Evaluation of long-term effects of a force-based approach on student understanding of chemical bonding
Catalin
Koro Arvidsson
Sven Erikson's High School, Sven Eriksonsgatan 1, 50180 Borås, Sweden. E-mail: catalin.koro.arvidsson@edu.boras.se
First published on 28th October 2025
Abstract
Chemical bonding is an abstract topic that students often find difficult to grasp. This study examines the long-term retention of knowledge about chemical bonding after instruction using a force-based approach, grounded on Coulomb's law. The study involved 15 students from an upper secondary school in Sweden. Using Bernstein's concept of vertical horizontal and vertical hierarchical discourse, students’ conceptual understanding was examined through semi-structured, task-based interviews one year after the instruction. The findings indicate that students who demonstrated solely, or primarily hierarchical discourse when solving the tasks demonstrated a wide range of cognitive strategies – from rote memorization to deep conceptual reasoning – when explaining how and why chemical bonds form. These results highlight the value of introducing fundamental Coulombic concepts early in the teaching of chemical bonding, as all chemical bonds can be linked to these principles. Regardless of knowledge discourse used during the interview, many students struggled to articulate the concept of electronegativity and its role in explaining bond formation. Interestingly, while covalent bonding was often described using a hierarchical discourse, the students tended to shift to a more horizontal discourse when addressing ionic bonding, frequently relying in the octet rule as the main explanatory tool. This trend underscores the need to place greater emphasis on electrostatic interaction between subatomic particles and ions to foster a force-based understanding of bonding and move beyond oversimplified explanations of covalent and ionic bonds.
Introduction
It is well established that subjects within the natural sciences pose significant challenges for upper secondary students (Bennett and Hogarth, 2009; Anderhag et al., 2016), with chemistry identified as one of the most difficult disciplines (Musengimana et al., 2021). When students are not provided with adequate tools to comprehend science subjects at a sufficiently advanced level during upper secondary education, they may subsequently risk failing in science, engineering, and mathematics programs at the college or university level (Seymour and Hewett, 1997; Chen 2015). Therefore, it is essential to identify specific areas within scientific topics that students find particularly challenging and to develop didactic strategies that promote deeper understanding and improved academic outcomes. This study serves as one example by investigating how an alternative approach to teach chemical bonding can improve students’ comprehension of the topic.The topic of chemical bonding within chemistry is particularly challenging for students to understand (Tsapralis et al., 2019). This difficulty is exacerbated by the presence of alternative concepts, which often arise from the use of oversimplified instructional models to explain bond formation (Hunter et al., 2022). Additionally, the reliance on rote memorization instead of developing higher-order thinking skills and effective study strategies at the upper secondary school level further complicates the learning process (Muteti et al., 2021). Grasping chemical bonding is essential for solving complex, context-based problems in chemistry, such as describing solubility, predicting chemical reactions, and analyzing molecular structures (Broman and Parchmann, 2014). Thus, chemical bonding is considered a foundational aspect of chemistry education.
Traditionally, education about chemical bonding in upper-secondary school relies on a dichotomized view of ionic- and covalent bonds, in which bonding is a result of either electron transfer or electron sharing, respectively (Hurst, 2002; Bergqvist et al., 2013). This dichotomization likely contributes to an overemphasis on the octet framework, promoting the belief that atoms actively “seek” to attain a “full outer shell” or noble gas configuration. Such an interpretation is simplistic and anthropomorphic, often leading students to rely on rote memorization rather than developing a deep conceptual understanding. According to Ausubel's assumptive learning theory, learning is placed in a continuum between rote learning and meaningful learning, spanning between low to high degrees of concept retention (Ausubel, 1968). According to this theory, meaningful learning takes place when students anchor new knowledge to already relevant prior knowledge. On the other hand, concepts that are learnt by rote learning are characterized by limited retention because of little or no incorporation of the new information into the existing knowledge structure. Nevertheless, rote learning is commonly used in school chemistry education, including when teaching chemical bonding (Zoller and Pushkin, 2007; Sevian and Talanquer, 2014).
However, in education about chemical bonding, meaningful learning can be recognized when instruction goes beyond simple presentation of factual knowledge, such as learning ionic and covalent bonds as separate phenomena without a fundamental description of why chemical bonding occurs. To prevent rapid loss of memorized facts learnt by rote learning, chemical bonding can be taught by using a force-based approach, which was first described by Levy Nahum and colleagues (2008). In the force-based approach, chemical bonds are not presented as fundamentally different, as they share the same underlying principles. In specific, chemical bonding occurs due to electrostatic forces between charged subatomic particles and is grounded on Coulomb's law. Accordingly, charged particles interact due to simultaneous attractive and repulsive forces. At a specific interatomic distance, the atoms reach equilibrium, reducing the potential energy of the system and resulting in the formation of chemical bonds (Segal, 1989). In line with Ausubel's assumptive theory (Ausubel, 1968), students can use the force-based approach to anchor new knowledge about various chemical bonds, generating meaningful learning that is not as easily forgotten as dichotomized memorized facts.
Although chemical bonding is an abstract and complex topic traditionally taught through a dichotomous, fact-focused approach (Levy Nahum et al., 2007) that can lead to confusion or eventual forgetting, few studies have examined how well students retain this knowledge over the long term after instructions ends. Long-term case studies by Taber (2003) and Joki and Aksela (2018) suggest that knowledge of chemical bonding can be stored in long-term memory if students develop an understanding of its formation from a Coulombic, force-based perspective, where the octet rule is regarded merely as a “rule of thumb”. However, these studies include only one and eight students, respectively, which highlights the need to test the long-term effects of using the force-based approach as an alternative to traditional education about chemical bonding.
This study expands upon the author's previously published work (Koro Arvidsson, 2025), which investigated the effect of teaching chemical bonding by using a force-based approach at an upper secondary school level and will implement a delayed post-test, which according to Ding and Harskamp (2011) is essential for evaluating long-term retention.
Aim
This study will follow up on the previous research by conducting task-based interviews with students one year after they received instruction on chemical bonding using a force-based approach. The primary aim is to evaluate the extent to which students can apply Coulombic concepts related to chemical bonding to solve new problems, as well as to assess their recall of prior instruction on the topic. Additionally, the study seeks to identify alternative conceptions that emerge one year after instruction, which can inform the development of improved didactic approaches for teaching chemical bonding.The research questions explored in this study are as follows:
1. What knowledge discourses do students use one year after the force-based instructions on chemical bonding?
2. What scientific and alternative conceptions do students demonstrate one year after instruction on this topic?
Theoretical framework
In this study Bernstein's theory of the pedagogical discourse is used as a theoretical framework to analyze students’ content knowledge about chemical bonds (Bernstein, 1999, 2000). The pedagogic discourse describes the communication in the classroom, specifically what and how the knowledge is taught by the teacher and how the students process it. According to Bernstein (2000), the content taught in schools is often characterized by vertical discourse, which means that it is hierarchically organized and comprises formal and explicit language. On the other hand, a horizontal discourse is context-dependent knowledge that is learnt informally, characterized by everyday knowledge that is learnt segmentally. However, the vertical discourse can be divided into a horizontal and a hierarchical part (Bernstein, 1999). If the education is characterized by a vertical hierarchical discourse (or hierarchically organized knowledge) new knowledge builds cumulatively on previous knowledge, leading to increased abstraction. As one example, the force-based approach, described earlier, uses a fundamental theory (Coulomb's law) that can be applied to all chemical bonds. On the other hand, a vertical horizontal discourse (or segmented knowledge within a vertical structure) is still academic and specialized, but not tightly integrated, generating knowledge presented as complementary or contrasting. In addition, the discourse contains segments of less academical language. For instance, segments of horizontal discourse can be recontextualized and embedded within hierarchically structured subjects, either by students (Player-Koro, 2011) or by the teachers seeking to make specialized language more accessible to students (Ferreira and Morais, 2018). This can be applied to the education about chemical bonds by presenting chemical bonds dichotomously, without introducing the underlying force-based phenomena common for all chemical bonds and by using everyday language or anthropomorphic descriptions. However, the knowledge is vertical since the difference between various chemical bonds is compared by using correct, however limited, scientific models, which offers breadth without highlighting underlying patterns leading to bond formation.It is possible to believe that presentation of segmented (vertical horizontal) knowledge about chemical bonding encourages students to learn through rote learning, which includes memorization of divergent facts, without fully understanding their meaning or context. For instance, a student might memorize that an ionic bond is formed between a metal and a non-metal or that the difference in electronegativity between atoms is less than 1.8 but is unaware of why or how ions form chemical bonds. However, a student can possess a “partially memorized understanding”, meaning they can recite a sequence of reasoning without fully grasping it. For instance, a student may know that electronegativity refers to how strongly one atom attracts another atom's electrons, which demonstrates a higher level of understanding than mere rote memorization. Yet, the student may not comprehend that the difference in electronegativity between the two atoms arises from the electrostatic forces between both similarly and oppositely charged subatomic particles. Thus, there is no strict line between memorized facts and meaningful learning (according to Ausubel, 1968) or between a vertical horizontal and a vertical hierarchical discourse (Bernstein, 1999). Moreover, students may initially rely on rote memorization to recall that ionic bonds form between metals and non-metals, this does not preclude a deeper conceptual understanding later during the education. Declarative knowledge can serve as a foundation upon which more sophisticated mental models are constructed, allowing students to later explain the electrostatic principles that underpin chemical bonding. In the present study, the term hierarchical discourse is used as a shorthand for vertical hierarchical discourse, while horizontal discourse refers to vertical horizontal discourse. Within this framework, rote memorization is conceptualized as a component of horizontal discourse.
