Teaching and learning chemical bonding: research-based evidence for misconceptions and conceptual difficulties experienced by students in upper secondary schools and the effect of an enriched text†
Georgios
Tsaparlis
*a,
Eleni T.
Pappa
a and
Bill
Byers
b
aDepartment of Chemistry, University of Ioannina, Ioannina, Greece. E-mail: gtseper@cc.uoi.gr
bSchool of Health Sciences, Ulster University, Jordanstown, UK
First published on 18th June 2018
Abstract
Chemical bonding is a fundamental but complex topic, which has traditionally been associated with learning difficulties, misunderstandings, and misconceptions. This paper reviews some previous studies, concerning students’ conceptual difficulties and reports the findings from a research study with Greek students, which set out to examine their knowledge and understanding of a number of key concepts related to bonding. Three student samples were studied; one consisted of tenth-grade students from three public schools, the second contained first-year chemistry and biology students at the beginning of their university courses, and the third involved tenth-grade students from a prestigious private school. The students generally exhibited limited knowledge and possessed certain misconceptions, with the private school and the university students demonstrating better knowledge than the public school students. A quasi-experimental research design was employed using students from the private school, with some students used as a control group and others as a treatment group. The control group was taught using the standard Greek chemistry textbook, while the treatment group used enriched teaching material. It was found that while the two groups demonstrated similar performance for many bonding concepts, the treatment group did show superior knowledge with respect to a number of issues, such as the role of electrostatic interactions, electronegativity, and bond polarity.
“A ‘chemical bond’ is what holds neighbouring atoms together to form the intricate structures of the world. All chemical bonds result from changes in the distribution of electrons in the bonded atoms…” Peter Atkins (2014, p. 11)
Introduction
Student conceptual difficulties with chemical bonding
Chemical bonding is one of the most fundamental concepts in chemistry (Gillespie, 1997; Atkins, 1999, 2010; Taber and Coll, 2003; Holme and Murphy, 2012; Holme et al., 2015), and is directly associated with the understanding of many chemistry concepts taught in secondary schools and at university, such as chemical reactions and structure–property relationships. However, although the topic of chemical bonding is fundamental to the teaching of chemistry, it has proved very complex to organize and apply for curriculum designers, teachers, and students alike. Chemical bonding can be of several types (ionic, covalent, polar and nonpolar and metallic bonding) and it involves a number of concepts including molecule, atom, proton, neutron, electron, ion, cation, anion, attraction of opposite charges and repulsion of like charges.The conceptual difficulties that students encounter when trying to understand chemical bonding lead many of them to resort to rote learning and the formation of numerous misconceptions (e.g.Levy Nahum et al., 2007), many of which prove resistant to instruction (e.g.Nicoll, 2001; Taber and Coll, 2003; Özmen, 2004). It is perhaps not surprising then that the learning of chemical bonding has been the subject of many investigations, e.g.: Peterson et al. (1986, 1989), Butts and Smith (1987), Peterson and Treagust (1989), Taber (1994, 1997, 1998), Robinson (1998), Birk and Kurtz (1999), Tan and Treagust (1999), Barker and Millar (2000), Harrison and Treagust (2000), Boo and Watson (2001), Coll and Taylor (2001), Coll and Treagust (2001, 2002), Niaz (2001), Nicoll (2001), Taber and Coll (2003), Özmen (2004), Kronik et al. (2008), Othman et al. (2008) and Ünal et al. (2010). Ünal et al. (2006) carried out a detailed thematic review of chemical bonding studies by developing a matrix, which summarized their findings. The matrix incorporates the following themes: needs, aims, methods of exploring students’ conceptions, general knowledge claims, students’ alternative conceptions, implications and recommendations for teaching and learning, implications for curriculum development, and suggestions for future research.
It is well known that teachers rely on and closely follow textbooks in their everyday practice (Britton et al., 1993). However, when taught in the traditional way, the concept of chemical bonding can lead to misunderstandings by learners. Many of these misunderstandings result directly from the oversimplification of models used in student textbooks. In fact, misunderstandings can arise both from the presentation of a rather limited and sometimes incorrect picture of the issues related to chemical bonding and, also, because the way the topic is taught is often heavily influenced by the assessment procedures (Levy Nahum et al., 2010). Thus, the teachers’ main aim is often to prepare their students for future matriculation examinations. This can lead to a simplistic teaching approach (Kronik et al., 2008, Fig. 1, p. 1680), with superficial teaching, resulting in shallow learning and ultimately producing misconceptions and pseudo-conceptions.
In a seminal paper, based on an interview study with 16–19 year-old UK students, Taber (1998) proposed an ‘alternative conceptual framework’ for chemical bonding, according to which the ‘octet rule’ is used by the students as an explanatory framework for chemical stability and reactivity. Unfortunately, however, this can be responsible for “the development of a spread of ‘misconceptions’”. A common feature is that students adopt an anthropomorphic approach in their explanations of chemical bonding, thinking that atoms need or desire to fill their shells so that they obtain octets of outer electrons and acquire noble gas electronic configurations.‡ The key characteristic of student thinking in this UK study was that species with ‘full’ valence electron shells (usually octets of electrons) have an inherent stability. This was seen as a driving force for chemical reactions, with reactions occurring to allow atoms to fill their shells. Taber (1999) considered this framework as a largely coherent theory-like basis for thinking about chemical stability, change and bonding (for a review, see Taber, 2013). In addition, he has provided both a theoretical background (Taber, 2002a) and classroom tools (Taber, 2002b) for diagnosing relevant student misconceptions and testing the popularity of the ideas elicited in the 1998 interview study. Focusing on student reasoning about chemical substances and reactions (including bonding), Talanquer (2013) maintains that in building their explanations, students fail to recognize leading causal agents or any other type of causal mechanism and demonstrate a cognitive bias toward teleological explanations, which are pervasive and resilient to conceptual change.
Following the paper by Levy Nahum et al. (2007) (see also: Kronik et al., 2008), which developed “a new teaching approach (the ‘bottom up’ approach) for the chemical bonding concept aligned with current scientific and pedagogical knowledge”, there has been a revival of interest in teaching and learning of the topic of chemical bonding. Levy Nahum et al. (2013) have reviewed their work on chemical bonding.
Tsaparlis and Pappa (2012) examined fourteen general chemistry textbooks and identified similarities and differences with regard to the following aspects: presentation order of bonds (covalent and ionic bonds); placement and method of presentation of intermolecular bonds; the bond continuum; physical states and their relationship to the types of bonding; metallic bonding; the octet rule; electronegativity and bond polarity; and coordinate bonding. Almost all the books presented covalent and ionic bonding before intermolecular bonding, with a majority of books introducing the ionic bond before discussing covalent bonding. Most books refer to the covalent and ionic bonding continuum, though some continued to treat the two topics completely independently. Electronegativity and bond polarity are presented within covalent and ionic bonding, which most books seem to consider as the only true types of chemical bonding, with intermolecular bonds often being merely referred to as forces. While consideration of metallic bonding is included in the majority of these textbooks, many fail to even mention the coordinate bond. Finally, regarding intermolecular bonds, the books follow different orders of presentation, with most attention often given to hydrogen bonding.
Taber et al. (2012) used a diagnostic instrument, previously used in England (Taber, 1994, 1997, 1998), to elicit the conceptions about bonding in NaCl (commonly used as the teaching example of an ionic compound) from two samples of students setting out on university courses in Greece and Turkey. The study reported that students in these two national contexts displayed high levels of support for statements based upon the alternative ‘molecular’ conceptual framework for thinking about ionic bonding identified in the English context, reinforcing the cross-national nature of the challenge to the effective teaching of abstract models of chemistry at the submicroscopic level. The study was later repeated in Croatia (Vladušić et al., 2016) and the findings were consistent with the previous ones.
Yayon et al. (2012) developed and tested a matrix, consisting of strands, to represent: (a) a systematic organization of the conceptual knowledge on chemical bonding required at high school level and (b) a tool for representing students’ conceptual knowledge of this topic. Using various instruments, students’ conceptual knowledge of chemical bonding was assessed and mapped onto the matrix, generating graphical representations of their knowledge. The matrix contains three strands: (i) the structure of matter at the nanoscopic level; (ii) electrostatic interactions between charged entities; and (iii) energy aspects related to bonding. Each strand consists of hierarchically ordered cells that contain fine grain elements of knowledge: the unbound atom; one bond between two atoms; a single molecule; many bonds of the same type (lattice); many bonds of different types; states of matter; overview. However, the matrix deals only with canonical knowledge, so misconceptions were not considered.
Bergqvist et al. (2013) analysed models of chemical bonding presented in chemistry textbooks at upper secondary level in Sweden and investigated themes in the textbooks which are related to students’ alternative conceptions and difficulties in understanding chemical bonding. The authors quoted examples of students’ alternative conceptions and difficulties in understanding for the different types of chemical bonding as these have been reported in the research literature, followed by possible sources of these difficulties according to the literature. The findings indicated that the models of chemical bonding represented in school textbooks are likely to be a cause of students’ alternative conceptions and difficulties in understanding. They therefore, suggested that there was a need to close the gap between textbook writers and current educational research.
Wang and Barrow (2013) carried out a comparison of the conceptual frameworks of chemical bonding possessed by undergraduate general chemistry students with high and low levels of content knowledge. It was found that high content knowledge students viewed chemical bonding as an electrostatic force between two atoms and justified bond formation with the octet rule and stability; in addition, they were comfortable with exceptions to the octet rule, and associated bond polarity with differences in electronegativity. On the other hand, low content knowledge students viewed chemical bonding as some type of (attractive) force between two atoms and justified bond formation with the octet rule and stability or with teleological explanations. They tended to be uncomfortable with exceptions to the octet rule and to possess misconceptions, but, like the high content knowledge students, they associated bond polarity with differences in electronegativity.
