VanessaBarker,ScienceandTechnologyGroup,UniversityofLondonInstituteofEducation,LondonWC1H0AL;,DepartmentofEducationalStudies,UniversityofYork,YorkYO105DD; Students’reasoningaboutbasicchemicalthe

30  Download (0)

Full text



Students’ reasoning about basic chemical

thermodynamics and chemical bonding: what changes occur during a context-based post-16 chemistry course?

Vanessa Barker, Science and Technology Group, University of London Institute of Education, London WC1H 0AL; e-mail: and Robin Millar, Department of Educational Studies, University of York, York YO10 5DD; e-mail:

A longitudinal study of 250 students following the Salters Advanced Chemistry (SAC) course probed a range of chemical ideas including the exothermicity of bond formation and the development of thinking about covalent, ionic and intermolecular bonds. Students responded to the same diagnostic questions on three occasions: at the start, after eight months and sixteen months of a twenty-month course. At the start, many students demonstrated misunderstandings about these chemical ideas, but in general their understanding improved as the course progressed. By the end of the study, about half knew that bond making is exothermic. Initially, few described covalent bonds accurately or understood hydrogen bond- ing. A majority gave responses at the final survey which were in line with ideas and language a chemist may use. Students attributed changes to the use of context-based materials including a drip-feed approach which allowed their understanding to develop over time. However, some aspects of chemical bonding, including ionic bonding and intermolecular bonds other than hydrogen bonds remained problematic for students despite explicit teaching. The findings have implications for post-16 chemistry teaching, suggesting that a review of teaching strategies is needed in some areas.


In recent years, courses adopting a context-based approach to science teaching at the secondary (11-18 year old) level have been developed in several different countries attracting international attention. By presenting scientific concepts in everyday situations, course authors aim to promote students’ enthusiasm and motivation for science. The Salters project in the UK has involved development of Salters’ Chemistry (SC), Salters’ Science (SS), Salters Advanced Chemistry (SAC) and currently under development, Salters Horners Advanced Physics (SHAP) (University of York Science Education Group 1989, 1990± 2, Burton et al. 1994a, 1994b, 1994c and 1994d, Swinbank 1997, Edexcel Foundation 1998). Other examples include The Supported Learning in Physics Project (Whitelegg and Edwards 1997), ChemComm (American Chemical Society 1988), the Chemical Education for Public Understanding Programme (CEPUP) (1991) and PRIME Science (The PRIME Science Education Group 1998) in the US, and the Physics Curriculum Development Project (PLON) (Eijkelhof and

International Journal of Science Education ISSN 0950-0693 print/ISSN 1464-5289 online#2000 Taylor & Francis Ltd


Kortland 1988) in the Netherlands. The approach raises questions about the effec- tiveness of learning science in this context-led way rather than through the more usual form of curriculum organization and sequencing based on the major divi- sions and sub-divisions of the science subject. This study, completed in 1995 (Barker 1994 and Barker and Millar 1996a, 1996b and 1996c), focused on students following the Salters Advanced Chemistry (SAC) course and sought answers to two research questions: What level of understanding do beginning A level students have about basic chemical ideas? and in what ways is student learning influenced by the context-theory approach?

Their thinking about a variety of basic chemical ideas was investigated. This article presents data on students’ understanding of basic thermodynamics and chemical bonding. An earlier paper (Barker and Millar 1999) reported findings concerning students’ thinking about other aspects of chemical reactions.

The Salters Advanced Chemistry (SAC) course

At the age of 16 students in England and Wales select usually three subjects for study at Advanced (A) level. A level courses take about 20 months to complete and are regarded as preparation for university entrance. SAC (Burton et al. 1994a, 1994b, 1994c and 1994d) was developed by the University of York Science Education Group to provide a stimulating account of chemistry for 16± 18 year olds by emphasizing industrial and real-life applications of chemistry. This, it was hoped, would help to increase the numbers of students electing to study chemistry beyond the end of compulsory education at age 16 by helping to retain students’

motivation for studying chemistry and so aid the quality and future supply of research and industrial chemists.