Methods
Context of the study
This is a qualitative follow-up study that is part of a larger research project investigating the effect of teaching chemical bonding by using a fundamental theory based on Coulomb's law (the force-based approach). The students in the present study are recruited from the first part of the research project, which has already been published (Koro Arvidsson, 2025). This study was carried out in one Swedish upper secondary school in the Technology program. In Sweden, students choose their secondary school program themselves, and the Technology program prepares students for a future career as engineers and/or scientists. Even though many students do not aim for studying chemistry at a university level, it is mandatory to read chemistry in this program. The participating students studied in the second year and were taught about chemical bonding one year before the data to the present study was collected. The long-term effect of a force-based approach when teaching chemical bonding was investigated by using a delayed semi structured task-based interview. At the time of the interview, the students had not received any additional instruction on intramolecular chemical bonding since their lessons a year earlier. However, they had studied physics and other areas of chemistry, primarily on chemical calculations, chemical equilibrium and reaction rate.The lessons about chemical bonding
Ionic and covalent bonds were introduced to the students in four lessons, which are described in the author's previous work (Koro Arvidsson, 2025). This section presents an overview of the lesson content that was presented to the students a year before the delayed task-based interviews were conducted.During the first lesson, Coulombic interactions were introduced as the common foundation of bond formation. The reason behind bond formation was explained as atoms tending to reach a lower energy state. Both the Bohr model and the quantum mechanical model of the atom were introduced.
The second lesson focused on the formation of ionic bonds. The teaching emphasized the attractive forces between oppositely charged ions and explained that ions form lattice structures due to the interplay of attractive and repulsive forces. Consistent with lesson one, the underlying reason for bond formation was described as the atoms’ tendency to reach a lower potential energy state. The octet framework was introduced as one way in which atoms can achieve lower energy during ion and ionic bond formation, but it was emphasized as a role of thumb rather than the fundamental reason for bond formation. When describing ion formation using the octet framework, the process was illustrated with two separate atoms – such as magnesium and oxygen – showing the transfer and acceptance of electrons. This focus on two separate atoms reacting to form an ionic bond represents a simplified model that would not likely occur spontaneously. However, it serves as a useful first step toward understanding Coulombic interactions both between subatomic particles during ion formation and between oppositely charged ions in the formation of ionic bonds. Accordingly, student responses that reflected what had been taught were treated as correct during the written test and delayed interviews.
The third lesson introduced single, double, and triple covalent bonds. The students were asked to find differences and similarities between ionic and covalent bonds, with the aim of avoiding a dichotomous presentation of the bonding types. Both the quantum mechanical and the Bohr models were used to illustrate covalent bonding, and the octet framework was again described as a rule of thumb.
Finally, the fourth lesson focused on the formation of polar covalent bonds. Nonpolar covalent bonds, polar covalent bonds and ionic bonds were presented as part of a continuum, and the Pauling's electronegativity scale was introduced to support this concept.
Student groups
According to the first study in this project, 40 out of 67 participating students at an upper secondary school level used a hierarchical knowledge discourse when describing how and why chemical bonding occurred after being taught the force-based approach when teaching this subject (Koro Arvidsson, 2025). Generally, the students using solely or mostly a hierarchical knowledge discourse achieved a higher final grade in the chemistry course. In this study, students from the previously published work were recruited to investigate their knowledge of chemical bonding one year after instruction. The students were 16–17 years old at the time and studied their second year at the upper secondary school. Based on the students’ answers to a written test post-education, the students were categorized into different student groups. The groups were based on the degree of hierarchical discourse used when answering the test questions a year ago. In Student group 1, students used solely a hierarchical knowledge discourse whereas students in group 4 used solely a horizontal knowledge discourse. In this study, 15 out of the 67 students were recruited to the long-term study. Those recruited were from Student group 1–3 and are further described in Table 1.| Student group based on the answers to the written test | Discourse used during the written test | Students |
|---|---|---|
| Student group 1 | Solely hierarchical discourse | S6, S9, S10, S11, S14 |
| Student group 2 | Predominantly hierarchical discourse | S1, S2, S3, S5, S8 |
| Student group 3 | Predominantly horizontal discourse | S4, S7, S12, S13, S15 |
The 15 students were divided into three groups based on their answers to the written test completed before the present study. Student group 1 included five students who used only a hierarchical knowledge discourse when describing chemical bonding. Student group 2 consisted of five students who used both a horizontal and a hierarchical knowledge discourse but mainly the latter. Student group 3 included five students who mostly used horizontal discourse. Students who exclusively used a horizontal knowledge discourse at the written test were excluded because many failed the chemistry course and likely wouldn't recall enough to provide useful data. All students participating in this study had passed the course.
Design of the task-based interview
The purpose of the task-based interviews was to investigate the students’ knowledge discourses and possible alternative conceptions from their instructions on chemical bonding a year earlier. The task-based interview was designed in a way that the interviewees interacted both with the interviewing researcher (the first author of this paper) and tasks designed for the study (Goldin, 2000). Task-based interviews are commonly used in mathematics education (Maher and Sigley, 2014) but have also been used in chemistry education (e.g.Broman and Parchmann, 2014; Burrows et al., 2021), with the intention to elicit the existing knowledge and ways of reasoning when encountering mathematical or scientific problems, respectively. In this study, the interview was semi-structured, allowing the researcher to ask follow-up questions and explore emergent ideas during the conversation. Thus, the students could be asked to describe a solution to a task further or in another way with the intention to address their conceptual knowledge. In the interviews, the students started the problem-solving process on their own, without interaction with the interviewer. The interviewer then provided hints or asked the student to clarify their answer to encourage higher-level reasoning, which is effective in solving open-ended chemistry problems (Broman and Parchmann, 2014; Broman et al., 2018). For instance, if a student used the octet rule when writing and describing how chemical bonds are formed between two chlorine atoms, the student was asked if there were any forces involved during the bond formation and, if so, if there was any force between any special subatomic particles.The interview questions were designed as mostly open-ended questions, except for questions 16 and 19, which consisted of closed-ended questions with only one correct answer (see appendix). The open-ended questions allowed the students to select from among that person's repertoire of possible responses, which enabled the researcher to get full information about the student's knowledge about a specific topic and the mental strategies to solve a task (Patton, 1987). To make the students comfortable to explain their thoughts about specific questions, the researcher often asked for the student's thoughts about a phenomenon. For instance, by asking “Why do you think chemical bonds are formed?” instead of simply asking “Why do chemical bonds form?”. This prompts the student to form their own explanation, rather than just recall memorized facts. Furthermore, the phrasing is more conversational and less evaluative, creating a safe environment for students to explore ideas and make mistakes since the researcher do not ask for a correct answer.
To help remember the conversations, the interviews were audio recorded. The students were informed that they could choose to decline or withdraw from participating at any time and that all collected data were anonymized. All participating students gave their signed informed consent.