Hazzi and Dumon (2014) examined the extent to which Algerian undergraduate students were able to integrate the Lewis and quantum models of atoms and molecules. It was found that the students were not able to correctly describe covalent bonds in a Lewis structure using the concepts of the quantum model. According to the authors, students’ conceptions of atomic or hybrid orbitals act as a pedagogical impediment to learning, so the students use the electron pair framework, which is based on the use of Lewis model paired valence electrons, to form covalent bonds. In addition, the students believed that a covalent bond arises from the sharing of one electron (either s or p but not spn) from each atom to give one σ electron pair, and thought that σ bonding occurs only in single bonds.
Luxford and Bretz (2014) conducted interviews with 28 high school physical science and chemistry students, and general chemistry university students in order to investigate their understanding and misconceptions about covalent and ionic bonding representations through analysis of both student-created and expert-generated multiple representations used to communicate bonding concepts such as Lewis structures, formulas, space-filling models, and 3D manipulatives. Misconceptions were identified regarding four themes: (i) periodic trends; (ii) electrostatic interactions; (iii) the octet rule; and (iv) surface features associated with the representations (for instance, lines indicate bonds or spacing of the dots between atoms indicates equal sharing). Furthermore, a concept inventory, the Bonding Representations Inventory (BRI), was developed to quantify the prevalence of these misconceptions. The BRI was administered to 1072 high school chemistry, advanced placement chemistry, and general chemistry students across the United States, and content validity, concurrent validity, and reliability of the instrument were determined. The BRI was subsequently tested successfully in the Slovak educational system, where misconceptions were identified, and the results were compared with those in the USA (Vrabec and Prokŝa, 2016).
Burrows and Reid Mooring (2015) used concept mapping to uncover undergraduate general chemistry students’ knowledge structures about chemical bonding concepts and detected several common misconceptions concerning electronegativity, with a weaker understanding of electronegativity among students with low concept map scores than for students with high concept map scores. Further, a lack of understanding of electronegativity led to a misunderstanding of polar covalent bonding.
Cheng and Oon (2016) developed an instrument and carried out a very large (n = 3006) survey of year 10–12 students on their understanding of metallic bonding. The instrument probed into students’ understanding of metallic bonding as: (a) a submicro structure of metals; (b) a process in which individual metal atoms lose their outermost shell electrons to form a ‘sea of electrons’ and octet metal cations; and (c) an all directional electrostatic force between delocalized electrons and metal cations. According to the findings, the structure, process and interaction understandings were unidimensional and in an increasing order of difficulty. The instrument also assessed students’ explanations for the malleability of metals.
Nimmermark et al. (2016) carried out a large study with almost 700 Swedish and South African chemistry students from upper secondary school and first-term university levels, and reported that only about 20% of students had a clear grasp of the concept of bond energetics on coming to university, often confusing individual bond formation and breakage with exo- and endothermic processes. Also, an over-use of ball and stick models can lead to students having difficulty with visualization of the correct shape of molecules. Students who had encountered the VSEPR model in secondary school tended to have a better grasp of molecular shapes. It was also reported that an understanding of the quantum mechanical model of bonding was beneficial to undergraduate students.
Ballester Pérez et al. (2017) investigated Spanish students’ understanding of chemical bonding and their related misconceptions. The students came from secondary schools and from first-year undergraduate chemistry and pharmacy courses, while the topics studied included the interpretation of some properties of substances (colour, boiling points, solubility and conductivity), intra- and intermolecular forces, hydrogen bonding, covalent and molecular networks, geometry and polarity of molecules. The main misconceptions were: attributing macroscopic properties to particles; incorrect prediction of boiling points; perceiving ionic compounds as being formed by molecules; misunderstanding the nature of the hydrogen bond and assuming that it is present in any molecule containing hydrogen together with nitrogen, oxygen or fluorine, regardless of whether the hydrogen atom is directly bonded to these atoms or not; confusing the geometry of a molecule with the distribution of electron pairs around the central atom; and predicting the polarity of molecules incorrectly.
In conclusion, the prevailing approach to teaching chemical bonding continues to lead to learning difficulties and misunderstandings for many students. The following are some typical examples:
(a) The ionic bond is referred to as just a transfer of electrons between separate atoms in order to acquire full valence shells.
(b) The model of the covalent bond is described as an electron-pair that is shared between two atoms.
(c) The octet rule is often used by the students as an explanatory framework for chemical stability and as a prerequisite for a ‘proper’ bond.
(d) Covalent and ionic bonds are often presented in isolation, as bonds that share and bonds that transfer electrons respectively.
(e) Bond polarity is directly linked to the covalent bond.
(f) Covalent and ionic bonds are described as ‘real’ chemical bonds, while most intermolecular bonds are simply referred to as forces.
The present study
This study used a written assessment instrument to evaluate the understanding of a number of basic chemical concepts, such as electronegativity, Lewis structures, the octet rule, Coulomb's forces, bond polarity, etc., used in the study of ionic and covalent (polar and nonpolar) bonding. Other types of bonding were not considered in this work, although they are included in some of the studies reviewed above. Three samples of students, chosen for convenience, were studied: one consisted of tenth-grade students (age 15–16) from three public schools; the second was from a tenth-grade class at a prestigious private school; and the third consisted of first-year chemistry and biology students at the beginning of their university courses (age 17–18). Our research questions were as follows:(1) To what extent do Greek upper secondary school students share a number of common misunderstandings and misconceptions about chemical bonding?
(2) What is the effect of an enriched educational text, based on the standard Greek chemistry textbook, on preventing the various misunderstandings and misconceptions in the case of high-achieving, tenth-grade students?
In a follow-up publication to this work, we will consider a spiral curriculum and learning progressions in the teaching of chemical bonding, taking into account recommendations from the research literature and our findings/answers to the two research questions posed above.
Method
Educational material
With the exception of the treatment group of students (see below), all students of our study were taught using the standard chemistry textbook for the tenth grade (Liodakis et al., 2012), produced by the former “Greek Pedagogic Institute” (currently: “Institute of Educational Policy”), which is used throughout Greece, for their study. Teachers are required to follow this text closely, although they are also encouraged to include their own personal knowledge and views into their teaching. We label this text, which includes “Exercises and Problems”, the ‘standard text’. See Appendix 1 (ESI†) for a reproduction of this material in English. The translation from Greek into English was carried out by one author (GT) and checked by another (BB).The treatment group (TG) was taught using a modified/enriched text, which was based on the standard text, had the same format, and included the same exercises and problems at the end, but only for the topic of chemical bonding. This material was written by one of the authors (GT), who took into account, on the one hand the education research literature on the topic of chemical bonding, and on the other hand his personal scientific knowledge of this topic. Special attention was paid to the findings of the Tsaparlis and Pappa (2012) study of the way in which general chemistry textbooks presented the topic of chemical bonding (see above).
Two chemistry teachers had previously trialled an initial version of the material with their tenth grade classes, and provided feedback about its strengths and weaknesses. Their feedback was taken into account in writing a second version, which was subsequently reviewed by the two teachers and by a third chemistry teacher. The comments of these three reviewers were used in revising the material to provide the final text, which we label as the ‘enriched text’. Each student in the TG was supplied with a booklet containing a printed copy in colour of this enriched text (in Greek). However, these students also had available the standard textbook, so they could readily consult and compare both texts. See Appendix 2 (ESI†) for a reproduction of the enriched text in English. The translation from Greek into English was again carried out by one author (GT) and checked by another (BB).
Regarding the basic features and differences between the standard and enriched texts, we should first acknowledge that the enriched text involves a considerable lengthening from 10 pages in the original Greek text, not including exercises and problems (9 pages in the English version, see Appendix 1, ESI†), to about 15 pages (both in Greek and in English), that is, an increase by a factor ca. 1½. We also note the following in brief (see Appendix 3 (ESI†) for a detailed comparison):
• Electrostatic interactions: The standard text does not mention anywhere the term ‘electrostatic interactions’, but it refers to electrostatic Coulomb forces. The enriched text also does NOT refer to ‘electrostatic interactions’ but does mention ‘Coulomb's law’, ‘forces of electrostatic nature (Coulomb forces)’, and ‘Coulombic interactions’.
• Octet rule: The enriched text includes a more detailed discussion about the octet rule, with emphasis on the fact that the octet rule does not “impose” the formation of an ionic bond, but is able to explain the charges on the ions and the stoichiometry of ionic compounds.
• Electronegativity. The enriched text gives a more detailed coverage, including the use of a quantitative scale (the Pauling scale) to this topic, and its contribution to the understanding of the continuum between ionic and covalent bonding.
• Bond polarity. While the standard text does not use the term ‘bond polarity’ or ‘polarity’, the enriched text refers explicitly to this concept. Both texts, however, refer extensively to nonpolar and polar bonds.
Type of questions
Questions can be classified, according to their structure, into two broad types: closed, closed-type or closed-ended (also called objective or fixed-response questions), and open, open-type or open-ended questions. Closed-type questions usually only test for recall/reproduction of information, and not for comprehension. There are numerous types of closed-type questions (Macintosh and Morrison, 1969; Pappa and Tsaparlis, 2011). In our work, the first part of the assessment instrument included closed-type multiple-choice questions, while the second part included a number of disjunction-answer (on an either/or basis, e.g. right or wrong) coupled with explanation open-type questions; the second part was concluded with two metacognitive questions. Open-type questions allow for checking higher-level learning, such as correlation of information, conceptual understanding, and critical thinking. They allow the student to use his/her knowledge and ideas, and the evaluator to infer the degree to which the student understands the information.Pearson and Johnson (1978) proposed a classification that is particularly suitable for the evaluation of questions (see also Pappa and Tsaparlis, 2011). Their classification refers to three kinds of knowledge: declarative knowledge, procedural knowledge, and metacognition. Declarative knowledge is knowledge of what, while procedural knowledge is knowledge of how. Procedural knowledge involves higher-level learning, requiring justification for the preference of one process instead of another, the making of hypotheses, observation, recording of data, and arriving at conclusions. Finally, metacognition aims to identify the extent to which the student has acquired autonomous thinking and action (Schneider, 2004).
Pappa and Tsaparlis (2011) evaluated the questions contained in the chapters on chemical bonding of a number of general chemistry textbooks, and reported that closed-type questions, by far, outnumbered the open-type ones. Questions were mainly of the ‘short answer’ type, with the majority testing for declarative knowledge and fewer questions testing for procedural knowledge. A complete lack of metacognitive questions was found.