SAC adopts a novel, context-led approach to A level chemistry teaching. The course comprises thirteen theoretical units each made up of a ‘storyline’, ‘chemical ideas’ and ‘activities’. Each unit requires approximately twenty hours of teaching time. These units include all the nationally agreed subject core for Chemistry (SEAC 1993). The course also includes a structured visit to the chemical industry and an extended individual practical investigation. The chemical ` storylines’ pro- vide contextual settings for chemistry reflected in the unit titles, for example

` Developing Fuels’ . Three important consequences arise from this structure:

first, chemical ideas are introduced only as contexts demand, thus breaking down the traditional physical, inorganic and organic divisions of chemistry. This encourages students to draw several aspects of the subject together to understand a specific chemical context. Second, students only learn the chemistry required to understand each storyline, so any one chemical topic is delivered in a ` drip-feed’

fashion through several units which are taught in a prescribed order. This feature has an advantage from a research perspective in that a population of students drawn from different schools and colleges receive the same material at approxi- mately the same time. Third, chemical ideas are revisited as the course proceeds, allowing students’ understanding to develop over a longer time period than is possible with a ` traditional’ type course. These tend to deliver each chemical topic in one block of lessons, with contexts used as illustrations after the ideas have been taught rather than as pretexts for introducing them. The data collected during the study explore the strengths and illustrate some weaknesses associated with these strategies.


The treatment of chemical bonding and thermodynamics in SAC

Treatment of chemical bonding and thermodynamics illustrate the key features of the SAC approach. Different aspects of chemical bonding feature in many units, so this is an excellent example of ‘drip-feeding’ information. In the first unit, ‘The elements of life’, students learn how covalent and ionic bonds form, how to repre- sent simple covalent molecules using dot/cross diagrams and the meaning of bond polarity. In unit 4, ‘The Atmosphere’, bond polarity is revisited in the context of heterolytic fission. Unit 5, ‘The Polymer Revolution’, uses intermolecular bonds including hydrogen bonds to help explain polymer properties; and in unit 8,

` Engineering Proteins’ , students learn about the shapes of simple covalent mol- ecules. Knowledge about covalent and ionic bonds is assumed in several other units, where ideas such as molecular stability, lattice enthalpy, enthalpy of solution and variation in boiling point are discussed. Thus, students revisit chemical bond- ing on many occasions allowing knowledge to be reinforced and developed through application to a variety of contexts. New ideas or aspects of the topic are added at each point, so students build up the knowledge required over a long period of time.

Basic thermodynamics concepts are introduced in unit 2, ` Developing Fuels’ , which is taught within the first three months. The storyline describes how petrol became the prime world fuel and provides a strong context-based link for teaching and learning basic thermodynamics ideas. Students learn that bond breaking is endothermic and bond making is exothermic, how to perform Hess’ Law calcula- tions and meet entropy in a qualitative sense. Aspects of thermodynamics feature in several other units. Ideas about energy changes are developed in unit 3, ` From minerals to elements’ . Bond breaking is met again in unit 4, ` The Atmosphere’ ; and enthalpy change of combustion is used to compare hydrogen and petrol in unit 7, ` Using Sunlight’ . These units are taught within the first twelve months of the course. In unit 13, ` The Oceans’ , students learn about entropy in more detail.

These later references offer students the opportunity to revisit ideas presented in earlier units and permit consolidation of understanding.

Students’ understanding of chemical bonding and thermodynamics

We provide a brief review of literature relating to these areas of chemistry. We note that both have received relatively little attention from researchers compared to other topics.


The basic chemical idea associated with thermodynamics probed in this study is that energy is released when chemical bonds form. Ross (1993) notes that many students think the opposite, that energy is released when chemical bonds break. He suggests this arises because of the strong association students develop between fuels and energy. This prompts them to learn the phrase ` fuels contain energy’

almost by rote, and then associate chemical bonds with an energy-storing role.