Analysis of the data
To investigate students’ knowledge discourse, the transcripts from the interviews were analyzed using an iterative process involving multiple readings of each transcript. The student responses were subsequently coded by using a thematic analysis (Boyatziz, 1998; Braun and Clarke, 2006). This study is theoretically grounded on Bernstein's (1999) theory of vertical horizontal and vertical hierarchical discourse, with a particular focus on the students’ knowledge one year after conducting the written test. The analytical tool originally developed to identify knowledge discourses at that time was slightly revised to align with the current student responses and is presented in Table 2. For a detailed description of the analytical tool's development, see Koro Arvidsson (2025). The codes derived from the transcripts were organized into five overreaching categories: (1) descriptions of the structure of atoms, molecules and/or ionic bonds; (2) whether ionic and covalent bonds were described dichotomously or through similar force-based explanations; (3) students’ explanations of covalent bond formation; (4) students’ explanations of ionic bond formation, and; (5) whether students conceptualized non-polar covalent bonds, polar covalent bonds, and ionic bonds as a continuum. From these five categories, subcategories were identified by either horizontal or hierarchical, in line with Bernstein's (1999) theoretical model. These subcategories were labelled as horizontal (Ho1–5) or hierarchical (Hi1–5) and are detailed in Table 2. A more comprehensive description of the categories and subcategories is described in the following sections.| Categories | Horizontal discourse | Hierarchical discourse |
|---|---|---|
| 1. Description of the structure of atoms, molecules, and/or ionic bonds | Ho 1 : Describing atoms, molecules, or ionic bonds using the Bohr model | Hi 1 : Describing atoms, molecules, or ionic bonds using Coulomb's law |
| The student explains the structure of atoms, ionic bonds, or molecules by using Bohr's atomic model without explaining the particles’ interaction caused by attractive and repulsive forces. Example of student answers: | The student explains the structure of atoms, ionic bonds, or molecules by using a force-based model. Example of student answers: | |
| 1. Electrons are circulating a nucleus/protons. | 1. Repulsive and attractive forces between particles within an atom keep electrons and the nucleus at a certain distance and/or | |
| 2. Electrons are circulating several nuclei when forming covalent bonds. | 2. Electrostatic attraction between oppositely charged particles contributes to the cohesion of atoms and/or ions. | |
| 3. Ionic bonds form crystals. | ||
| 2. Description of ionic- and covalent bonds as a different (dichotomous) or a similar model | Ho 2 : Dichotomous explanation of ionic- and covalent bonds or explanations without using electrostatic forces of interaction. | Hi 2 : Coulombic explanation for consistency. |
| Example of student answers: | The student describes why both ionic- and covalent bonds are formed with a similar, hierarchical explanatory model by mentioning all the following: | |
| 1. By transfer or sharing of electrons, without being able to describe why this happens. | 1. Ionic and covalent bonds are formed because of electrostatic forces, even if only the attraction force between oppositely charged particles is mentioned when describing the formation of ionic bonds, | |
| 2. When two or more non-metal atoms bond, they form a covalent bond. | 2. the bond formation results in a lower energy level, | |
| 3. When one metal and one non-metal atom bond, they form an ionic bond. | 3. the student can also describe the difference between ionic- and covalent bonds. | |
| 4. Covalent bonds are formed due to attraction between oppositely charged particles (electrons and protons), but oppositely charged ions do not attract. | ||
| 5. Covalent bonds and ionic bonds are formed due to the desire of achieve noble gas structure or to be less stable or to be less reactive. | ||
| 3. Description of the formation of covalent bonds | Ho 3 : Simple description of the formation of covalent bonds | Hi 3 : Coulombic description of the formation of covalent bonds |
| Describes why covalent bonds are formed with a simple description model, without showing any understanding of Coulombic principles. Example of student answers: | The student describes why covalent bonds are formed by connecting the minimum-energy principle, electrostatic forces and the octet rule by describing all the following: | |
| 1. The student describes that the electrons are forcing atoms to stick together. | 1. The student describes how electrostatic forces/attractions between particles result in chemical bond formation, | |
| 2. Atoms share electrons to reach a noble gas structure. | 2. the student also mentions that covalent bond formation results in a lower energy level, | |
| 3. Covalent bonds are formed since the difference in the electronegativity between the atoms is zero. | 3. noble gas formation is one way of reaching a lower energy level. | |
| 4. Description of the formation of ionic bonds | Ho 4 : Simple description of the formation of ionic bonds | Hi 4 : Coulombic description of the formation of ionic bonds |
| The student describes why ionic bonds are formed with a simple description model, without showing any understanding of Coulombic principles. Example of student answers: | The student describes why ionic bonds are formed by connecting the minimum-energy principle, electrostatic forces and the octet rule by describing all of the following: | |
| 1. Ionic bonds are formed because of the transfer or receiving electrons. | 1. The student describes how ionic bonds are formed because of electrostatic attractions between oppositely charged ions; | |
| 2. Ionic bonds are formed because atoms strive to achieve a noble gas structure. | 2. The student also mentions that ionic bond formation results in a lower energy level; | |
| 3. Ionic bonds are formed because oppositely charged ions attract each other. | 3. Noble gas formation is one way of reaching a lower energy level. | |
| 4. When one metal and one non-metal atom bond they form an ionic bond. | ||
| 5. Description of chemical bonds as a continuum scale | Ho 5 : Chemical bonds are not described as a continuum scale | Hi 5 : Chemical bonds are described as a continuum scale |
| The student knows that different atoms have different electronegativity but cannot describe the connection between electronegativity and the formation of nonpolar covalent bonds, polar covalent bonds and ionic bonds as a continuum scale. | The student can describe the connection between electronegativity and the formation of nonpolar covalent bonds, polar covalent bonds and ionic bonds as a continuum scale. |
The interviews were conducted in Swedish, the students’ native language. Following the thematic analysis described above, the transcripts were translated into English by the first author. Translation between languages always present challenges, as certain nuances of the original interviews may be lost. To mitigate this risk, the translations were reviewed multiple times and, in some cases, back translated to ensure that no major linguistic inaccuracies remained.
Results and discussion
The students’ knowledge discourses one year after instruction
This section addresses the research question “what knowledge discourses do students use one year after the force-based instructions on chemical bonding?”. The students’ responses to the written test, performed after the lessons about chemical bonding, were compared with the students’ responses to the delayed task-based interview. However, even though the delayed interviews provide valuable data on the students’ long-term retention, it is important to note that the results between the written test and the interviews are not directly comparable, since the method when collecting data differed. Hence, it is not clear if the changes observed between the written test and the interview are a result of knowledge-retention over time or the change of methodology.Students’ interview responses about chemical bonding were analyzed using a theoretical thematic analysis as described previously. The categories identified during the analysis were (1) the structure of atoms, molecules and/or ionic bonds; (2) description of ionic- and covalent bonds as a different (dichotomous) or similar model; (3) description of the formation of covalent bonds; (4) description of the formation of ionic bonds; (5) description of chemical bonds as a continuum scale. Each category was further divided into horizontal or hierarchical knowledge discourses (Table 2). Based on the prevalence and frequency of each subcategory identified in the students' responses during the task-based interviews, the descriptions provided by the students were classified into four response groups called Response groups A–D. A summary of interesting and recurrent themes in the students’ answers to the delayed task-based interview is outlined in the appendix (Tables 4–6). In response group A, the students’ descriptions involved all the hierarchical subcategories (Hi1–Hi5, Table 2) when describing their solutions to the tasks during the task-based interview. In Response group B, the students used 3–4 subcategories of hierarchical description when solving the tasks (Table 2). In Response group C students’ description involved 3–4 subcategories of a horizontal discourse when describing the solutions of the tasks, whereas the students in Response group D showed solely a horizontal discourse (Ho1–Ho5, Table 2) when solving the tasks during the interview.
The students’ knowledge discourses used at the written test and the delayed task-based interview are illustrated in Fig. 1. The results of this study indicate that out of five students, two from Student group 1, who solely used hierarchical discourse during the written test demonstrated a decrease in hierarchical discourse by the delayed task-based interview. Three students in this group used solely a hierarchical discourse in both the written test and the delayed task-based interview. In Student group 2, two students displayed an increase in hierarchical knowledge at the interview compared to their performance on the written test one year prior. The remaining students in this group used either the same level of discourse (one student) or a decrease in hierarchical discourse (two students) at the interview. In Student group 3, one student utilized hierarchical discourse to a greater extent during the task-based interview compared to the written test performed a year earlier. However, the other students in this group either showed a decline or no change in their discourse during the interview.
 | ||
| Fig. 1 The knowledge discourse used by each student between the written test, completed one year ago and the task-based interview performed one year post instruction. The categorization of students in Student group 1, 2, and 3 is based in the student answers to the written test. The categorization of Response group A–D is based on the student answers to the delayed task-based interview. | ||
Interestingly, some students improved their hierarchical discourse one year after the learning. According to the data previously published by the author (Koro Arvidsson, 2025), which describes the categorization of the participating students into Student group 1–3, several students in Student group 2 used all the subcategories of a hierarchical discourse (described in Table 2) when solving the written tasks. However, students that used an anthropomorphic language when describing bond formation were ultimately placed in Student group 2 instead of Student group 1. According to unpublished data, the students enrolled in this study were placed in Student group 2 one year ago due to their description of that atom's “desire” to receive lower energy. The usage of anthropomorphic expressions may be interpreted as a horizontal discourse due to the simplified “everyday” language used. It is possible that an anthropomorphic view of bond formation may pose problems for student comprehension. Specifically, if a student believes that an atom's desire is the driving force behind bond formation, they may not seek other levels of explanation. However, anthropomorphic language may also function as useful metaphors in aiding communication and understanding (Taber and Watts, 1996). As illustrated by the following quote, the students in this study were allowed to describe whether they really believed that atoms have a “desire” or other human feelings during the task-based interview.
Student (S) 5: The atoms want to reach lower energy.
Researcher (R): Do they really want that, in the same way as humans want things?
S5: No! But they must… Lower energy is a result when they are attracted by each other.
Consequently, all students in Student group 2 showed that the usage of anthropomorphic language did not pose any problems for understanding the force-based (hierarchical) approach of chemical bonding. Given this, it is reasonable to assume that S5 and S8 used the same level of hierarchical knowledge one year post-instruction of this subject and that S1, S2, and S3 used a reduced level of hierarchical discourse.
Also, S7 in Student group 3 demonstrated a higher level of hierarchical discourse during the delayed task-based interview when compared to the written test. This improvement may indicate that the student gained a better understanding of chemical bonding after completing the written test. Another possible explanation is that oral examinations might be more beneficial for some students than written tests. A study by Burrows et al. (2021) suggests that upper-secondary students often exhibit higher-order thinking—beyond mere fact recall—during oral, task-based interviews in chemistry. These interviews can also help students to further develop their conceptual frameworks. Additionally, scaffolding techniques, such as providing hints during interviews, may influence students' responses (Broman and Parchmann, 2014). Therefore, it is possible that S7 could articulate the knowledge more effectively during the interview than in the written test.