Research instrument
In the case of the schools, a written assessment instrument (henceforth to be simply called the ‘test’) was distributed to the students after the completion of the topic of chemical bonding. To avoid collusion between adjacent students in class, two equivalent forms of the test, with different ordering of the concepts and the question were used.The test consisted of two parts. The 1st part had to be answered first, and gave the students a list of concepts that they had to identify according to whether:
• they are used or not for the description of bonds, and if so:
• what type or types of chemical bond they describe.
Next for each term, the students were asked to assign one of the following symbols:
• ‘IB’, if the term is used to describe the ionic bond only;
• ‘CB’, if the term is used to describe the covalent bond only;
• ‘IB&OB’, if the term is used to describe both ionic and covalent bonds;
• ∅, if the term is used to describe NEITHER ionic NOR covalent bonds;
• N/A: I don’t answer.
The concepts given to the pupils are listed in Table 1.
| Energy | Electronegativity | Density | Volume |
| Chemical bond | Mass | Covalent compound | Molecule |
| Ionic bond | Metal | Mass number | Valence electrons |
| Group | Nonmetal | Coulomb forces | Period |
| Covalent bond | Ion | Lewis structures | Melting point |
| Atomic radius | Crystal | Attractive forces | Percentage |
| Boiling point | Polarity | Repulsive forces | Amount |
| Octet rule | Ionic compound | Solubility |
The 1st part of the test was completed by asking students to use concepts listed in Table 1 to write up to two sentences for each of the following cases: (i) relating exclusively to the covalent bond, (ii) relating exclusively to the ionic bond, and (iii) relating to both types of bonds. A 20 minute timespan was allowed for completing the 1st part of the test.
In the 2nd part of the test, learners were given four questions (see Table 2). The first question entailed six disjunctive (Right/Wrong) statements about basic concepts regarding the chemical bond; in addition, students had to justify their answers. The second question concerned the criteria used for determining the type of bond. The six concepts of question 1 together with question 2 provide the seven items that are examined in Part 2 of the test of the current paper. Finally, there were two questions of the metacognitive type (questions 3 and 4 in Table 2); these required students to state the difficulties they had experienced in connection with the study of chemical bonds. A 20 minute timespan was also allowed for completing the 2nd part of the test.
| a As stated in the text, the six concepts of question 1 together with question 2 provide the seven items that are examined in Part 2 of the test used in the current study. |
|---|
| QUESTION 1. Characterize each of the following statements as correct (C) or wrong (W). Justify your answers. Provide examples where it is necessary. |
| Statement 1. In the ionic bond there are electrostatic interactions, while such interactions do not occur in the covalent bond. |
| Statement 2. The type of bond (ionic or covalent) between two atoms is determined by the valence electrons of the atoms. |
| Statement 3. Polarity occurs only in covalent bonds but not in ionic bonds. |
| Statement 4. Electronegativity plays a role only in covalent bond formation, but not in ionic bond formation. |
| Statement 5. The octet rule provides an explanation for chemical bond formation (ionic or covalent) between two atoms. |
| Statement 6. Lewis structures are used to represent covalent compounds, but not ionic compounds. |
| QUESTION 2. Which criteria could we use to determine if two atoms will combine to form an ionic or a covalent bond? |
| QUESTION 3. Refer to 1 or 2 difficulties, which you encountered in your study of the chapter on chemical bonding. |
| QUESTION 4. Which bond, ionic or covalent, did you find more difficult to understand? Suggest a possible reason for this. |
The level of justifying answers was evaluated using the SOLO taxonomy (Biggs and Collins, 1982; Biggs, 1999), which distinguishes among five levels of understanding: (1) prestructural, (2) unistructural, (3) multistructural, (4) relational, and (5) extended abstract. We will report the findings from this analysis in another publication, but suffice to comment here that the justifications offered by most students for their choices were at the lowest two levels (prestructural, and unistructural) of the taxonomy. Only for one question (that on the criteria for bond-type formation, which just required the enumeration of a number of pieces of knowledge) were there performances that exceeded the second level, while just nine students from among the 46 high-achieving students of the treatment group achieved an average over the second SOLO level. Few answers reached the third (multistructural) level.
Ethical considerations
Tenth-grade students answered the test as part of their chemistry course, that is, they were not given any choice over taking part. The teachers judged this was a valuable exercise to include in their classes, while students were given feedback on the items after the test. In addition, the school director was made aware of the aim of the study and institutional ethics were observed throughout. The tasks of the test were different from the exercises and problems that were included in both the standard and the enriched texts (see end of Appendix 1, ESI†). However, it should be noted that the instructions given by the Greek Ministry of Education to teachers do not prescribe a fixed structure for coursework tests, thus giving teachers freedom to design their own questions. The authors advised the teachers to make students aware of the fact that the questions aimed to assess their understanding of some difficult concepts, such as electronegativity, the octet rule, and bond polarity and that it was hoped that the feedback would contribute to future improvements in the teaching of these concepts. Students were also informed that this test was not to replace the term test and performances in the test would not contribute to their chemistry coursework marks. Students were required to provide their names when answering the test, to encourage meaningful participation, however neither their names nor their performances were to be revealed to any third parties. In the case of the university students who participated in the study, they were told that they would be taking part in a research study of their own free will. Students agreed to take part with no sense that they were being pressured to do so. In their case, also, neither students’ names nor their performances were to be revealed to any third parties. Performance in the test did not contribute to their course mark.Student samples
The original survey was conducted in the school year 2012–2013 using three samples taken for convenience. The first sample consisted of 66 tenth-grade students from three public schools (PuS) in Greece, one was in a provincial city centre and two in semi-urban towns, with 22 pupils taken from each school (total sample of PuS: 66). The second sample consisted of 72 first-year undergraduate students from two departments at a provincial Greek university: chemistry (CHE, n = 34)§ and biology (BIO, n = 38). The third sample consisted of 49 tenth-grade students from a prestigious private school (PrS) in a major Greek city (this sample will be identified below as the control group [PrS(CG) or simply CG] for comparison with a treatment group [PrS(TG) or simply TG]).¶The rationale for including the two university departments in our study was as follows. The students were at the very beginning of their university studies, so their knowledge about chemical bonding derived from their overall studies in secondary education. Because of the subjects of their study (chemistry and biology), all these students had studied chemistry as an advanced course and chemistry was among the subjects examined in their matriculation examinations.
Before proceeding, it is also essential to explain why we have treated, for convenience, the three public schools as a single sample and the two university departments as a single sample. There was a large variation in individual performance for students at the public schools. This could be due (at least in part) to random factors (such as guessing) plus the added fact that we are dealing with small samples (n = 22) from each school. The averaged performances for each school over the seven items of the 2nd part of the test of this study were found to be 34.4, 44.8, and 46.1%, showing a similar overall performance for the second and third schools, with the first school lagging behind. So our treatment of the three schools as a single group (PuS, n = 66) can be justified by the fact that they had the lowest performances among our three samples, but, in any case, our sampling frame should not be assumed to be representative of the average Greek public school.
In the case of the students from the two university departments, there was again a variation of performances which could also be due (at least in part) to random factors [in addition, the samples (n = 34 and 38 respectively) were again not large]. However, the average performance for each sample in the seven items was about the same: 55.9% for CHE and 57.5% for BIO. For this reason and for convenience, we combined the two university departments into a single group (university students, US). Note that the corresponding average (54.3%) for the control group from the private school (see below) is similar to that for the university students (56.8%). We have not combined these two samples (US and PrS), both because they correspond to quite different populations in terms of origin and age and also because of the differing demands for knowledge retention: the school students were tested on completion of the teaching, whereas the university students had received their teaching some considerable time before testing.
In the case of the tenth-grade students from the private school, the survey was conducted in two parts: in school year 2012–2013 for the control group [PrS(CG) or simply CG, n = 49]; and in school year 2015–2016 for the treatment group [PrS(TG) or simply TG, n = 46]. The teacher had a master's degree in chemistry education and new educational technologies, had taught at this particular school for sixteen years, and in addition had five years of experience as a private tutor. Students in the CG were taught the topic of chemical bonding using the standard textbook (see Appendix 1, ESI†), while students in the TG were taught this topic making use of the enriched text (see Appendix 2, ESI†).
At this point, it is important to explain exactly what we mean by the term ‘prestigious private school’. The school in question has outstanding facilities including laboratories, is renowned for the quality of both students and teachers, and provides a range of extracurricular activities. Because, of the tuition fees, the school draws its students from relatively high-income families. We provide above, in Table 3, data for the coursework, final examination and cumulative yearly student evaluation in the chemistry course, and the overall average yearly mark achieved for all courses, to demonstrate the quality of students at this private school. Table 3 includes school performance data for the students at the school for both CG and TG samples. Note the lower marks in the final end-of-year written examination in chemistry, compared to those for continuous assessment based on overall course participation, including oral and written test performance.
| Chemistry: continuous assessment | Chemistry: final examination | Chemistry: cumulative yearly mean marka | Overall yearly mean mark | |
|---|---|---|---|---|
| a Average of continuous assessment and final examination. b Excluding students with a cumulative mean mark in chemistry below 80%. | ||||
| Total sample, n = 53/46 | ||||
| Average (%) | 92.5/90.2 | 85.8/76.5 | 89.2/83.4 | 84.7/87.0 |
| Std deviation (%) | 6.7/10.4 | 13.4/17.8 | 9.3/13.2 | 7.4/9.2 |
| Reducedb sample, n = 47/33 | ||||
| Average (%) | 93.7/94.8 | 89.4/85.2 | 91.6/90.0 | 85.8/91.2 |
| Std deviation (%) | 5.9/4.2 | 7.7/9.4 | 6.1/6.0 | 6.8/5.3 |
Table 3 considers two samples of students for both the CG and the TG: the total sample (n = 53 and 46 respectively),|| and a reduced sample (n = 47 and 33 respectively), in which the 6 and 13 students respectively with a cumulative yearly mean mark in chemistry below 80% have been excluded. We observe that 89% of the students in the case of the CG and 72% in the case of the TG were high achievers, while achievements by the remaining students (more students in the case of the TG) were more modest.