Work by Andersson (1984), Schollum (1981) and BouJaoude (1991) among others show that students aged 14± 15 tend to describe burning in very primitive ways,


and may associate energy release with ` flames’ or simply ` burning’ . Early teaching about chemical bonding provides a more scientific-sounding notion: that the energy comes from breaking bonds. Boo (1998) suggests that this idea may be linked to the macroscopic everyday experience that energy is needed to make or do something, and so to ` make bonds’ must therefore require energy which is released when the bonds are broken.

Ross (1993) suggests that describing combustion as a ` fuel-oxygen system’

may assist students, as this means that the fuel and the energy generated on its combustion must be perceived in the context of a chemical reaction. Students can be encouraged to think about how the products are formed rather than focusing on the starting molecules alone.

Chemical bonding

In this study we report students’ ideas about covalent, ionic and intermolecular bonding. We consider earlier work in three separate sections.

Covalent bonds

Taber (1993a and b) reports an intensive case study carried out with one student,

` Annie’ , which charts her progress towards understanding that chemical bonds involve electrostatic attractions between atomic nuclei and electrons. Initially, Annie described covalent bonding as atoms ` pulling together’ . At the mid-point of the study, she associated covalent bonds with electron sharing and acquisition of full electron orbitals. These ideas reflect increasing sophistication of her ideas and may point to stages of development among students more generally.

Peterson and Treagust (1989, Peterson 1993) probed the understanding of 17 year old chemistry students and found that about 23%of them thought that elec- trons were shared equally in all covalent bonds. As almost without exception all covalent bonds exhibit some degree of polarity, this is an important misconception.

These authors also note that 60% of this group and 55% of first year university chemists could not correctly position the electron pair between hydrogen and fluorine.

Boo (1998) notes that some students think a single covalent bond comprises one electron alone, in the same way an apple may be shared between two people.

Ionic bonds

Misconceptions about ionic bonds common to several studies are noted. The notion that ionic substances exist as discrete molecules was found by Butts and Smith (1987), Taber (1993a) and Boo (1998) among chemists aged 17± 19 years.

Boo also notes that this idea influences students’ thinking about how ionic com- pounds behave; for example, a solution of sodium chloride comprises water mol- ecules and sodium chloride molecules, while hydrochloric acid contains hydrogen chloride molecules. In each case the molecules are made from ions bonded together.

Butts and Smith (1987) and Boo (1998) also report some students describe the bond between sodium and chlorine as covalent. This finding is corroborated by Boo (1998).


Boo (1998) notes that students who view ionic compounds as molecular think that the intramolecular ionic bond will be stronger than other bonds if the mol- ecules are placed close together. This finding is similar to that reported by Taber (1994), who describes a ` molecular framework’ characterizing the development of thinking about ionic bonds among five interviewees. This features three key mis- conceptions: that as a sodium atom can donate only one electron, so it can form only one ionic bond; that the sodium ion formed is bonded only to the chloride ion produced on receiving the electron; that the ions may be attracted by forces to other ions, but that these attractions are not ionic bonds.

Intermolecular bonds

‘Annie’ was asked (Taber 1993a) to describe the bonding she perceived between hydrogen fluoride molecules drawn in a chain arrangement, showing distorted electron clouds which were touching one another. She did not think there was any bonding between the molecules. By the end of Taber’ s study, completed post- teaching, Annie demonstrated much clearer understanding of hydrogen bonding.

She also knew about temporary induced dipole-dipole bonds (van der Waals’

` forces’ ) and incorrectly thought these would be present in sodium chloride.

Peterson and Treagust (1989) also investigated students’ thinking about intermolecular bonds. They found that about 23%of students thought that inter- molecular bonds were positioned within a covalent molecule. They also note that one-third of their sample confused the relative strengths of inter- and intramole- cular bonds, saying that ` strong intermolecular forces exist in a continuous covalent network’ (Peterson and Treagust 1989: 460). In his later work, Peterson (1993) reports that 36% of first year university chemists thought that silicon carbide had a high melting point because of ` strong intermolecular forces’ .