Hierarchical knowledge discourse entails an understanding of multiple advanced scientific concepts of a phenomenon. Therefore, it is not surprising that students may forget one or more of these aspects when solving chemical tasks over time. To examine the impact of a hierarchical knowledge discourse on the long-term understanding of chemical bond formation, it is essential to determine which specific aspects students retain. It is equally important to determine which aspects they are more likely to forget from their instructions and if the students use alternative conceptions when explaining why chemical bonds occur. This will be addressed in the following section.
Students’ scientific and alternative conceptions of chemical bonding one year after instruction
This section addresses the question “What scientific and alternative conceptions do students demonstrate one year after instruction on this topic?”. In addition, the alternative conceptions are contrasted with accepted scientific conceptions used by the students. Hopefully, the answers to this question will indicate specific areas that teachers should pay attention to when planning and teaching lessons concerning chemical bonding.The data in the appendix (Tables 4–6) summarize the themes of the students’ answers in the Response groups A, B, C, and D. The students’ answers mirror the components of the subcategories in Table 2. However, the components are presented separately to illustrate which specific parts the student used when solving the tasks. As one example, the constituent components within subcategory Hi3 and Hi4 are separated into the following two descriptions: that bond formation results in a lowering of potential energy and that chemical bond formation occurs because of electrostatic forces.
Response group A
Response group A consists of five students who used all subcategories of hierarchical knowledge discourse described in Table 2 when solving the tasks during the interview. The following quotes illustrates how the students in this group utilize rote memorization, a horizontal knowledge discourse and a hierarchical knowledge discourse to describe the formation of chemical bonds. Firstly, S5 used rote memorization to recall what kinds of elements form ionic bonds, which is illustrated below:R: This is a substance called magnesium oxide. Can you please draw and explain how it is formed?
S5: Yes, a metal and a non-metal form an ionic bond.
Second, S5 used a horizontal discourse when describing how ionic bonds are formed, by the transmission and reception of electrons to get a “full” valence shell. The student also mentioned that the two orbitals in ionic bonds are separate, which indicates that the student is familiar with the quantum mechanical model of the atoms and can apply it to the description of bond formation. The student's models and schematic representation of magnesium oxide are presented in Fig. 2.
 | ||
| Fig. 2 Student drawings of magnesium oxide: (A) a simplified Bohr model of oxygen and magnesium, in which the valence electrons of the magnesium atom have been transferred to the oxygen atom, (B) attraction between a magnesium ion and an oxide ion, and (C) schematic representation of ionic bond formation between Mg2+ and O2−. | ||
R: Yes. So, how is an ionic bond formed?
S5: (The student draws two simplified Bohr models of oxygen and magnesium, showing only the valence electrons, Fig. 2A. The figure illustrates how the valence electrons of magnesium are transferred to oxygen). And that one will take electrons (points at the drawn oxygen atom).
R: Yes! So, which one will receive electrons of these two [atoms]?
S5: Ehm… the oxygen.
R: Yes. And which one will transmit electrons?
S5: The magnesium.
R: And what happens then?
S5: Ehm… there will be two ions that will connect and form an uncharged particle (the student draws Fig. 2B.
R: Yes. And you said something about the atomic shells before. About the electrons in the shells. What happens there?
S5: The electrons will be in separate orbitals.
R: Yes. How many valence electrons does the oxygen atom have then?
S5: It will get eight, and the magnesium will have ten since it loses two.
R: That is right.
S5: So, they will have full valence shells.
Finally, the student used a hierarchical knowledge discourse when describing why ionic bonds are formed. The description emphasized that ionic bonds form to reach a lower energy state, resulting from the transfer of electrons driven by differences in the attraction forces between the two atoms. Also, by the electrostatic attraction between oppositely charged ions.
R: What is the underlying reason for the atoms [oxygen and magnesium] to bond?
S5: Because it will lower their energy. Or, like, first they will raise their energy… or the oxygen will get a full outer shell which lowers the energy. But it will cost energy to transfer an electron, which means that magnesium will increase its energy at first, but when they [magnesium and oxygen] are bond together the energy will be lower in total. That is why magnesium is transferring an electron, to get lower energy as a result.
R: Are there any forces leading to the formation of magnesium oxide?
S5: Yes… opposite charges attract.
R: Yes. Do you mean that the ions attract?
S5: Yes, they attract each other (point at Fig. 2B).
R: Why does magnesium and oxygen form an ionic bond and not a covalent bond?
S5: Because the difference between the atoms’ electronegativity is huge. A covalent bond is formed when the difference between the electronegativity of two atoms is less than 0.8, or whatever number it was. But here (points on the drawn model of magnesium oxide) the difference between the atoms’ electronegativity is more than 0.8, or whatever it was, since they are placed on opposite sides of the periodic table. According to this (points at the Pauling scale) the electronegativity of magnesium is 1.2 and the electronegativity of oxygen is 3.5.
The student did not mention the correct limit value (which is approximately 1.8) when using the Pauling scale of electronegativity to describe the difference between covalent bonds and ionic bonds. However, it is important to note that this value only serves as a “rule of thumb” and cannot be used to predict all chemical bonds. Understanding electronegativity requires a hierarchical knowledge structure, whereas simply recalling a memorized number relies on rote memorization, which represents a horizontal knowledge structure. Still, the quote above does not demonstrate that S5 understands what electronegativity is and, consequently, why covalent or ionic bonds form. However, the student refers to this concept elsewhere in the interview – for instance, when explaining why hydrogen fluoride contains polar covalent bonds with the following words:
S5: It is a polar covalent bond. If you look at this (points at the Pauling scale) you can see that fluor has stronger electronegativity and will pull the electron pair to a higher degree than hydrogen.
Research by Taber and Watts (2000) and Levy Nahum et al. (2007) indicates that many students tend to rely on rote memorization when learning about chemical bonding, which may contribute to the development of misconceptions and confusion regarding the formation of different types of chemical bonds. Even though there were no such confusions among the students in Response group A, one of the students (S6) had forgotten the name covalent bond. However, the student was able to use a hierarchical knowledge discourse when describing the formation of covalent bonds, e.g. when describing hydrogen gas, which illustrates that the student retains deep conceptual knowledge about bond formation.
S6: Two hydrogen atoms form hydrogen gas by sharing electrons.
R: Right.
S6: The two atoms attract until they reach a specific distance, when they share electrons.
R: How is hydrogen gas formed in detail?
S6: It is because… the nuclei are seeking after the electrons.
R: Yes, you may say that.
S6: It is adhesion forces and repulsive forces here (points at the model of the hydrogen-gas molecule).
R: Yes.
S6: And they have a name… electro…
R: Electrostatic forces?
S6: Yes, electrostatic forces!
Furthermore, the student used the same discourse when describing the formation of chlorine gas:
S6: The adhesion force between the nuclei and the electrons makes the atoms come closer and when they are close enough repulsive forces make them stay at a certain distance.
And furthermore, when describing the formation of all gas molecules:
S6: The two nuclei repel each other when they come to close. The atoms stay at this distance, where the energy is lowest.
Another student (S8) in response group A used all hierarchical knowledge structures according to Table 2 when describing chemical bonding. Interestingly, at one point the student used a simplified description of bond formation when explaining why covalent bonds are formed between some atoms and not between others.
R: Why is it an ionic bond between magnesium and oxygen and a covalent bond between the two atoms in chlorine gas?
S8: Well, I guess that if magnesium has two [electrons] in the outer shell and oxygen has six [electrons], then it won’t work to share pairs of electrons between them. Because magnesium needs six more electrons to fill the outer shell and then it is easier to transfer two [electrons] instead of sharing six [electrons] with another atom. If it transfers two [electrons] it will become more stable, and when oxygen receives the electrons, it will become more stable too. Then, the particles [ions] are charged and lose even more energy when they bond.
R: So, you say that depending on how many electrons the atoms have, different types of bonds are formed depending on whether it is easiest to share electrons or transfer and receive electrons?
S8: Yes.
The core idea of the student's answer was that the “ease” of electron transfer versus sharing influences bond type, which reflects a heuristic rather than scientific explanation of bond formation. Although this specific answer lacks information about electrostatic forces, proximity and energy in bond formation, the student used these scientific concepts when solving other tasks during the interview (see Tables 4–6 in the appendix).
In conclusion, all students in response group A used the force-based approach when describing why ionic- and covalent bonds are formed one year after learning a hierarchical knowledge structure about chemical bonding. The students showed a wide range of discourses to explain chemical bonding, from rote memorization to a hierarchical knowledge structure. Rote memorization and other heuristic explanations were used by the students as mental tools when solving chemistry problems, even though the students had understood the underlying theory behind the memorized facts.
Response group B
Response group B consists of three students that used primarily a hierarchical knowledge discourse (Table 2) when solving the tasks during the delayed task-based interview. Interestingly, none of the students in this group correctly described covalent bonds and ionic bonds as a continuous scale, e.g. by using Pauling's scale of electronegativity. As an example, one student (S7) was asked to select the correct model of hydrogen oxide from three different options (see Fig. 5 in the appendix).S7: The last one.
R: Why do you think it is the last option?
S7: Because it shows the electrons in the middle. It just looks better.
R: Do you remember the word electrostatic forces?
S7: Yes, I have heard the word. I guess it has something to do with electronegativity
R: Yes.
S7: But I do not remember what kind of force it is.
R: OK. But what is electronegativity?