Comparison of the results obtained by the CG and TG (whole samples) by means of the statistic t for independent samples (two-tailed tests), using the separate variances formula, gives the following values: for the chemistry continuous assessment, t = 1.29, which is not statistically significant; for the final chemistry examination, t = 2.90, which is significant at the 0.01 level; and for the cumulative chemistry mark, t = 2.49, which is statistically significant at the 0.05 and 0.02 significance levels but not at the 0.01 level; finally, for the overall yearly mean mark, t = 1.35, which is not significant.
Note that it is results in the chemistry final examination, which have resulted in a significant difference in favour of the CG, but the difference might not be due to any actual difference in teaching between the CG and the TG but could reflect more demanding exam questions in 2015–2016 compared to 2012–2013. The two PrS samples (the CG and the TG) can be considered as more or less equivalent for a number of other reasons: (i) they were from the same school; (ii) the selection method and the method of awarding bursaries to part of the students were the same, (iii) they had the same chemistry teacher, (iv) they were chosen randomly from the tenth-grade populations of school years 2012–2013 and 2015–2016 respectively; and (v) no changes in the curriculum, the chemistry textbook or the school programme occurred between 2012–2013 and 2015–2016. In any case, the possibility that the CG had a small superiority with respect to the TG can only serve to strengthen our findings from the comparison between the CG and the TG.
Results and discussion: 1st part of the assessment instrument
At the outset, it is necessary to admit, that, with the exception of the comparison between our control and treatment groups from the private school, there is no justification for trying to rationalize differences between the groups examined in this study. Our grouping together of the three public schools (first sample) and of the two university departments (second sample) was done only for convenience.1st part of the assessment instrument, all samples
We begin by dividing the concepts listed in Table 1 into four general classes:• Concepts that refer to tautologies (chemical bond, ionic bond, covalent bond, ionic compound, covalent compound).
• Concepts that have no relevance to chemical bonding (mass, volume, density, percentage amount).
• Concepts that are directly relevant to bond-type and bond formation (e.g. energy, electronegativity, octet rule).
• Concepts that are only indirectly relevant to bond-type and bond formation (e.g. mass number, metal, non-metal, solubility).
Knowledge of the first two classes was used to test the equivalence of the control and treatment groups and the general reliability of our data.
Abstentions: Examining the frequency of the option N/A (I don't answer), we note that the abstentions varied considerably from topic to topic but overall levels were disappointingly high, suggesting a lack of understanding. BIO and TG students seem to use this option the least, with frequencies <10%, while levels for CHE were high. PuS students also exhibited several high rates of abstention.
Knowledge of tautological concepts (chemical bond, ionic bond, covalent bond, ionic compound, covalent compound): With the exception of the PuS in the case of the chemical bond concept, correct responses ranged from 84 to 100%, with 17 out of the 20 cases being above 90%. In the case of the chemical bond concept (where as correct we assumed the IB&CB option), PuS had only 45% correct answers, with 22% opting for the ∅ answer, that is, the term ‘chemical bond’ is neither used to describe ionic nor covalent bonds. This finding shows that for many students the adjective ‘chemical’ differentiated a bond from both ‘ionic’ and ‘covalent’ bonds, and demonstrates the overall lower level of knowledge of the PuS students.
Knowledge of concepts having no relevance to chemical bonding (mass, volume, density, percentage amount). Frequencies of the expected ∅ choice ranged from 65 to almost 90% (96% for TG, with the PuS having the lowest frequencies).
Knowledge of concepts directly relevant to bond formation Table 4a gives the results for the first part of the test for knowledge of the concepts directly relevant to bond-type and bond formation.
| Knowledge element | Public schools | University students | Private school (CG) | Private school (TG) |
|---|---|---|---|---|
| a All types of force are involved in both IB and CB. b All entries are relevant for both IB and CB. | ||||
| (a) | ||||
| Ion (IB) | 57.6 (22.7 IB&CB) | 66.7 (23.6 IB&CB) | 59.2 (28.6 IB&CB) | 80.4 (13.0 IB&CB) |
| Crystal (IB) | 66.7 (13.6 CB) | 69.4 (11.1 IB&CB, 9.7 CB) | 89.8 | 91.3 |
| Molecule (CB) | 43.9 (21.2 ∅, 18.2 IB&CB) | 34.7 (43.1 IB&CB, 11.1 ∅) | 63.3 (18.4 IB&CB, 10.2 ∅) | 84.4 |
| Metal (IB) | 39.4 (25.8 CB, 22.7 IB&CB) | 66.7 (22.2 IB&CB) | 83.7 | 82.6 |
| Nonmetal (IB&CB) | 28.8 (33.3 IB, 27.3 CB) | 33.3 (47.2 CB, 12.5 IB) | 30.6 (55.1 CB, 10.2 N/A) | 17.4 (76.1 ∅) |
| (b) | ||||
| Coulomb forces (IB&CB) | 30.8 (21.5 ∅, 21.5 IB, 13.9 CB, 12.3 N/A) | 26.4 (29.2 IB, 19.4 ∅, 13.9 N/A,11.1 CB) | 8.2 (61.2 IB, 18.4 ∅) | 87.0 |
| Attractive forces (IB&CB) | 66.7 (12.1 IB) | 83.3 (8.3 IB) | 71.4 (12.2 IB, 10.2 CB) | 91.3 |
| Repulsive forces (IB&CB) | 27.3 (30.3 ∅, 13.6 IB, 13.6 CB, 15.2 N/A) | 48.6 (16.7 CB, 15.3 IB, 15.3 ∅) | 4.1 (65.3 ∅, 16.3 CB, 10.2 IB) | 17.4 (76.1 ∅) |
| (c) | ||||
| Energy (IB&CB) | 29.8 (50.8 ∅) | 73.6 (16.7 ∅) | 69.4 (22.4 ∅) | 82.6 |
| Valence electrons (IB&CB) | 53.8 (30.8 IB) | 58.3 (16.7 CB, 13.9 IB) | 87.8 | 89.1 |
| Octet rule (IB&CB) | 47.0 (21.2 IB, 12.1 ∅, 10.6 CB) | 59.7 (19.4 CB) | 77.6 (12.2 ∅ 8.2 IB) | 84.8 |
| Electronegativity (IB&CB) | 19.7 (39.4 CB, 13.6 IB, 13.6 ∅, 13.6 N/A) | 49.3 (23.9 CB, 14.1 IB, 8.5 ∅) | 36.7 (30.6 ∅, 16.3 CB, 12.2 IB) | 95.7 |
| Polarity (IB&CB) | 28.8 (39.4 CB, 16.7 ∅) | 29.2 (41.7 CB, 19.4 IB) | 12.2 (65.3 CB, 12.2 ∅) | 80.4 |
| Lewis structures (IB&CB) | 28.4 (52.2 CB) | 68.1 (20.8 CB) | 87.8 | 73.9 |
Ion. The expected answer for this concept was that it is relevant to IB only. However, for PuS and CG less than 60% of the students provided the correct answer. Considerable percentages (>20%) of answers favoured the option IB&CB in all cases (except for TG: 13.0%). The TG had the highest percentage correct answers (80.4%). It seems likely that the idea that a polar covalent bond possesses some ionic character was the cause of these results.
Crystal. The expected answer for this concept was that it is relevant to IB only. (Of course, crystals do exist in the case of covalent compounds too, but this issue was not included in our lessons.) PuS and US had relatively low frequencies (close to 70%), while both PrS samples had the highest frequencies of correct answers (≈90%).
Molecule. The expected answer for this concept was that it was relevant to CB only. However, the responses from both US (34.7%) and PuS (43.9%) were disappointing. Better results were found in the cases of CG (63.3%) and TG (84.4%).
Metal. The expected answer for this concept was that it was relevant to IB only. However, a high percentage of correct answers was found only for CG (83.7%) and TG (82.6%); lower frequencies of correct answer were found for the other two samples, with IB&CB providing relatively high percentages of choices. (Of course, there are many examples of metals being involved in covalent bonding e.g. aluminium trichloride and a variety of organometallic compounds, but these had not been introduced into the teaching at this stage.)
Nonmetal. The expected answer for this concept was that it was relevant to both IB and CB but low percentages of the correct choice were found, in all cases. This disappointing finding is not easy to explain but the issue clearly needs to be addressed in the teaching.
Electrostatic forces. Table 4b gives the results for the first part of the test for entries about electrostatic forces that are relevant to bond-type and bond formation.
Coulomb forces. The expected answer occurred with a high frequency only for the TG (87.0%), while for all other samples, frequencies were very low. Many students associated these forces with IB only, showing the connection with the attractions between ions. Indeed, the standard text discusses these forces only in connection with the IB. Frequencies of the ∅ option were disturbingly high.
Attractive forces. Here we found high levels of the correct answer, especially for US and TG.
Repulsive forces. Surprisingly, here we find very low percentages of the correct answer, in all cases. The high frequency of the ∅ option is noteworthy. It would appear that most students associate the concept of bonding only with attractive forces, and therefore ignore the repulsive forces (electron–electron and nucleus–nucleus repulsions). Clearly, this finding suggests that this likely misconception needs to be carefully considered when teaching chemical bonding.
It should be noted that both the standard school textbook and our enriched text contain regular references to Coulomb forces/Coulomb interactions, with repulsive forces being mentioned to a lesser extent. Nonetheless, the school textbook does contain an explicit reference to repulsive forces, at one point: “A chemical bond is created when the structural units of matter (atoms, molecules, or ions) come close enough together, so that the attractive forces that are developed between them (e.g., between the nucleus of one atom and the electrons of another atom) overcome the repulsive forces involved (i.e. between the nuclei and between their electrons)” (p. 1). Although the enriched text does not include such a reference, when considering a crystal lattice it is explicitly stated that “oppositely charged ions will attract each other, while ions with charges of the same sign will repel each other, through Coulombic interactions” (p. 7).