Peterson et al. (1989) noted that some 17 and 18 year old chemists thought that intramolecular bonds break on change of state, rather than intermolecular bonds.

The longitudinal study

The progress of 250 SAC students drawn from thirty six different schools and colleges in the UK was tracked by asking them to complete the same diagnostic questionnaire comprising twenty three questions on three occasions - at the start, after eight months and after sixteen months of the twenty month course.

Questionnaires were sent to schools for completion by students under examination conditions. Teachers administered the questionnaires without warning in a lesson within a specified two week period. Students were not permitted to take question- naires home or discuss or debate the paper with each other or their teacher. They were not given any feedback on their responses. The questionnaire was designed to require a maximum of one hour to complete and was presented in a booklet format on coloured paper to ensure a distinctive appearance. Each question was given a name rather than a number to help give a ‘user-friendly’ and non-threatening style. Question names are used throughout this paper. Response rates to most questions were over 80% on all three occasions, indicating that students were able to complete the questionnaire within the timescale suggested. Data were also collected relating to the students’ progress through SAC. These indicated


that by the time of the second and third surveys the great majority of students in the study had completed the same course units and so could reasonably be expected to have had similar learning experiences.

Students’ responses to eight of the twenty three questions are reported here.

‘Methane’ and ‘Energy change’ investigated ideas about basic thermodynamics.

The question ‘Chemical bonds’ investigated ideas about covalent bonds and hydrogen bonds, so data are reported separately under these headings. The for- mation of stable covalent molecules was explored using ‘Methane molecules’. The question ` Sodium and chlorine’ probed ionic bond formation, while ` Hydrogen chloride’ explored ideas about the formation of ions from a covalent molecule. The questions ` Boiling and Chlorides’ permitted investigation of ideas about the roles of intermolecular bonds in changes of state. Six questions were devised for this study, while ` Boiling’ was adapted from Osborne and Cosgrove (1983) and

` Chlorides’ from a University of London A level Chemistry examination paper set in June 1990. Figures 1± 4 give representations of the questions as presented to the respondents. Data relating to the questions are presented in tables 1± 9.

The expected responses and chemical ideas probed are given as questions are discussed.

Twenty four students whose thinking appeared to change markedly after the second survey were selected for interview to validate written responses and to probe reasons for these changes. Students selected for interview had given written responses which were representative of the cohort as a whole. They were interviewed one-to-one with the researcher and their permission was sought to record the interviews for research purposes. Respondents were told in advance about the interview, but were not given any warning about the content. They were invited to re-read their answers given in the first and second surveys and to comment on changes in their thinking. The interviews confirmed that students’

responses had been interpreted correctly and so ensured that the analysis of written answers was accurate. The interviews were not carried out as a free-standing study and so were not subjected to rigorous analysis. Selected extracts from these interviews are reported here to help highlight and clarify points noted in the tabulated data.

Describing changes in students’ understanding

In analysing the data a novel coding strategy was devised which was applied con- sistently across all questions in the test paper. This section begins by noting differences between ‘Methane and Energy change’. These are used to inform the comments about development of the coding system.

In ‘Methane’, the chemical equation is given together with a detailed enthalpy level diagram showing the formation of products from the reactants. Students were asked specific questions focusing on aspects of the information presented.

Responses were given in terms of the chemical idea being probed. The chemical reaction was not used in students’ answers.

In ` Energy change’ , respondents were given the equation for the reaction together with much less detailed enthalpy level diagrams. The question was more open than ` Methane’ and accordingly a wider range of responses was gen- erated. In some cases students responded in terms of the chemical reaction, rather than using ideas about thermodynamics.


The existence of responses using ideas about the chemical reaction rather than the chemical idea under test is significant and provided a key to devising the codings. A distinction was made between students’ understanding of two aspects of the questions: the chemical idea being probed (aspect 1) and the chemical reaction or event used (aspect 2). Application of this distinction helped to clarify responses which were neither completely in accordance with the expected answers nor indicative of misunderstanding the chemical idea being probed.