S7: It measures the ability of the atoms to react. The higher electronegativity the higher…
R: … the higher reactivity?
S7: Yes, exactly.
Even if S7 described electronegativity as the atom's ability to react (which may be interpreted as rote memorization) the student did not solve the above task by using the Pauling scale, possibly because the student lacked full understanding of the concept electronegativity (i.e. the atoms tendency to attract electrons when forming chemical bonds). Even though S7 did not remember the concept of electrostatic force, the student explained why ionic- and covalent bonds are formed by correctly explaining the interaction between subatomic particles (see Tables 4–6 in the appendix).
Another student in this group (S1) expressed an alternative conception of electronegativity and electrostatic force between subatomic particles during bond formation.
S1: Magnesium wants to transfer two electrons to get a full outer shell. A full other shell means noble gas configuration, which is a stable state.
R: Right.
S1: Yes. And it is the same with the oxygen [atom]. But instead of transferring electrons it will receive the electrons transferred from magnesium and it will get noble gas configuration, which is a stable state and lower energy.
R: Right, is there any force involved here?
S1: (S1 study the Pauling scale) OK… so… we calculate the delta between magnesium and oxygen. Or was it delta?
R: Yes, delta means the difference.
S1: Yes, the difference between magnesium and oxygen. That means 3.5–1.2, which is 2.3, right?
R: Right.
S1: So, oxygen has a higher electronegativity when compared to magnesium, which means that it will attract in a more powerful way.
R: What do you mean it will attract?
S1: To magnesium. It [the oxygen atom] will attract the most.
R: Yes, but what particles will it attract? Is it the whole magnesium atom or some part of it?
S1: Ehm… it is… I think that the oxygen's electrons in the outer shell will attract the nucleus of the magnesium.
In contrast to S7 and S9 in Response group B, S1 remembered that electronegativity describes the ability of subatomic particles to attract other (oppositely charged) subatomic particles. However, instead of describing the attractive force between the oxygen atom's nucleus and the magnesium atom's electrons – which ultimately leads to the formation of ions – the student referred to electronegativity as an attractive force between the oxygen's electrons and magnesium's nucleus. It is unclear whether the student believe that this force holds magnesium and oxygen together or if the student describes the formation of ions in an alternative way. It is, however, evident that S1 is uncertain about how electronegativity is connected to ionic bond formation. As reported by others (e.g., Luxford and Bretz, 2013), it is concerning when students do not recall the usage and theoretical foundation of the Pauling scale, despite its importance in illustrating how all chemical bonds arise from the same fundamental forces. This lack of understanding may contribute to a dichotomous perception of covalent and ionic bonds (Taber, 1998).
As stated above, S1 focused heavily on the octet framework when explaining the formation of ionic bonds. The student described how ions are created through the transfer of electrons from a metal to a non-metal, resulting in noble gas configuration and lower potential energy. However, at this stage, the student did not mention the electrostatic force between the oppositely charged ions. In contrast, the student suggested that the ions no longer interact after achieving the noble gas configuration.
T: So, you said that magnesium transfer two electrons to oxygen when forming magnesium oxide. What happens then?
S1: Oxygen becomes an ion, so it will get two minus.
R: Yes, and what about that one (points at the Bohr model drawn by the student, Fig. 3)?
 | ||
| Fig. 3 A student drawing of a Bohr model showing the formation of a magnesium ion and an oxygen ion. | ||
S1: This will be Mg 2+.
R: So, we will have two ions, one is positive, and one is negative.
S1: Right. Two ions, one is positively charged, and one is negatively charged. I don’t think… I don’t know if I remember, but I don’t think that they will bond. I think that it will just transmit electrons… but then, they have reached a certain state, and they want to go in different directions.
This inadequate explanation of bond formation is also seen in the student group using mostly hierarchical knowledge discourse (i.e. Response group C) during the task-based interview and will be discussed later.
Taken together, none of the students in Response group B correctly described covalent bonds and ionic bonds as a continuum by correctly using the Pauling's scale, which indicates that these students do not fully understand how the same underlying forces generate different chemical bonds. This suggests that significant emphasis should be placed on the concept of electronegativity and its crucial role in bond formation during the teaching of this topic.
Response group C
Response group C consists of five students who used primarily a horizontal knowledge discourse (Table 2) when describing chemical bonding during the delayed interview. Three out of five students in this group failed to mention that covalent and ionic bonds are formed due to the natural tendency of systems to move towards lower potential energy. Instead, the students mentioned that atoms strive to reach stability and lower reactivity, which forces them to form chemical bonds.R: Do you remember why two hydrogen atoms form a bond?
S11: Ehm… to make them more stable.
R: Yes.
S11: To get a full outer shell. That is what they strive for.
R: Yes.
S11: Then they will be more stable.
R: All right.
S11: If they have a full outer shell, they are not prone to react anymore.
R: OK, good. Does this lead to something else, when they are stable?
S11: They won’t react. I mean, when they are stable, they don’t want to react since… they don’t have any need to form something new or something else. They don’t have the need to react.
S2 and S3 in the same Response group gave similar explanations as S11. One example is when S2 described why all chemical bonds are formed in general:
S2: It is because the atoms in nature are seeking for noble gas configuration (…) That is the driving force leading to bond formation.
R: Right. So, you mean that it is the reaching of noble gas configuration that is the driving force? Or is there anything else?
S2: Yes, generally atoms want to be stable. They do not want to be reactive. They want to reach as low reactivity as possible and one way to do so is to form noble gas configuration.
The instructions provided to the students in this study primarily focused on the tendency of physical systems to achieve lower energy states when teaching about chemical bonding, which have been published in the author's previous study (Koro Arvidsson, 2025). However, students in response group C often described atoms as “seeking stability,” which demonstrates a horizontal knowledge discourse. This perspective is problematic, as the concept of stability does not consistently align with a low-energy state. For instance, adenosine triphosphate (ATP) is considered a relatively stable molecule despite its high-energy phosphate bonds. Such discrepancies highlight the limitations of using “stability” as an explanatory framework. Atoms do not seek stability in a purposive sense; rather, chemical bonding occurs because it results in a more energetically favorable configuration. Thus, emphasizing the principle of energy minimization – rather than relying on vague notions of stability – offers a more accurate and conceptually robust understanding of chemical bonding.
S2 and S3 stood out from the rest of the students in Response group C by applying Coulomb's law when describing the formation of both ionic and covalent bonds (see Tables 4 and 5 in the appendix). However, since they did not mention the lowering of energy during bond formation, which is the fundamental reason why all chemical bonds are formed, their explanations were not seen as hierarchical (Hi3 and Hi4, Table 2). In contrast to the other students in this Response group, S2 and S3 did not confuse the names of different chemical bonds and were able to provide a correct, though not fully hierarchical, description of covalent and ionic bonds. It is possible that recalling Coulomb's law supported their understanding by offering a unifying conceptual framework. When students can integrate new knowledge into a coherent mental model, they are less reliant on memorize disconnected facts (Ausubel, 1968). This may highlight the pedagogical value of introducing Coulomb's law when teaching chemical bonding.
In contrast, S11, S12 and S15 described the formation of ionic bonds by strongly emphasizing the octet framework instead of describing electrostatic attraction and/or repulsion as the driving force behind bond formation. Interestingly, two of these students (S11 and S12) did not mention that two oppositely charged ions attract. One example is when S12 describes the formation of magnesium oxide. The student draws a Bohr model showing a magnesium atom transferring two electrons to an oxygen atom but does not explain how the resulting ions attract each other.
R: Do you remember if there is a bond between magnesium and oxygen?
S12: It is an ionic bond, isn’t it?
R: It is. Why do you think that it is an ionic bond?
S12: Because magnesium is a metal, and oxygen is not a metal.
R: Yes, a metal and a non-metal form an ionic bond. Absolutely. What else is typical of ionic bonds?
S12: Salt is an ionic bond. Salt is NaCl, and sodium is a metal.
R: Right. Do you think that there is a force between magnesium and oxygen?
S12: They [the magnesium atom's valence electrons] are forced to the oxygen (points at the drawn model).
R: Yes, the magnesium atom's electrons are attracted to the oxygen atom.
S12: Yes, and electrons are also attracted to the opposite atom in chlorine gas (points at a drawn Bohr model of chlorine gas). But not to that degree that the electrons jump to the other atom. Instead, they share [electrons].
R: Yes, good! Do you remember the Pauling scale?
S12: Yes, but I do not remember how to use it.
R: That is OK. Why do you think another bond is formed here (points at the model of chlorine gas) as compared to here (points at the model of magnesium oxide)?
S12: Because it [magnesium oxide] is a salt.
R: Yes.
S12: So, it is a crystal. And here (points at the model of chlorine gas) there is something else.
S12 relied on rote memorization when explaining the properties of ionic bonds, stating that they involve a metal and a non-metal and that they form salts and crystals. However, the student failed to describe the attractive forces that bring the oppositely charged ions together. As a result, S12 only explained the creation of ions using the octet rule. This was also evident in S11's explanation of sodium chloride formation, which similarly relied on the octet framework without addressing the bonding between ions.
R: What happens then, after sodium has transferred an electron to chlorine?
S11: They will get a full other shell. Since they have eight electrons, they are stable, and the shell is full.
R: Right, and what happens then?