Table 4c lists the results for knowledge of other concepts directly relevant to bond formation: energy, valence electrons, the octet rule, electronegativity, polarity, and Lewis structures.
Energy: With the exception of PuS (where only 29.8% of the students chose the correct option and 50.8% chose ∅), frequencies were good with US and CG (≈70%), and the TG achieving the highest percentage of correct answers (82.6%).
Valence electrons: Performance here was moderate, with only about 55% correct answer for PuS and US, but much higher (near 90%) for the two PrS samples.
Octet rule. Results are similar to those for valence electrons (see above).
Electronegativity. Results were poor to moderate, except for TG, who recorded an excellent percentage of correct answers (95.7).
Polarity. Results demonstrated similar trends to those for electronegativity (see above).
Lewis structures. With the exception of PuS, where we found a very low frequency of correct answers (28.4%, with 52.2% opting for the CB only response), the other results were satisfactory, though not high, except for CG (87.8%).
| Knowledge element | Public schools | University | Private school (CG) | Private school (TG) |
|---|---|---|---|---|
| a All entries are relevant for both IB and CB. | ||||
| Group (IB&CB) | 47.0 (28.8 ∅, 15.2 IB) | 33.3 (48.6 ∅, 11.1 N/A) | 75.5 (22.4 ∅) | 82.6 |
| Period (IB&CB) | 41.5 (24.6 ∅, 13.8 IB, 10.8 N/A) | 27.8 (61.1 ∅, 8.3 N/A) | 71.4 (26.5 ∅) | 89.1 |
| Mass number (IB&CB) | 30.8 (30.8 ∅, 20.0 IB) | 25.0 (62.5 ∅) | 34.7 (57.1 ∅) | 87.0 |
| Atomic radius (IB&CB) | 45.4 (25.8 ∅, 16.7 IB) | 38.9 (33.3 ∅, 19.4 CB) | 42.9 (42.9 ∅) | 91.3 |
| Solubility (IB&CB) | 12.1 (47.0 ∅, 24.2 IB) | 18.1 (56.9 ∅, 13.9 IB) | 36.7 (49.0 ∅) | 69.6 (23.9 ∅) |
| Melting point (IB&CB) | 36.4 (28.8 ∅, 16.7 N/A, 12.1 CB) | 26.4 (56.9 ∅, 11.1 N/A) | 79.6 (12.2 IB) | 73.9 (15.2 ∅) |
| Boiling point (IB&CB) | 36.4 (30.3 ∅, 15.2 N/A, 10.6 IB) | 22.2 (56.9 ∅, 16.7 N/A) | 69.4 (12.2 ∅, 10.2 IB) | 73.9 (13.0 CB, 13.0 ∅) |
A similar pattern of results, showing low frequencies of correct answers (and considerable percentages choosing the option ∅), is found, with both PuS and US. The results are similar for the CG in some cases, but those for the TG are generally satisfactory (>69%).
Sentences composed by the students about chemical bonding
The 1st part of the test was completed by asking students to use concepts from the list in Table 1 to compose and write, up to, two sentences for each of the following cases: (i) relating exclusively to the covalent bond, (ii) relating exclusively to the ionic bond, and (iii) relating to both types of bonds. A general comment that arises from inspecting the resulting sentences is that even if the students correctly associated concepts, such as electronegativity, polarity, Coulomb forces, and Lewis structures, with both types of bonding, few of them selected these concepts to write sentences about. This suggests a preference for concepts that required minimal effort or difficulty to understand and use. For this reason, we decided that it would not be useful to explore this part of our data further. Rather we merely provide examples of such sentences (see Table 6), which demonstrate the above issue and employ the terms ‘polarity’ and ‘Lewis structures’. For reasons of plausibility, we show the sentences written by the two students, S21 and S29, from the treatment group of the private school, who obtained the highest overall marks in the SOLO taxonomy (in addition, student S21 obtained the highest cumulative chemistry mark and the highest overall yearly mark in the class).| Student S21 | Student S29 | |
|---|---|---|
| Covalent bond, sentence 1 | The covalent bond forms molecules and not crystals, as does the ionic [bond]. | Covalent bonds are formed between molecules of elements (sic)”. |
| Covalent bond, sentence 2 | Covalent bond forms only between two non-metal elements. | Covalent bonds are of two types: polar and nonpolar. / Covalent bonds are usually formed between non-metal elements. |
| Ionic bond, sentence 1 | The ionic bond consists of a metal and a non-metal element, that is, it is a bond between ions and not [between] neutrally charged atoms. | The ionic bonds form crystals, of which structural units are ions. |
| Ionic bond, sentence 2 | Ionic compounds are extended aggregates [of corpuscles], that is, crystals, and (they) do not consist of molecules, as covalent [compounds] do. | Ionic bonds form usually between one non-metal and one metal elements. |
| Ionic and covalent bond, sentence 1 | Polarity appears/occurs between hetero nuclear atoms with an electronegativity difference, therefore (it) appears/occurs in both the ionic and the (polar) covalent bond. | The octet rule helps in forming ionic and covalent bonds. |
| Ionic and covalent bond, sentence 2 | Lewis structures are used for the description of both the ionic and the covalent [bond]. | Ionic and covalent bonds are written with Lewis structures. |
Comparison of the control and treatment groups from the private school for the 1st part of the assessment instrument
A 2-sample z-test is the appropriate statistical criterion (Spiegel, 1961) to compare two sample populations. The p-values for two-tailed comparisons are calculated and compared to the 95% (p = 0.05) significance level.| Knowledge element | z (p value) | Knowledge element | z (p value) | Knowledge element | z (p value) |
|---|---|---|---|---|---|
| Note: (in this and other tables that follow) S*: statistically significant difference at p < 0.05; S**: statistically significant difference at p < 0.01; NS: statistically not significant. | |||||
| Ion | +2.242 (0.0251) S* | Coulomb forces | +7.697 (<0.01) S** | Octet rule | +0.8955 (0.3681) NS |
| Crystal | +0.2495 (0.8026) NS | Attractive forces | +2.473 (0.0135) S* | Electronegativity | +6.036 (<0.01) S** |
| Molecule | +2.329 (0.0198) S* | Repulsive forces | +2.110 (0.0349) S* | Polarity | +6.674 (<0.01) S** |
| Metal | −0.1432 (0.8887) NS | Energy | +1.501 (0.1336) NS | Lewis structures | −1.728 (0.0836) NS |
| Nonmetal | −1.501 (0.1336) NS | Valence electrons | +0.198 (0.8415) NS | ||
The remaining nine concepts in Table 7 are relevant to the description of both types of bond. Data for the concepts of ‘energy’, ‘valence electrons’, ‘octet rule’, and ‘Lewis structures’, reveal only none statistically significant differences between CG and TG. Results for the concepts of ‘electronegativity’, ‘polarity’, and ‘Coulomb forces’ are of more interest: the TG outperformed the CG, with the TG frequencies exceeding 80% and differences being statistically significant. This finding is consistent with the fact, that, while the enriched text emphasized this property for these concepts, the standard textbook did not. The CG students provided a very low percentage of correct answers for both ‘polarity’ (12.2%) and ‘Coulomb forces’ (8.2%). The majority of CG students (65.3%) associated the concept of ‘polarity’ only with the covalent bond, while in the case of ‘Coulomb forces’ the majority of students (61.2%) associated the concept only with the ionic bond. The fact that the standard text discusses polarity only in connection with the covalent bond and Coulomb forces only in connection with the ionic bond probably explains these findings. On the other hand, ‘attractive forces’ were used to discuss both types of bonds for both the CG and the TG, and percentages of correct answers were high for both groups (71.4% for CG and 91.3% for TG, with the difference being statistically significant at the 0.05 level).
| Entry | z (p value) |
|---|---|
| Group | +0.848 (0.3953) NS |
| Period | +2.154 (0.0316) S* |
| Mass number | +5.201 (<0.01) S** |
| Atomic radius | +4.989 (<0.01) S** |
| Solubility | 3.210 (<0.01) S** |
| Melting point | −0.6582 (0.5092) NS |
| Boiling point | +0.486 (0.6241) NS |
Results and discussion: 2nd part of the assessment instrument
2nd part of the assessment instrument, all samples
Table 9 shows the percentages of correct answers to the questions posed in Table 2, concerning seven items of relevance to chemical bonding, for all samples. We consider the appropriate answer to be ‘correct’ for Statement 2 of question 1: valence electrons can determine the type of bond (ionic or covalent) that occurs between two atoms. However, the correct option for the remaining five statements in question 1 (Table 2) should be ‘wrong’. Electrostatic interactions and polarity occur in BOTH types of bond; electronegativity plays a role in BOTH types of bond; the octet rule DOES NOT provide an explanation for chemical bond formation (ionic or covalent) between two atoms; and Lewis structures are used to represent BOTH covalent and ionic compounds. Average values (plus standard deviations) for performance in the whole of the seven items are provided at the bottom of the table, which allows direct comparison among the four samples. We observe that the treatment group had by far the highest performance. (Note that the latter performance climbs to 92.0% when we ignore performance in the question about the octet rule.) Average values for all samples, for each question, are listed in the final column of the table, to give a general feeling of the difficulty of each question. However, given that the samples themselves were chosen for convenience, and were therefore not representative of the sectors, there is no justification for carrying out any statistical analysis on the data listed in Table 9 to compare our four samples.| Public schools (n = 66) | University students (n = 72) | Private school (CG) (n = 49) | Private school (TG) (n = 46) | Total (n = 233) | |
|---|---|---|---|---|---|
| Electrostatic interactions | 42.4 | 50.0 | 24.5 | 87.0 | 49.8 |
| Valence electrons | 77.3 | 65.3 | 79.6 | 89.1 | 76.4 |
| Octet rule | 7.6 | 11.1 | 8.2 | 2.2 | 7.7 |
| Electronegativity | 43.9 | 55.6 | 71.4 | 91.3 | 62.7 |
| Bond polarity | 36.4 | 40.3 | 18.4 | 91.3 | 44.7 |
| Lewis structures | 30.3 | 83.3 | 100 | 97.9 | 74.7 |
| Criteria for bond-type formation | 54.5 | 91.7 | 89.8 | 95.6 | 81.1 |
| Total (standard deviation in parentheses) | 41.8 (20.6) | 56.7 (17.3) | 56.0 (11.6) | 79.2 (11.9) | 56.8 (20.8) |
Considering performances in the individual questions suggests that the question on the octet rule was by far the most difficult for all four of the samples, while the questions on ‘electrostatic interactions’ and on ‘bond polarity’ also proved difficult. The questions on the ‘criteria for bond-type formation’, the ‘valence electrons’ and the ‘Lewis structures’, on the other hand, seem to have been relatively easy. Specific comments about these findings are given below, with reference, where appropriate, to the contents of the standard and enriched texts. The basic features and differences of the two texts were described in subsection ‘Educational material’, but for a detailed comparison, see Appendix 3 (ESI†).