Hence, responses were grouped into five main categories denoted P, Q, R, S and T in tables 1 and 2. A summary of the types of evidence placed in each follows:

. P Evidence for understanding aspect 1, no evidence for misunderstanding aspect 2;

. Q Partial evidence for understanding of aspect 1, no evidence for misun- derstanding of aspect 2;

. R Evidence for understanding aspect 2, no evidence for misunderstanding aspect 1;

. S Evidence for misunderstanding aspect 2, evidence for understanding aspect 1;

. T Evidence for misunderstanding aspect 1; and . U Uncodeable responses, including no responses.

In the case of ‘Methane’, responses were placed only into categories P, Q, T and U. This also applies to ‘Covalent bonds’, ‘Methane molecules’, ‘Hydrogen bonds’,

‘Boiling and chlorides’. In answering these, categories R and S were not needed, as students always responded in terms of aspect 1, in most cases because no chemical reaction featured in the question. For ‘Energy change’, all categories were used, because some responses focused on the chemical reaction between sodium and chlorine and did not mention aspect 1. In ` Sodium and chlorine’ and ` Hydrogen chloride’ , students also gave responses involving the chemical reactions. In cate- gory Q were placed responses which were incomplete or inconclusive and so gave only partial evidence. For ` Methane’ , for example, some students did not answer all three parts of the question, while for ` Energy change’ students may have selected a correct diagram, but failed to explain their choice. The scheme was verified by chemical educators and their comments on placing responses in specific categories were noted in preparation of the final codings.

The data tables give the levels of understanding represented in the five main categories and indicate the proportion of uncodeable responses at each stage.

Examination of the first column indicates the understanding exhibited by students beginning SAC. These students had taken the General Certificate of Secondary Education (GCSE) examinations several months prior to the data collection. Use of the three columns together permits us to report shifts in response levels between categories as the study proceeded. To aid this, codings were entered on a computer database, which was used to explore patterns of response across the three surveys.

These data point to influences of the context-theory approach on students’

learning as it is possible to tie changes in responses to the material presented prior to the administration of the surveys. À2tests were used to determine whether changes in the relative proportion of P-coded responses in the third survey were significant at the 0.05 or 0.01 level. Comments on values are made as questions are discussed.




Methane is shown in figure 1 and data relating to this question is given in table 1.

The chemical idea being tested is that energy is needed to break bonds between atoms, but is released when bonds form. The answers expected to each part were:

. Methane and oxygen are stable as a mixture, but some molecules will break apart if energy is supplied. This starts a reaction between them.

. Energy is released when carbon dioxide and water form as products of combustion. Some of the energy is used to break up more molecules of reactants. This will continue until the supply of one reagent is exhausted.

. The energy comes from bonds forming between atoms of oxygen, carbon and hydrogen to make molecules of carbon dioxide and water.

In analysis, the responses to part c were considered first as this probed the chemi- cal idea under investigation. Responses given to the first and second parts are shown in each box in the table.

A very large increase in P-coded responses is found, significant at the 0.01 level (À2ˆ 118:2). One interviewee explained her changed thinking at interview.

She was asked to explain where energy comes from in the reaction:

S: From the bonds that have been formed . . . you break a few bonds, they form other products, they give out more energy, break a few more bonds and that keeps on going.

I: . . . Here [1st survey] you tell me that it’s those bonds [in methane] that are broken which give out energy, and here you say that it’s energy released when - [interrupted]

S: . . . I came to this college not understanding this completely . . . I thought if you broke the bonds you’d give out energy, which wasn’t true, because it confused me at GCSE . . . because everyone said ‘Energy is stored in bonds’ . . .

She went on to identify the ‘Developing Fuels’ unit as the source of her changed thinking. In company with around 33% of the group, this student changed response from a misunderstanding-type answer to the correct idea. An additional 12% made the same change between second and third surveys. Inspection of the response database shows that about 32% give correct responses at the second and third surveys, suggesting that most retain their knowledge.

However, while these data are encouraging, inspection of the response data- base shows that almost 27% gave T-coded responses on all three occasions, sug- gesting that these students learned no new material to prompt changed thinking.