S11: Then they do not want to react anymore.
R: No.
S11: And then…. I guess that they will repel each other.
The above descriptions of ionic bond formation given by S11 and S12 (and by S1 in Response group B) provide examples of when the use of the octet framework turns into an obstacle for reaching full understanding of chemical bonding. If the octet framework serves as the underlying reason why chemical bonds form, and the student knows that ionic bonds are formed because of the process of transferring and receiving electrons, then it would be logical that the atoms are “satisfied” when the octet formation is completed, and no bonding will occur.
Interestingly, attraction between oppositely charged particles was more often used by the students in Response group C when describing covalent bonds than ionic bonds (see Tables 4 and 5 in the appendix). The difficulty of understanding electrostatic forces between ions among students has been reported by others (Taber, 1997, 1998; Taber and Coll, 2002; Bowe et al., 2022). Moreover, several studies have found that students use electron transfer and receiving as a sufficient description of ionic bond formation, rather than describing ionic bond formation because of electrostatic forces between oppositely charged ions (Doymus, 2008; Luxford and Bretz., 2014; Prodjosantoso et al., 2019). It has been suggested that this alternative conception may result from textbooks presenting ionic and covalent bonds dichotomously, without focusing on the force-based approach that leads to the formation of all bonds (Bergqvist et al., 2013; Bergqvist et al., 2016). However, even though the students in this study were introduced to the force-based approach when teaching chemical bonding, the students tended to not mention this fundamental aspect of bond formation and did only apply it to covalent bonds.
Taken together, the students in Response group C primarily used a horizontal discourse when describing both covalent and ionic bonds, placing strong emphasis on the octet framework as an explanatory model for bond formation. Furthermore, most of the students did not use the force-based explanation of bond formation regarding the description of ionic bonds. In addition, several students in this group mentioned that atoms strive to reach stability during bond formation. Thus, the results suggests that teaching chemical bonding without relying on stability or lower reactivity as explanatory models may be more effective, as these concepts do not adequately account for the underlying reason why chemical bonds occur.
Response group D
Response group D consists of two students (S4 and S13) who solely used a horizontal knowledge discourse (Table 2) when describing chemical bonds during the delayed task-based interview. At the written test, both students used a hierarchical discourse to a small degree, by being able to describe that all chemical bonds are formed because of electrostatic forces between oppositely charged particles. However, the rest of their understanding relied on a horizontal knowledge discourse (data not shown). During the delayed interview, it became evident that these students had reduced their previously limited hierarchical discourse to a purely horizontal one. During the interview, S13 described noble gas configuration as the sole reason for chemical bond formation and did not identify any underlying principles – such as Coulomb's law or the tendency of atoms to reach a lower potential energy state – as common factors in bonding processes. Like S8 in response group A, S13 described that the number of valence electrons in two reacting atoms decides the nature of the bond. Specifically, the bond formed depends on what is most “convenient” to form.R: Do chlorine gas and magnesium oxide consist of the same or different bonding?
S13: They are different.
R: Why are they different?
S13: Because the chlorine atoms… one of the atoms cannot give [electrons] to the other [chlorine atom]. So, they need to share [electrons] otherwise it will not work.
R: Right.
S13: But here (points at the drawn model of magnesium oxide), they [the atoms] cannot share electrons to make it work. The electrons are transferred.
R: Right, you mean that the atoms need to get eight electrons in the outer shell to form a bond, and the only way of doing so is to share electrons here (point at the drawn model of chlorine gas) or transfer and receive electrons here (points at the drawn model of magnesium oxide)?
S13: Yes.
As illustrated above, S13's response is an example of an incomplete horizontal explanation of why both ionic and covalent bonds are formed. The following quote further illustrates the student's description of the formation of magnesium oxide, followed by a description of the formation of chlorine gas:
S13: The two outer electrons in magnesium will go to oxygen. Then, oxygen will get a full outer shell.
R: Yes. What happens next?
S13: Ehm… then both magnesium and oxygen get full outer shells, and it will be a… Mg 2−
R: Yes?
S13: And oxygen will be O 2+
R: Yes, and what happens next?
S13: Isn’t there a reaction?
R: It is. And what does that mean?
S13: It means… (long silence)
R: Will they [the ions] go in different directions or will they stay here, positive and negative…?
S13: Oh yes… right… some atoms share, and some atoms transfer, but then I don’t know what happens.
R: OK. But do you think the reaction is done after transmitting and receiving electrons?
S13: No. I don’t remember.
Similarly to S1 in Response group B as well as S11, and S12 in Response group C, also S13 used the octet framework when describing the formation of ions (even though the concept “ion” is not mentioned and the resulting ions are given incorrect charge). Furthermore, S13 did not describe that oppositely charged ions attract when forming an ionic bond. The uncertainty about charges and their interaction is also evident when S13 described a model of hydrogen gas.
S13: The [hydrogen] atoms attract and form H 2 . It is the protons that attract.
R: Do you mean that the protons from the two hydrogen atoms attract?
S13: Yes, because they are positive.
R: And what are the dots around the nucleus?
S13: That is the electrons that go around. But it is the nuclei that attract.
The misconception regarding the attraction of two positively charged nuclei has been described by others (Shahani and Jenkinson, 2016) and might indicate that rote memorization from the instructions one year before the interview led to confusion about the details of the memorized facts.
In contrast to S13, who described both ionic and covalent bonding using a horizontal discourse, S4 did not recall what either an ionic or a covalent bond is, or how they are formed. However, the student knew that bond formation occurs due to some kind of force as illustrated in the following quote:
T: Is there any force that makes them [the two atoms] bond?
S4: Yes, there is an attraction force that attracts the electrons.
T: Do you remember what it is that attracts the electrons?
S4: Ehm… no I don’t know really.
T: OK, but you do remember that there is something that attracts the electrons.
S4: Yes, perhaps it has something to do with the nucleus, but I am not sure.
During the written test a year ago, S4 described that both ionic and covalent bonds were formed because of electrostatic forces (subcategory Hi2 in Table 2) but used a horizontal discourse when describing the formation of covalent and ionic bonds in detail (data not shown). However, during the interview a year after S4 did not describe formation of bonds by using either a horizontal or a vertical discourse (see appendix, Tables 4 and 5). This might indicate that learning about chemical bonding through a horizontal discourse may have been forgotten, or that the change in methodology during data collection influenced the student to describe chemical bonding in less detail.
Didactic consequences for instruction of chemical bonding
A force-based approach to teaching chemical bonding has been proposed as an alternative method to help address common alternative conceptions about this topic (Levy Nahum et al., 2008). However, to the best of the author's knowledge, the force-based approach has been examined in only a limited number of studies (Joki et al., 2015; Venkataraman, 2017; Zohar and Levy, 2019). This study investigates the long-term effects of the intervention. Even though the force-based approach can be considered a simplified model, omitting certain quantum mechanical details, previous studies have suggested that it may be too demanding for high school students (Levy Nahum et al., 2008; Stevens et al., 2010). In this study, while the participating students were able to describe the fundamental reasons for chemical bonding using a hierarchical discourse directly after instruction, several alternative conceptions appeared after one year and are listed in Table 3.| Row | Challenging topics | Alternative conceptions | Possible consequences of the alternative conceptions | Suggestions of didactic approaches to address the topic |
|---|---|---|---|---|
| 1 | The concept of electronegativity and the usage of the Pauling scale. | The concept of electronegativity describes atoms’ ability to react. | Students may perceive ionic and covalent bonding as a strict dichotomy. | Emphasize the concept of electronegativity for instance by alternative teaching methods (Ucar, et al., 2017; Jones and Spencer, 2018; Danckwardt-Lillieström et al., 2020) and/or by incorporating electronegativity values when introducing non-polar covalent bonds (Dhindsa and Treagust, 2014). |
| The concept of electronegativity describes the strength of attraction between two ions/atoms in ionic bonds. | ||||
| 2 | Ions, and subsequently ionic bonds, are formed because of electrostatic forces between charged particles, which results in decreased potential energy of the system. | Ions are formed because of atoms tendency to form octets. | Overemphasis of the octet framework may lead students to overlook the role of attraction between oppositely charged ions. | Covalent bonds may be introduced prior ionic bonds (Dhindsa and Treagust, 2014). |
| Strongly emphasize that electrostatic attraction between the nucleus and other atoms electrons is the driving force when ions are formed. | ||||
| 3 | Chemical bonds are formed because of atoms tendency to reach lower energy. | Atoms strive to reach stability trough noble gas configuration. | The concept of stability may be misapplied, leading to incorrect generalization about why chemical bonds are formed. | The concept of stability should be avoided since it does not consistently align with a low-energy state. |
