be categorised in five different categories (Van Driel & Verloop, 2002:1257):

(i) a model is directly related to the target it represents;

(ii) a model is simply a tool through which research is conducted or information is gathered; in this instance it is about a target that cannot be accessed with a direct approach;

(iii) a model resembles the target, and therefore allows the researcher to derive a hypothesis merely from studying the model;

(iv) category four shows differences between models and the target and in this instance models are usually kept as simple as possible;

(v) the fifth category pertains to the designing of the model and different compromises that must be made in order for the model to be scientifically correct so that the researcher can use it to derive certain choices.

Using scientifically correct models can improve a learners' understanding of different concepts, especially in subjects such as Chemistry (Nakhleh & Postek, 2008). Whitlock et al. (cited by Du Plooy, 2012:14) agrees that models are part of science methods and are thus important in science education.

3.4.4 Other types of illustrations

A picture is a type of illustration that can be a general term to describe graphics, illustrations, and photographs or maps (Mayer cited by Akaygun & Jones, 2014:785). Tversky (cited by Akaygun & Jones, 2014:785) also maintains that written language is developed from pictures in order to better convey abstract concepts. The term “pictures”, therefore, includes illustrations and graphics and is a broad term.

Paintings are also a way of conveying information and Akaygun and Jones (2014) state that according to history, paintings and engravings were used long before written language and therefore, in some cases, represent meaning better than written language.

(39)

30

3.5 Visualisation and the formation of mental models

According to Clement (cited by Borges, Tecnico & Gilbert, 1998: 362) direct model building is closely related to reasoning in every-day life. Model building is also characteristic of the work of scientists (Zwickl, Hu, Finkelstein & Lewandowski, 2015). This section of Chapter 3 focuses on how visualisations affect the formation of mental models in learners. Section 3.5.1 first describes the different types of models and section 3.5.2 gives more detail about mental models.

3.5.1 Types of models

The five different types of models are mental models, expressed models, consensus models, historical models and teaching models (Du Plooy, 2012:12). A mental model is the way that people see or think about a specific concept (Gentner & Stevens cited by Borges et al., 1998: 361). A consensus model is a model that is used when a community wants an explanation for a particular aspect or phenomenon (Hart, 2008:529). A working model is a model constructed from understanding and as the understanding grows the model changes (Borges et al., 1998: 362). Greca and Moreira (1997) maintains that a working model is a model that explains a mental model through perceptions, proportions and imagination.

3.5.2 Mental models

As previously stated a mental model is the way people perceive a specific concept (Gentner & Stevens (cited by Borges et al., 1998: 361)). Greca and Moreira (1997) describes mental models as “structural analogues” of, among others images, language and symbols and sees the world from a specific point of view. A mental model represents a small area of reality and is used to make predictions and draw conclusions, and is often used to solve problems (Molitor et al., 1989:10). Akaygun and Jones (2014) state that textual as well as pictorial tools can be used to represent a mental model.

Learners bring their own mental models into the classroom with them and thus use it to understand the world around them (Greca and Moreira, 1997). Grosslight et al. (cited by Du Plooy, 2012:1) maintains that a mental model or representation creates