Also, about 39% were T-coded for the first and second surveys and 30% for the second and third. This suggests that the misunderstandings are tenacious. One possible reason for the resistance to change is demonstrated in the following inter- view extract in which a student is arguing that bond breaking is both exo- and endothermic:

I: . . . how can energy be put in to break bonds, and yet you say energy comes from breaking bonds?

S: [Pause] How do you - [stops]

I: . . . what you’re saying [in your written answers] is that activation energy means you’ve got to put energy in to break bonds.

S: Yes.


Figure 1. Questions ‘Methane’ and ‘Energy change’.


I: If you’ve got to put energy in to break bonds, how can you also say here [in part 3] that energy comes from broken bonds?

S: . . . when you put something into it, it breaks the bonds releasing energy, and that energy that’s been released breaks bonds next and continues the reaction. . . This student thinks that energy is needed to break bonds initially, but once bro- ken, energy is released. This causes him problems when later in the interview he attempted to explain what happens when bonds form. He realizes that burning

Table 1. Changes in SAC students’ responses to ‘Methane’.

RC Description 1st % 2nd % 3rd %

P1 Energy is from bond formation 5.6 43.6 46.4

P2 Products are more stable than reactants 0.4 1.6 3.2

Part a: EA supplied / Energy splits molecules Part b: reaction is exothermic / chain reaction / O2

is unlimited

P Total 6.0 45.2 49.6

Q1 Answers as above to a and b only or to part c: 4.0 8.4 9.2 Energy is from bond making and breaking

Q Total 4.0 8.4 9.2

T1a Energy is from bonds in CH4 - 1.2 1.6

T1b Energy is from bond breaking | Energy splits molecules 4.0 6.8 4.4

T1c Energy is stored in CH4 12.8 16.0 18.8

Part a: as above (P) / Energy speeds up reaction Part b: as above (P) / CH4releases energy / excess energy is used

T1 Total 16.8 24.0 24.8

T2 CH4is energy store made from animals/sun 6.4 1.2 0.8 Part a: heat required / flammable H2present

Part b: CH4is fuel / flammable / O2available

T3 Energy is from CH4| Need O2, fuel & 5.6 0.8 0.4 heat | Fire ‘triangle’ kept going

T4a From burning CH4| CH4always burns in air / until 14.4 5.2 1.6 it runs out

T4b Heat energy from burning 4.8 2.8 2.4

T4c Heat is given out / from the flame 6.8 1.2 1.2

Part a: as above (P) / Heat required

Part b: as above (P) / Heat of burning / flame keeps reaction going

T4 Total 26.0 9.2 5.2

T5 From exo reaction | O2needs spark | gas has 2 8.4 2.0 2.4 flammable elements in it / is a hydrocarbon

T6 Only 1 part answered 18.8 7.2 6.8

T Total 82.0 45.6 40.4

U1 Uncodeable for all three parts 2.0 - -

U2 No response 6.0 0.8 0.8

U Uncodeable 8.0 0.8 0.8

Overall total 100.0 100.0 100.0

n ˆ 250

Key: The symbol / denotes ‘‘or’’; that is, alternative answers which are equally acceptable.

The notation (P) means that responses were identical to those coded P.


fuels releases energy into the environment, but is unable to link the exothermic character of the reaction with bond formation. After this he is asked:

I: So you still think that when you break bonds you give out energy?

S: Yes.

The key problem is that students find it difficult to appreciate that energy is released by bond formation, not bond breaking. This erroneous reasoning seems to begin pre-16, as at this stage students are frequently taught that fuels are ‘energy stores’. When at A level students are faced with the idea that bonds break to enable a reaction to occur, the most plausible explanation is that energy comes from the bonds and is ‘released’ on breaking in the way one breaks an eggshell to release the contents. These data indicate that about one-fifth of SAC students do not move from this thinking. Table 1 shows that about 19% give the response ` energy is stored in methane’ (T1c) at the third survey.