| Response group (delayed task-based interview) | Student | Previous group (Written test) | Description of ionic bonds by using rote memorization | Horizontal description of ionic bond formation other than rote memorization | Ionic bonds are formed because of electrostatic forces (attraction and/or repulsion between particles) | Ionic bond formation results in lower potential energy | Oppositely charged ions do not attract |
|---|---|---|---|---|---|---|---|
| Response group A 5 hierarchical subcategories | S5 | Mostly hierarchical | x | x | x | x | |
| S6 | Solely hierarchical | x | x | x | x | ||
| S8 | Mostly hierarchical | x | x | x | x | ||
| S10 | Solely hierarchical | x | x | x | x | ||
| S14 | Solely hierarchical | x | x | x | x | ||
| Response group B 3–4 hierarchical subcategories | S1 | Mostly hierarchical | x | x | x | ||
| S7 | Mostly hierarchical | x | x | ||||
| S9 | Solely hierarchical | x | x | x | x | ||
| Response group C 3–4 horizontal subcategories | S2 | Mostly hierarchical | x | x | |||
| S3 | Mostly hierarchical | x | x | x | |||
| S11 | Solely hierarchical | x | x | x | |||
| S12 | Mostly horizontal | x | x | x | x | ||
| S15 | Mostly horizontal | x | x | x | |||
| Response group D solely horizontal subcategories | S13 | Mostly horizontal | x | x | |||
| S4 | Mostly horizontal |
| Response group (delayed task-based interview) | Student | Previous group (written test) | Description of covalent bonds by using rote memorization | Horizontal description of covalent bond formation other than rote memorization | Covalent bonds are formed because of electrostatic forces (attraction and/or repulsion between particles) | Covalent bond formation results in lower potential energy | Attraction of several nuclei results in covalent bond formation |
|---|---|---|---|---|---|---|---|
| Response group A 5 hierarchical subcategories | S5 | Mostly hierarchical | x | x | x | x | |
| S6 | Solely hierarchical | x | x | x | x | ||
| S8 | Mostly hierarchical | x | x | x | x | ||
| S10 | Solely hierarchical | x | x | x | x | ||
| S14 | Solely hierarchical | x | x | x | x | ||
| Response group B 3–4 hierarchical subcategories | S1 | Mostly hierarchical | x | x | x | x | |
| S7 | Mostly hierarchical | x | x | x | x | ||
| S9 | Solely hierarchical | x | x | x | x | ||
| Response group C 3–4 horizontal subcategories | S2 | Mostly hierarchical | x | x | x | ||
| S3 | Mostly hierarchical | x | x | x | |||
| S11 | Solely hierarchical | x | |||||
| S12 | Mostly horizontal | x | x | x | x | ||
| S15 | Mostly horizontal | x | x | ||||
| Response group D solely horizontal subcategories | S13 | Mostly horizontal | x | x | |||
| S4 | Mostly horizontal |
| Response group (delayed task-based interview)) | Student | Previous group (written test) | Atoms' active strive to reach noble gas configuration results in the formation of ionic and covalent bonds | Electrostatic attraction between oppositely charged particles results in ionic and covalent bond formation | Electrostatic attraction between a positively charged nucleus and negatively charged electrons results in both ionic and covalent bond formation | Repulsion between similarly charged particles keeps bonding atoms at a specific distance from each other | Atoms' tendency to reach lower potential energy results in the formation of ionic and covalent bonds | Nonpolar covalent bonds, polar covalent bonds and ionic bonds are described as a continuum scale |
|---|---|---|---|---|---|---|---|---|
| Response group A 5 hierarchical subcategories | S5 | Mostly hierarchical | x | x | x | x | ||
| S6 | Solely hierarchical | x | x | x | ||||
| S8 | Mostly hierarchical | x | x | x | ||||
| S10 | Solely hierarchical | x | x | x | x | |||
| S14 | Solely hierarchical | x | x | x | ||||
| Response group B 3–4 hierarchical subcategories | S1 | Mostly hierarchical | x | |||||
| S7 | Mostly hierarchical | x | x | |||||
| S9 | Solely hierarchical | x | x | x | ||||
| Response group C 3–4 horizontal subcategories | S2 | Mostly hierarchical | x | x | ||||
| S3 | Mostly hierarchical | x | x | x | ||||
| S11 | Solely hierarchical | x | x | |||||
| S12 | Mostly horizontal | x | ||||||
| S15 | Mostly horizontal | x | ||||||
| Response group D solely horizontal sub-categories | S13 | Mostly horizontal | x | |||||
| S4 | Mostly horizontal | x |
Specifically, several students demonstrated difficulty in understanding the concept electronegativity and applying the Pauling scale (Table 3, Row 1), which might foster a dichotomous perception of ionic and covalent bonding. During the instructions about chemical bonding only one lesson (90 minutes long) focused on how non-polar covalent bonds, polar covalent bonds and ionic bonds can be understood as a continuum scale, as proposed by Levy Nahum et al. (2008). This section may require more time to process, perhaps by applying alternative teaching methods, such as drama (Danckwardt-Lillieström et al., 2020), 3D-visualization (Jones and Spencer, 2018), or haptic augmented simulations (Ucar et al., 2017). Dhindsa and Treagust (2014) proposed an instructional approach in which covalent bonds are introduced before polar covalent and, subsequently, ionic bonds. This sequencing allows for the early introduction of the concept of electronegativity and highlights that increasing differences in electronegativity between bonded atoms correspond to a decreasing extent of orbital overlap, with ionic bonding representing the extreme case of maximal polarity. Moreover, this approach may prevent an overemphasis of the octet role during ion formation, prior ionic bonding, which is another alternative conception stated by some of the participating students (Table 3, row 2). It is possible that comparatively little time was devoted to teaching ionic compounds relative to covalent bonds and molecular structures. The short timeframe and limited opportunities to repetition may also have affected the students’ long-term ability to apply this knowledge. Finally, several students in the present study preferred to express that chemical bonds are formed due to their tendency to find stability (Table 3, row 3) instead of lower energy, which is more scientifical correct. Thus, one suggestion is to avoid the concept stability during instruction and solely highlight that chemical bonding minimize the system's energy through electrostatic forces.
General findings and conclusions
Most students in this study demonstrated difficulty to explain certain scientific aspects of chemical bonding one year after instruction, regardless of the type of discourse used one year ago. However, those who employed solely or primarily hierarchical discourse during the delayed task-based interview exhibited a broad repertoire of mental tools for solving chemical problems and explaining how and why chemical bonds form. While it is common for students to forget some content without continued practice, the findings in the present study underscore the importance of both depth and quality of learning – what is understood deeply is more likely to be retained over time.This study also sheds light on alternative conceptions that may appear a year post instruction. Notably, even students demonstrating a predominantly hierarchical knowledge discourse during the interviews struggled to explain the concept of electronegativity and its role in explaining bond formation. Additionally, while covalent bonding was often described using a hierarchical discourse, student tended to use a more horizontal discourse when discussing ionic bonding, frequently relying in the octet rule as the primary explanatory framework. This pattern highlights the importance to strongly emphasize electrostatic interaction between both subatomic particles and ions to help students move beyond dichotomous explanations of covalent versus ionic bonding.
Conflicts of interest
There are no conflicts to declare.Data availability
Data collection from the participating students are not available due to ethical reasons.Appendix
Questions in the delayed task-based interview
2. Did you think something was particularly difficult to understand?
3. Do you have any suggestions of what your teacher could do to make the lectures more fun or interesting?
5. How would you describe an ion?
6. How would you describe a molecule?
7. Which chemical bonds do you remember?
8. Can you give any examples of chemical substances and what intramolecular chemical bonds they consist of?
9.
a. How would you describe an intramolecular chemical bond? Make a drawing and explain.
b. What particles are involved?
c. How do the particles behave?
10. The researcher shows Fig. 4 to the student.
 | ||
| Fig. 4 A Bohr model (a) and a quantum mechanical model (b) showing two hydrogen atoms approaching each other. | ||
a. How would you describe the models shown in Fig. 4a and b?
b. Do you see any difference between the models?
c. Do you see any similarities between the models?
d. Which model do you think describe chemical bonding the best?
11. Why do you think chemical bonds are formed?
12. Do you think chemical bonds are formed in the same way or in different ways?
a. How do you think magnesium oxide (MgO) is formed? Draw a model and explain.
b. Why do you choose to draw a model in this way (e.g. by drawing eight valence electrons)?
c. What is the reason behind the bounding between the magnesium ion and the oxygen ion? (The particles are only called ions if the student know that ions are formed).
d. Is there any force involved in the formation of magnesium oxide?
(If the student does not know that magnesium oxide consists of an ionic bond, the student is asked to explain the formation of sodium chloride).
14.
a. How do you think chlorine gas is formed? Draw a model and explain.
b. Why do you choose to draw a model in this way (e.g. by drawing eight valence electrons)?
c. What is the reason behind the bounding between the chlorine atoms?
d. Is there any force involved in the formation of chlorine gas?
(If the student does not know that chlorine gas consists of a covalent bond, the student is asked to explain the formation of hydrogen gas).
16. Which intramolecular bonds do you think the following chemical substances consist of?
a. Water
b. Sodium chloride (NaCl)
c. A piece of iron
d. Oxygen gas
17. Can you please explain how you determine the different bonds in the chemical substances in question 16?
18. What do you think is the reason why intramolecular chemical bonds are formed?
19. Which model in Fig. 5 do you think describe hydrogen fluoride the best? Please motivate your answer.
 | ||
| Fig. 5 Three models of hydrogen fluoride. | ||
References
- Anderhag P., Wickman P.-O., Bergqvist K., Jakobson B., Hamza K. M. and Säljö R., (2016), Why do secondary school students lose their interest in science? Or does it never emerge? A possible and overlooked explanation, Sci. Educ., 100(5), 791–813.