Electrostatic interactions: The question concerning electrostatic interactions produced poor performances, except for the TG. The standard text does not mention anywhere the term ‘electrostatic interactions’, but it does refer to ‘forces of an electrostatic nature/Coulomb forces’. The enriched text also did NOT refer to ‘electrostatic interactions’, but referred to ‘Coulomb's law’, ‘forces of electrostatic nature (Coulomb forces)’, and to ‘Coulombic interactions’.
Valence electrons: Good performances were found for the first three samples, while the performance by the TG was very good.
Octet rule: Here we found a very small percentage of correct answers for the role of the octet rule, even for the TG. The standard textbook contained relatively little relevant information compared with the enriched text. It is therefore surprising and disappointing that even the students of the TG performed so badly in this question.
Electronegativity: The first three samples had relatively low performances (44, 56, 71% respectively), but the TG performed very well (91%). These findings can be justified by comparing the coverage of electronegativity in the standard and the enriched texts, with the latter giving a more detailed coverage, including the use of a quantitative scale (the Pauling scale), for this property. The enriched text discusses electronegativity with reference to nonpolar covalent bonding, nonpolar and polar covalent bonds, the ionic bond, the continuum between ionic and covalent bonding, and the percentage ionic and covalent character of a bond.
Bond polarity: The findings are similar to those for the electronegativity concept, with the first three samples having low performances (36, 40, and 18% respectively) and the TG giving a very high performance (91%, the same as for electronegativity). It is noteworthy that the standard text does not use the term ‘bond polarity’ or ‘polarity’, while the enriched text refers explicitly to this concept. Both texts, however, refer extensively to nonpolar and polar bonds. Of particular relevance is the following statement in the enriched text in relation to polar covalent and ionic bonds:
“Chemical bonds between atoms of different elements are polar, with a partially ionic and a partially covalent character. If the covalent character is dominant, we have a polar covalent bond, if the ionic character dominates, we have ionic bonding”. (p. 15)
Lewis structures: With the exception of the public schools (30%), all other samples had high performances (83, 100, 98%).
Criteria for bond-type formation: Public schools were again an exception, with a relatively low performance (54%), while all other samples performed well (92, 90, 96%).
Comparison of the control and treatment groups from the private school for the 2nd part of the assessment instrument
Table 10 shows the statistical data (z scores and corresponding p values) for a direct comparison of the percentages of correct answers between the CG and TG from the private school for the 2nd part of the test. It can be seen that the TG produced a statistically significantly superior performance for the questions on electrostatic interactions, electronegativity, and bond polarity, though no significant differences were found in the cases of valence electrons, the octet rule, Lewis structures or criteria for bond formation.| z score (p value) | |
|---|---|
| Electrostatic interactions | 6.116 (<0.01) S** |
| Valence electrons | 1.269 (0.2041) NS |
| Octet rule | 1.305 (0.1902) NS |
| Electronegativity | 2.473 (0.0135) S* |
| Bond polarity | 7.121 (0.01) S** |
| Lewis structures | 1.020 (0.3077) NS |
| Criteria for bond-type | 1.080 (0.2801) NS |
Performances in the 2nd part of the assessment instrument for the reduced sample and the remaining students from the treatment group
Table 11 shows the data for performances in the 2nd part of the test for the ‘Reduced sample’ (n = 33) and the ‘Remaining students’ (n = 13) of the TG (see Section ‘Student samples’). Performance data for the whole TG sample (n = 46) are repeated in the table. Recall that the reduced sample included students with a cumulative mark in school chemistry over 80%. These 33 students had a mean cumulative mark in school chemistry of 90.0% (s.d. 6.0%) and an overall yearly school mean mark of 91.2% (s.d. 5.3%) (see Table 3). The remaining 13 students had a mean cumulative mark in chemistry 65.5% (s.d. 9.0%) and an overall yearly mean mark 75.5% (s.d. 7.4). From Table 11, we observe a difference between the two groups in the crucial/more demanding concepts of ‘electrostatic interactions’, ‘valence electrons’, and ‘electronegativity’, but not in the less demanding topics of ‘Lewis structures’ and ‘criteria for bond formation’. Surprisingly, no difference is detected for ‘bond polarity’. The situation with the ‘octet rule’ is somewhat different (we note again the over confidence of the better-achieving students in the function of this construct). Needless to add, given the small sample in the case of the 'Remaining students', no meaningful statistical comparisons were possible.| Private school (TG) total sample (n = 46) | Private school (TG) reduced samplea (n = 33) | Private school (TG) remaining sampleb (n = 13) | |
|---|---|---|---|
| a Students with cummulative mark in chemistry >80%. b Students with cumulative mark in chemistry <80%. | |||
| Electrostatic interactions | 87.0 | 90.9 | 77.1 |
| Valence electrons | 89.1 | 93.6 | 77.7 |
| Octet rule | 2.2 | 0.0 | 7.8 |
| Electronegativity | 91.3 | 97.0 | 76.8 |
| Bond polarity | 91.3 | 90.9 | 92.3 |
| Lewis structures | 97.9 | 100 | 92.6 |
| Criteria for bond-type formation | 95.6 | 97.0 | 92.0 |
| Total (standard deviation in parentheses) | 79.2 (11.9) | 81.3 (10.0) | 73.8 (14.3) |
Metacognitive questions and students’ opinions about the enriched text
Detailed results for the two metacognitive questions will be presented and discussed in a follow-up publication to this work. The opinions of the students in the treatment group from the private school, about the enriched text are, however, of current interest. Over a third of these students were explicitly critical of the booklet, making negative comments about a number of issues: “Large volume of information”; “too densely written”; “many definitions”; “too difficult”; “everything was difficult”; “a lot of difficult and complicated theory”; “the coverage was very fast, contributing to the conceptual difficulties”, etc. Although there are grounds for believing that understanding of a number of key concepts had been improved as a result of using the enriched text, it will be important to consider these criticisms.Conclusions and the way forward
As our samples themselves were not representative of the sectors that they represent, the results obtained should not be used to infer anything about the types of institution examined in general. With this limitation in mind, let us now consider the answers to the two research questions of this study:Research question (1): To what extent do Greek upper secondary school students share a number of common misunderstandings and misconceptions about chemical bonding?
List of identified misconceptions and conceptual difficulties
• A considerable number of PuS students thought that a ‘chemical bond’ was different from either IB or CB (that is, the adjective ‘chemical’ was thought to differentiate a bond from both IB and CB).
• The idea that a polar covalent bond possesses some ionic character leads to the misconception that ‘ions’ are involved in both IB and CB
• The idea that the concept of ‘molecule’ is relevant to both IB and CB (and not only to CB) is widespread. The relevancy of the concept of ‘metal’ to IB only was not universally accepted.
• Few students recognized that the concept of ‘nonmetal’ was relevant to both types of bond.
• While students appeared to have little difficulty accepting that ‘attractive forces’ were involved in both types of bonding, many students associated Coulomb forces with IB only.
• A majority of students appeared to believe that ‘repulsive forces’ are not relevant to chemical bonding.
• A large majority of students were not only wrong in believing that the ‘octet rule’ could be used to explaining chemical stability and reactivity, but also mistakenly assumed that it could be used to predict whether atoms will form ionic or covalent bonds.
• The direct connection of the concepts of ‘electronegativity’ and ‘bond polarity’ to both IB and CB was not universally recognized, though students in general accepted that ‘Lewis structures’ were of relevance to both types of bond.
• Many students failed to recognize any connection between the concepts of ‘group’, ‘period’, ‘melting point’, ‘boiling point’, and ‘solubility’ to chemical bonding.
In general, weaknesses were evident for a large number of the examined concepts, especially in the case of the public school students (PuS). Only in the cases of ‘criteria for bond formation’, ‘valence electrons’ and ‘Lewis structures’ did the concepts appear to be well understood and even here the performances of the PuS students were limited. There is clearly much room for improvement.
Comparing our findings with those of previous studies, we note that the key feature of student thinking in the UK study by Taber (1998) that a ‘full’ valence electron shells (usually octets of electrons) gives atoms stability was apparent in our samples too. Wang and Barrow (2013) reported a similar finding. Indeed the ‘octet rule’ clearly remains a major source of student misconceptions. Luxford and Bretz (2014) identified misconceptions about periodic trends, electrostatic interactions, and the octet rule in their study. On the other hand, Burrows and Reid Mooring (2015) noted a lack of understanding of electronegativity and misunderstanding of polar covalent bonding. Finally, incorrect predictions regarding boiling points and the polarity of molecules were among the findings reported in the study by Ballester Pérez et al. (2017).
Remarkably, in many cases, the control group (CG) from the PrS showed similar results to those obtained by the university students. This can be justified by the fact that we are considering a prestigious high school, with high-level students and high quality teaching, while the university freshmen represented a specific group with a longer exposure to chemistry over all three upper secondary grades.
Focusing attention on the 1st part of our assessment instrument (the ‘test’), we note that for many PuS students the adjective ‘chemical’ appeared to differentiate a bond from ‘ionic’ and ‘covalent’ bonds. The concept of ionic character for a polar covalent bond seems likely to have led many students to select the option IB&CB. Only the TG students from the private school appeared to appreciate the importance of Coulomb forces to both types of bonding. Many students only associated these forces with ionic bonding, apparently because of the attractions between ions. The standard textbook only discusses these forces in connection with the ionic bond, and this seems likely to be responsible for this finding. Knowledge of repulsive forces, which many students failed to link with chemical bonding, was very disappointing.
In contrast, knowledge of attractive forces, which most respondents correctly associated with both types of bond, was good. Electronegativity, polarity and Lewis structures (the last not by the PrS students), were assumed by many students to be of relevance only to covalent compounds. Finally, when considering knowledge of concepts that are only indirectly relevant to bond formation, we found low frequencies of correct choices (and correspondingly significant percentages opting for the ∅ choice), for group, period, melting point, and boiling point. However, the TG students provided satisfactory percentages of correct answers for all concepts, except solubility.
Research question (2): What is the effect of an enriched educational text, based on the standard Greek chemistry textbook, on preventing the various misunderstandings and misconceptions in the case of high-achieving, tenth-grade students?
The quasi-experimental research design of the 1st part of the test, which sought to test knowledge of many concepts that relate directly to bond formation, detected no statistically significant differences between the control group (CG) and the treatment group (TG) for crystal, metal, energy, valence electrons, the octet rule, and Lewis structures. On the other hand, statistically significant differences in knowledge were detected for the concepts: molecule, non-metal, electronegativity, polarity and types of electrostatic forces (Coulomb, repulsive and attractive forces, although for the final concept performance was very low for all samples).
The findings for the concepts of electronegativity, bond polarity, and Coulomb forces were consistent with the fact that the enriched text places emphasis on the fact that these concepts relate to both types of bonding, while the standard textbook fails to do this. Consequently, the majority of CG students associated polarity only with covalent bonding and Coulomb forces only with ionic bonding. Considering concepts that are only indirectly relevant to bond formation, TG students demonstrated a superior knowledge of the concepts of period, mass number, atomic radius, and solubility, but not of the concepts of group, melting point and boiling point.
Turning to the 2nd part of the test, we note that the question about electrostatic interactions resulted in low performances, except for the TG. Very small percentages of correct answers were found for the role of the octet rule in bond formation, even for the TG, despite the fact that the modified/enriched text clearly indicated the limitations of the octet rule when discussing bonding. It would appear that the ‘octet framework’ is not only being used as a heuristic device, but is also serving a teleological function, with students regularly but incorrectly seeking to use it as an explanatory tool (Talanquer, 2007).
Very high levels of correct answers (91%) were found with the TG for the concepts of electronegativity and bond polarity, with much poorer performances being found for the other three groups. These findings can be justified by comparing the coverage of the concepts in the standard and the enriched texts, with the latter giving a much more extended coverage. Finally, for Lewis structures and criteria for bond-type formation, performances were good except for the PuS students.
In conclusion, the TG produced a statistically significantly superior performance for the questions on electrostatic interactions, electronegativity, and bond polarity, though no significant differences were found in the cases of valence electrons, the octet rule, Lewis structures or criteria for bond formation. The extent of the problem with the octet rule and its resistance to change is a major significant finding of this study: students tend to treat the octet rule as a simple predictive algorithm even when the teaching specifically contradicts this approach, as was the case with the TG.
Final remarks
The diagnostic part of our study leads to the conclusion that many fundamental misconceptions are evident in our samples. The intervention introduced with the treatment group from the private school demonstrated mixed results, with statistically significant differences being detected for various concepts/knowledge elements, but not for others. Interestingly, the treatment group from the private school displayed high levels of performance (usually over 80%) in a large majority of the tested concepts and knowledge elements. On the other hand, pronounced weaknesses were apparent for a few concepts: the importance of non-metallic elements to both types of bonding, repulsive forces, solubility and the octet rule as explanation for chemical bond formation (ionic or covalent).The question we must ask ourselves is, can the role of the modified/enriched text be considered advantageous in dealing with fundamental student misconceptions concerning chemical bonding? While the answer could be yes, it could also be no, especially when we take into account that the intervention was carried out with a prestigious private school, whose students demonstrated high performances in school courses, including chemistry. One might well expect lower performances and higher persistence of misconceptions with less able students.
Understanding the topic of chemical bonding requires the appreciation of many critical details and some sophisticated reasoning, making it complicated and therefore difficult for a large majority of the general student population. In a subsequent publication, we will present data derived from the two metacognitive questions given to the treatment group of students from the private school. These data represent the opinions of students towards the topic as it was presented in the enriched text, and highlight a serious problem, revealing that over one third of the students expressed negative comments about a number of issues (too lengthy; too dense; too complicated; with a lot of details; a lot of definitions; a lot of examples). Content overload is indeed recognized to be a major problem in modern chemical education, leading to curricula that “are too often aggregations of isolated facts” and to the students lacking “a sense of why they should learn the required material” (Gilbert, 2006, pp. 957–958). This data, coupled with the supporting research-based evidence presented here lays a foundation for consideration of a spiral curriculum and of learning progressions in the teaching of chemical bonding. In a subsequent publication, we will look to review earlier studies which made proposals about the teaching of the topic of chemical bonding, and will provide our own proposals, which include defining lower, intermediate, and upper anchors of relevant scientific knowledge.
We would like to finish by considering a number of important issues that were raised by a reviewer of our original manuscript. In the design of the new learning materials, consideration should be given to the following:
• The principles underlining chemical bonding should be stated as basic ideas of our models, rather than as chemical facts. For instance, in the exercises and problems (pp. 10–12 of the standard text – see Appendix 1, ESI†), there is no sense of modelling, with ideas being presented merely as facts. On the other hand, it was recognized that there is occasional reference to modelling in the enriched text.
• Care should be taken to prevent common textbook- and teacher-generated misconceptions. For instance, the standard Greek textbook presents (both in the text and in the exercises), the formation of ionic compounds by electron transfer between single atoms (and this is also followed by the enriched text), leading to the common misconception of representing ionic compounds as molecules. This usual practice of trying to explain the existence of ions is unnecessary as ions exist in nature, so to discuss ionic substances, one just needs to discuss the process of crystallisation from aqueous solution as a result of evaporation of the solvent.
• Finally, caution should be exercised in considering Lewis structures of ionic compounds (such as those exposed in the column on page 10 of the enriched text): one should be comfortable with the notion of Lewis structures of sodium ions, and Lewis structures of chloride ions, but Lewis structures of the compound NaCl may not make sense. Such structures fail to represent the extended structure of ionic compounds, and can only lend themselves to formation of images of single ion-pairs (or molecules).
Conflicts of interest
There are no conflicts to declare.Acknowledgements
We thank the teachers of the schools for their contribution in carrying out this study. For the same reason, we thank Prof. Achilleas Garoufis of the University of Ioannina. Thanks are also due to Mr Evangelos Pyrgas for giving the enriched material the same format as that of the standard material.References
- Atkins P., (1999), Chemistry: the great ideas, Pure Appl. Chem., 71(6), 927–929.
- Atkins P., (2010), Chemistry's core ideas, Chemistry Education in New Zealand, 2010, 8–12.
- Atkins P., (2014), Physical chemistry, a very short introduction, Oxford, UK: Oxford University Press.
- Ballester Pérez J. R., Ballester Pérez M. E., Calatayud M. L., García-Lopera R., Sabater Montesinos J. V. and Trilles Gil E., (2017), Student's misconceptions on chemical bonding: a comparative study between high school and first year university students, Asian Journal of Education and e-Learning, 5(1), Asian Online Journals (http://www.ajouronline.com).
- Barker V. and Millar R., (2000), Students' reasoning about basic chemical thermodynamics and chemical bonding: what changes occur during a context-based post-16 chemistry course? Int. J. Sci. Educ., 22(11), 1171–1200.
- Bergqvist A., Drechsler M., De Jong O. and Chang Rundgren S.-N. C., (2013), Representations of chemical bonding models in school textbooks – help or hindrance for understanding? Chem. Educ. Res. Pract., 14(4), 589–606.
- Biggs J. B., (1999), Teaching for quality learning at university, Open University Press.
- Biggs J. B. and Collins K. F., (1982), Evaluating the quality of learning. The SOLO Taxonomy, New York, Academic Press.
- Birk J. P. and Kurtz M. J., (1999), Effect of experience on retention and elimination of misconceptions about molecular structure and bonding, J. Chem. Educ., 76(1), 124.
- Boo H.-K. and Watson J. R., (2001), Progression in high school students’ (aged 16–18) conceptualizations about chemical reactions in solution, Sci. Educ., 85(5), 568–585.
- Britton B. K., Woodward A. and Binkley M. (ed.), (1993), Learning from textbooks theory and practice, Hillsdale, New Jersey Hove and London: Lawrence Erlbaum Associates.
- Burrows N. L. and Reid Mooring S., (2015), Using concept mapping to uncover students’ knowledge structures of chemical bonding concepts, Chem. Educ. Res. Pract., 16(1), 53–66.
- Butts B. and Smith R., (1987), HSC chemistry students' understanding of the structure and properties of molecular and ionic compounds, Res. Sci. Educ., 17(1), 192–201.
- Cheng M. M. W. and Oon P.-T., (2016), Understanding metallic bonding: structure, process and interaction by Rasch analysis, Int. J. Sci. Educ., 38(12), 1923–1944.
- Coll R. K. and Taylor N., (2001), Alternative conceptions of chemical bonding held by upper secondary and tertiary students, Res. Sci. Tech. Educ., 19(2), 171–191.
- Coll R. K. and Treagust D. F., (2001), Learners’ mental models of chemical bonding, Res. Sci. Educ., 31(3), 357–382.
- Coll R. K. and Treagust D. F., (2002), Exploring tertiary students’ understanding of covalent bonding, Res. Sci. Tech. Educ., 20(2), 241–267.
- Gilbert J. K., (2006), On the nature of “context” in chemical education, Int. J. Sci. Educ., 28(9), 957–976.
- Gillespie R. J., (1997), The great ideas of chemistry, J. Chem. Educ., 74(7), 862–864.
- Harrison A. G. and Treagust D. F., (2000), Learning about atoms, molecules, and chemical bonds: a case study of multiple model use in grade 11 chemistry, Sci. Educ., 84(3), 352–381.