- Ausubel D., (1968), Educational Psychology: A Cognitive View, New York: Holt, Rinehart, and Winston.
- Bennett J. and Hogarth S., (2009), Would you want to talk to a scientist at a party? High school students’ attitudes to school science and to science, Int. J. Sci. Educ., 31(14), 1975–1998.
- Bergqvist A., Drechsler M., De Jong O. and 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.
- Bergqvist A., Drechsler M. and Rundgren S.-N. C., (2016), Upper Secondary Teachers’ Knowledge for Teaching Chemical Bonding Models, Int. J. Sci. Educ.38(2), 298–318.
- Bernstein B., (1999), Vertical and horizontal discourse: an essay, Brit. J. Sociol. Educ., 20(2), 157–173.
- Bernstein B., (2000), [1996] Pedagogy, Symbolic Control and Identity. Theory, Research, Critique, London: Rowman & Littlefield.
- Bowe K. A., Bauer C. F., Wang Y. and Lewis S. E., (2022), When all you have is a covalent model of bonding, every substance is a molecule: a longitudinal study of student enactment of covalent and ionic bonding models, J. Chem. Educ.99, 2808–2820.
- Boyatziz R. E., (1998), Transforming qualitative information: thematic analysis and code development, London, Thousand Oaks, CA: Sage.
- Braun V. and Clarke V., (2006), Using thematic analysis in psychology, Qual. Res. Psychol., 3(2), 77–101.
- Broman K., Bernholt S. and Parchmann I., (2018), Using model-based scaffolds to support students solving context based chemistry problems, Int. J. Sci. Educ.40(10), 1176–1197.
- Broman K. and Parchmann I., (2014), Students’ application of chemical concepts when solving chemistry problems in different contexts, Chem. Educ. Res. Pract., 15, 516–529.
- Burrows N. L., Ouellet J. O., Joji J. and Man J., (2021), Alternative assessment to lab reports: a phenomenology study of undergraduate biochemistry students’ perceptions of interview assessment, J. Chem. Educ., 98, 1518–1528.
- Chen X., (2015), STEM attrition among high-performing college students: scope and potential causes, J. Technol. Sci. Educ., 5, 41–59.
- Danckwardt-Lillieström K., Andrée M. and Enghag M., (2020), The drama of chemistry – supporting student explorations of electronegativity and chemical bonding through creative drama in upper secondary school, Int. J. Sci. Educ., 42(11), 1862–1894.
- Ding N. and Harskamp E. G., (2011), Collaboration and peer tutoring in chemistry laboratory education, Int. J. Sci. Educ., 33(6), 839–863.
- Dhindsa H. D. and Treagust D. F., (2014), Prospective pedagogy for teaching chemical bonding for smart at sustainable learning, Chem. Educ. Res. Pract., 15, 435–446.
- Doymus K., (2008), Teaching Chemical Bonding through Jigsaw Cooperative Learning, Res. Sci. Technol. Educ., 26(1), 47–57.
- Ferreira S. and Morais A., (2018), Practical work in science education: study of different context of pedagogic practice, Res. Sci. Educ., 50, 1547–1574.
- Goldin G., (2000), A scientific perspective on structures, task-based interviews in mathematics education research, in Lesh R. andKelly A. E. (ed.), Research design in mathematics and science education, New York: Routledge, pp. 517–545.
- Hunter K. H., Rodriguez Jon-Marc G. and Becker Nicole M., (2022), A Review of Research on the Teaching and Learning of Chemical Bonding, J. Chem. Educ., 99, 2451–2464.
- Hurst M. O.,(2002), How we teach molecular structure to freshmen, J. Chem. Educ., 79(6), 763–764.
- Joki J. and Aksela M., (2018), The challenges of learning and teaching chemical bonding at different school levels using electrostatic interactions instead of the octet rule as a teaching model, Chem. Educ. Res. Pract., 19(3), 932–953.
- Joki J., Lavonen J., Juuti K. and Aksela M., (2015), Coloumbic interaction in Finnish middle school chemistry: a systematic perspective on students’ conceptual structure of schimcal bonding, Chem. Educ. Res. Pract., 16(4), 901–917.
- Jones O. A. H., and Spencer M. J. S., (2018), A Simplified Method for the 3D Printing of Molecular Models for Chemical Education, J. Chem. Educ., 95, 88–96.
- Koro Arvidsson C., (2025), Chemical bonding in Swedish upper secondary school education: a force-based teaching model for enhanced understanding, J. Chem. Educ., 26(1), 315–333.
- Levy Nahum T., Mamlok-Naaman R. and Hofstein A., (2007), Developing a new teaching approach for the chemical bonding concept aligned with current scientific and pedagogical knowledge, Sci. Educ., 91(4), 523–682.
- Levy Nahum T., Mamlok-Naaman R. and Hofstein A., (2008), A new “bottom-up” framework for teaching chemical bonding, J. Chem. Educ., 85(12), 1680–1685.
- Luxford C. J. and Bretz S. L., (2013) Moving beyond Definitions: What Student-Generated Models Reveal about Their Understanding of Covalent Bonding and Ionic Bonding, Chem. Educ. Res. Pract.14(2), 214–222.
- 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.
- Maher C. A. and Sigley R., (2014), Task-based interviews in mathematics education, in Lerman S. (ed.), Encyclopedia of Mathematics Education, Dordrecht: Springer.
- Musengimana J., Kampire E. and Ntawiha P., (2021), Factors affecting schools students’ attitudes toward learning chemistry: a review of literature. EURASIA J. Math., Sci. Technol. Educ., 17(1), 1–12.
- Muteti C. Z., Zarraga C., Jacob B. I., Mwarumba T. M., Nkhata D. B., Mwavita M., Mohanty S. and Mutambuki J. M., (2021), I realized what I was doing was not working: the influence of explicit teaching of metacognition on students’ study strategies in a general chemistry I course, Chem. Educ. Res. Pract., 22, 122–135.
- Patton M. Q., (1987). How to use qualitative methods in evaluation, Newbury Park, CA: Sage.
- Player-Koro C., (2011). Marginalising students’ understanding of mathematics through performative priorities: a Bernsteinian perspective, Ethn. Educ., 6(3), 325–340.
- Prodjosantoso A. K., Hertina A. M. and Irwanto, I., (2019). The Misconception Diagnosis on Ionic and Covalent Bonds Concepts with Three Tier Diagnostic Test, Int. J. Instr.12(1), 1477–1488.
- Segal B. G., (1989), Chemistry experiment and theory, 2nd edn, New York: John Wiley & Sons.
- Sevian H. and Talanquer V., (2014), Rethinking chemistry: a learning progression on chemical thinking, Chem. Educ. Res. Pract., 15(1), 10–23.
- Seymour E. H. and Hewett N. M., (1997), Talking about leaving: Why undergraduates leave the sciences, Talking about leaving: Why undergraduates leave the sciences, High. Educ.36, 115–116.
- Shahani V. M. and Jenkinson J., (2016), The Efficacy of Interactive Analogical Models in the Instruction of Bond Energy Curves in Undergraduate Chemistry, Chem. Educ. Res. Pract.17(2), 417–428.
- Stevens S. Y., Delgado C. and Krajcik J. S., (2010), Developing a hypothetical multidimensional learning progression for the nature of matter, J. Res. Sci. Teach., 47(6), 687–715.
- Taber K. S., (1997), Student understanding of ionic bonding: molecular versus electrostatic framework? 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., (2003), Lost without a trace or not brought to mind? – A case study of remembering and forgetting of collage studies, Chem. Educ. Res. Pract., 4(3), 249–277.
- Taber K. S. and Coll R., (2002), Bonding, in Gilbert J., De Jong O., Justi R., Treagust D. and Van Driel J. (ed.), Chemical education: towards research-based practice, Dordrecht: Kluwer Academic Publishers BV., pp. 213–234.
- Taber K. S. and Watts M., (1996), The secret life of chemical bond: students’ anthropomorphic and animistic references to bonding, Int. J. Sci. Educ.18(5). 557–568.
- Taber K. S. and Watts M., (2000). Learners’ explanations for chemical phenomena, Chem. Educ.: Res. Practice Europe, 1(3), 329–353.
- Tsapralis G., Pappa E. T. and Byers B., (2019). Proposed pedagogies for teaching and learning chemical bonding in secondary education, Chem. Teacher Int., 2(1), 401–420.
- Ucar E., Ustunel H., Civelek T. and Umut I., (2017), Effects of using a force feedback haptic augmented simulation on the attitudes of the gifted students towards studying chemical bonds in virtual reality environment, Behav. Inf. Technol., 36(5), 540–547.
- Venkataraman B., (2017), Emphasizing the significance of electrostatic interactions in chemical bonding, J. Chem. Educ., 94(3), 296–303.
- Zohar A. R. and Levy T., (2019). Attraction vs. repulsion – learning about forces and energy in chemical bonding with the ELU-Chem simulator, Chem. Educ. Res. Pract., 20(667), 667–684.
- Zoller U. and Pushkin D. B., (2007), Matching Higher-Order Cognitive Skills (HOCS) promotion goals with problembased laboratory practice in a freshman organic chemistry course, Chem. Educ. Res. Pract., 8(2), 153–171.
| This journal is © The Royal Society of Chemistry 2026 |