- Hazzi S. and Dumon A., (2014), Conceptual integration of covalent bond models by Algerian students, Chem. Educ. Res. Pract., 15(4), 675–688.
- Holme T. and Murphy K., (2012), The ACS Exams Institute undergraduate chemistry anchoring concepts map I: General chemistry, J. Chem. Educ., 89(6), 721–723.
- Holme T., Luxford C. and Murphy K., (2015), Updating the general chemistry anchoring concepts map, J. Chem. Educ., 92(6), 1115–1116.
- Kronik L., Levy Nahum T., Mamlok-Naaman R. and Hofstein A., (2008), A “bottom-up” framework for teaching chemical bonding, J. Chem. Educ., 85(12), 1680–1685.
- Levy Nahum T., Mamlok-Naaman R., Hofstein A. and Krajcik J., (2007), Developing a new teaching approach for the chemical bonding concept aligned with current scientific and pedagogical knowledge, Sci. Educ., 91(4), 579–603.
- Levy Nahum T., Mamlok-Naaman R., Hofstein A. and Taber K., (2010), Teaching and learning the concept of chemical bonding, Stud. Sci. Educ., 46(2), 179–207.
- Levy Nahum T., Mamlok-Naaman R. and Hofstein A., (2013), Teaching and learning of the chemical bonding concept: problems and some pedagogical issues and recommendations, in Tsaparlis G. and Sevian H. (ed.), Concepts of matter in science education, Dordrecht, the Netherlands: Springer, pp. 373–390.
- Liodakis S., Gakis D., Theodoropoulos D., Theodoropoulos P. and Kallis A., (2012), Chemistry for the 10th-grade general education, (in Greek), Athens, Greece: Greek Ministry of Education and Religious Affairs/Institute of Educational Policy/Diophantos, Institute of Technology, Computers, and Publications, pp. 52–61 and 74–77, http://ebooks.edu.gr/modules/document/file.php/DSGL111/%CE%94%CE%B9%CE%B4%CE%B1%CE%BA%CF%84%CE%B9%CE%BA%CF%8C%20%CE%A0%CE%B1%CE%BA%CE%AD%CF%84%CE%BF/%CE%92%CE%B9%CE%B2%CE%BB%CE%AF%CE%BF%20%CE%9C%CE%B1%CE%B8%CE%B7%CF%84%CE%AE/22-0215-02_Chimeia_A-Lyk_BM.pdf.
- Luxford C. J. and Bretz S. L., (2014), Development of the bonding representations inventory to identify student misconceptions about covalent and ionic bonding representations, J. Chem. Educ., 91(3), 312–320.
- Macintosh H. G. and Morrison R. B., (1969), Objective testing, London: University of London Press.
- Niaz M., (2001), A rational reconstruction of the origin of the covalent bond and its implications for general chemistry textbooks, Int. J. Sci. Educ., 23(6), 623–641.
- Nicoll G., (2001), A report of undergraduates’ bonding misconceptions, Int. J. Sci. Educ., 23(7), 707–730.
- Nimmermark A., Öhrström L., Mårtensson J. and Davidowitz B., (2016), Teaching of chemical bonding: a study of Swedish and South African students’ conceptions of bonding, Chem. Educ. Res. Pract., 17(4), 985–1005.
- Othman J., Treagust D. F. and Chandrasegaran A. L., (2008), An investigation into the relationship between students’ conceptions of the particulate nature of matter and their understanding of chemical bonding, Int. J. Sci. Educ., 30(11), 1531–1550.
- Özmen H., (2004), Some student misconceptions in chemistry: a literature review of chemical bonding, J. Sci. Educ. Technol., 13(2), 147–159.
- Pappa E. T. and Tsaparlis G., (2011), Evaluation of questions in general chemistry textbooks according to the form of the questions and the Question–Answer Relationship (QAR): the case of intra- and intermolecular chemical bonding, Chem. Educ. Res. Pract., 12(2), 262–270.
- Pearson P. D. and Johnson D. D., (1978), Teaching reading comprehension, New York: Holt, Rinehart and Winston.
- Peterson R. F. and Treagust D. F., (1989), Grade-12 students’misconceptions of covalent bonding and structure, J. Chem. Educ., 66(6), 459.
- Peterson R., Treagust D. F. and Garnett P., (1986), Identification of secondary students’ misconceptions of covalent bonding and structure concepts using a diagnostic instrument, Res. Sci. Educ., 16(1), 40–48.
- Peterson R. F., Treagust D. F. and Garnett P., (1989), Development and application of a diagnostic instrument to evaluate grade-11 and -12 students’ concepts of covalent bonding and structure following a course of instruction, J. Res. Sci. Teach., 26(4), 301–314.
- Robinson W. R., (1998), An alternative framework for chemical bonding, J. Chem. Educ., 75(9), 1074–1075.
- Schneider W., (2004), Metacognitive development: educational implications, in International Encyclopedia of the Social & Behavioral Sciences, pp. 9730–9733.
- Spiegel M. R., (1961), Theory and problems of statistics (Schaum's Ourline Series), New York: Mc-Graw-Hill.
- Taber K. S., (1994), Misunderstanding the ionic bond, Educ. Chem., 31(4), 100–103.
- Taber K. S., (1997), Student understanding of ionic bonding: molecular versus electrostatic thinking? Sch. Sci. Rev., 78(285), 85–95.
- Taber K. S., (1998), An alternative conceptual framework from chemistry education, Int. J. Sci. Educ., 20(5), 597–608.
- Taber K. S., (1999), Alternative conceptual frameworks in chemistry, Educ. Chem., 36(5), 135–137.
- Taber K. S., (2002a), Chemical misconceptions – Prevention, diagnosis and cure: Theoretical background, London: Royal Society of Chemistry, vol. 1.
- Taber K. S., (2002b), Chemical misconceptions – Prevention, diagnosis and cure: Classroom resources, London: Royal Society of Chemistry, vol. 2.
- Taber K. S., (2013), A common core to chemical conceptions: learners’ conceptions of chemical stability, change and bonding, in Tsaparlis G. and Sevian H. (ed.), Concepts of matter in science education, Dordrecht, The Netherlands: Springer, pp. 391–418.
- Taber K. S. and Coll R., (2003), Bonding, in Gilbert J. K., Jong O. D., Justy R., Treagust D. F. and Van Driel J. H. (ed.), Chemical education: towards research-based practice, Dordrect: Kluwer, pp. 213–234.
- Taber K. S., Tsaparlis G. and Nakiboĝlu C., (2012), Student conceptions of ionic bonding: patterns of thinking across three European contexts, Int. J. Sci. Educ., 34(18), 2843–2873.
- Talanquer V., (2007), Explanations and teleology in chemistry education, Int. J. Sci. Educ., 29(7), 853–870.
- Talanquer V., (2013), How do students reason about chemical substances and reactions? in Concepts of matter in science education, Tsaparlis G. and Sevian H. (ed.), Dordrechet, The Netherlands: Springer, pp. 331–346.
- Tan K.-C. D. and Treagust D. F., (1999), Evaluating students’ understanding of chemical bonding, Sch. Sci. Rev., 81(294), 75–83.
- Tsaparlis G. and Pappa E. T., (2012), Types of intra- and inter-molecular bonding: the case of general chemistry textbooks, in Bruguière C., Tiberghien A. and Clément P. (ed.), Proceedings of the ESERA 2011 Conference, Lyon, France, http://www.esera.org/publications/esera-conference-proceedings/.
- Ünal S., Çalik M., Ayas A. and Coll R., (2006), A review of chemical bonding studies: needs, aims, methods of exploring students’ conceptions, general knowledge claims and students’ alternative conceptions, Res. Sci. Tech. Educ., 24(2), 141–172.
- Ünal S., Coştu B. and Ayas A., (2010), Secondary school students’ misconceptions of covalent bonding, J. Turk. Sci. Educ., 7(2), 3–29.
- Vladušić R., Bucat R. B. and Ožić M., (2016), Understanding ionic bonding – a scan across the Croatian education system, Chem. Educ. Res. Pract.17(4), 685–699.
- Vrabec M. and Prokŝa M., (2016), Identifying misconceptions related to chemical bonding concepts in the Slovak school system using the bonding representations inventory as a diagnostic tool, J. Chem. Educ., 93(8), 1364–1370.
- Wang C.-Y. and Barrow L. C. R., (2013), Exploring conceptual frameworks of models of atomic structures and periodic variations, chemical bonding, and molecular shape and polarity: a comparison of undergraduate general chemistry students with high and low levels of content knowledge, Chem. Educ. Res. Pract., 14(1), 130–146.
- Yayon M., Mamlok-Naaman R. and Fortus D., (2012), Characterizing and representing students’ conceptual knowledge of chemical bonding, Chem. Educ. Res. Pract., 13(3), 248–267.
Footnotes |
| † Electronic supplementary information (ESI) available. See DOI: 10.1039/c8rp00035b |
| ‡ Anthropomorphism is present in the standard Greek text that was used with the control group of this study (see Appendix 1, ESI†): “Atoms of the other [than noble gas] elements are not “in the same boat”, that is, they do not have an octet of electrons in their outer shell (or a dyad of electrons in the case of the K shell) but try to acquire this structure, that is, to “look like” the noble gases” (p. 2); or “[in covalent bonding], the atoms, while essentially retaining their own electrons, simultaneously enter into an “agreement of co-ownership”, that is, they share common electron pairs between themselves” (p. 7). |
| § 37 answer sheets were returned by the CHE students, but 3 of them were exluded from analysis: two had left blank the 2nd part of the test and one had left the whole test completely blank. |
| ¶ 51 answer sheets were returned by the PrS(CG) students, but 2 of them were excluded from analysis as both had left the first part of the test entirely blank. |
| || The data on school performance for the CG derived from a list provided by the school, which contained a higher number of students (53) than the actual number (49) of students who were included in our study. We have not been able to identify the 4 extra students to exclude them from the analysis. However, this fact does not seriously affect the validity of the current comparison. |
| This journal is © The Royal Society of Chemistry 2018 |