Energy change

‘Energy change’ (figure 1) uses the reaction between sodium and chlorine to explore the chemical idea that an upwards arrow on an enthalpy level diagram represents energy absorbed when bonds are broken, while a downwards one repre- sents energy given out when bonds are made and the difference in lengths meas- ures the enthalpy change of reaction. The expected answer is that diagram A best represents the reaction. This shows the largest difference between the arrows representing a highly exothermic reaction. Diagram C was also accepted as correct, as no scale is included. Data relating to this question are given in table 2.

These data imply that the energy release on formation of an ionic bond is not well understood by many students by the end of their A level course. Comparing these data with those for covalent bond formation explored by ‘Methane’ suggests that ionic bond formation is more problematic for students. Nevertheless, the increase in correct responses is significant at the 0.01 level …À2ˆ 19:6†. One student explained his changed thinking at interview, explaining that by the second survey he had a meaning for the arrows: ‘I’d seen the diagrams before and this is a violent reaction giving energy out so the arrow would be going down.’ As with

‘Methane’, most of the increase in correct responses occurs between the first and second surveys following the early work on thermodynamics in SAC. Here, inspection of the response code database indicates that most of the increase arises from students moving from Q-coded responses to fully accurate explanations.

Q-coded responses are the most popular type through all three surveys. One reason for this may be that students learn that activation energy is involved in chemical reactions and so may select a diagram showing what they think represents the activation energy needed to break a chlorine-chlorine bond. This idea is taught in the ‘Developing Fuels’ unit. The response-code database indicates that about 6% who gave this response (Q1) at the second survey change to a fully correct explanation at the third. For these students, the ‘activation energy’ answer appears to be a half-way point in their thinking.

A relatively high proportion seem able to select a correct diagram but offer no explanation, or one which is uncodeable (Q4). The response code database shows that these are different students at each survey. One reason for this response may


be the absence of numerical data in the question, which prompts students to select diagram C as a guess, for example:

S: . . . I didn’t know how to class this - this reaction whether it was a large amount of energy or small, or medium.

I: . . . so you just opted for medium?

S: Yes! [laughs]

Relatively few students selected the completely incorrect diagram at any survey.

Most who think that the long arrow means ‘lots of energy’ (S2) change their minds after the first survey. This suggests that by the second survey, most associate

Table 2. Changes in SAC students’ responses to ‘Energy change’.

RC Description 1st % 2nd % 3rd %

P1 A/ C Low EA, energy released on bond formation 1.2 4.8 8.0 P2 A/ C Low EA, exothermic reaction /stable 10.8 18.4 19.6

compound forms

P3 A/ C electrons lost & gained / low ionisation energy 0.4 0.4 0.4 and high lattice enthalpy

P Total 12.4 23.6 28.4

Q1 A/ C Energy required to break Cl2bond / start reaction 4.4 21.2 14.4 Q2 A/ C Energy required to heat Na / two states react 1.2 2.4 0.4 Q3 A/ C Violent reaction / react easily / low EA required / - - 0.4

Q4 A/ C Uncodeable / No explanation 16.0 17.2 18.4

Q Total 21.6 40.8 33.6

R1 A/ C Na & Cl are reactive 1.6 2.8 2.4

R2 A/ C It is an exothermic reaction 6.8 8.4 8.0

R3 B Energy is conserved in the reaction 0.4 - -

R Total 8.8 11.2 10.4

S1a B/ C Reaction doesn’t give out much energy 1.6 1.6 2.4 S1b B Reaction needs lots of energy to start 4.4 4.0 3.6 S2 B electron transfer / equation misunderstood 3.2 0.4 1.2 S3 A/ C 2:1 moles reactants / 1 bond broken, 2 formed 5.2 1.2 3.6

S Total 14.4 7.2 10.8

T1 B Bond breaking gives out energy - - -

T2 B Reaction is in equilibrium / energy levels are 3.2 2.8 0.8 equal / no. of bonds broken = no. of bonds formed

T3a B Long arrows => lots of energy made / used / reaction 6.0 2.0 1.2 is violent / Small gap => low energy input

T3b A / C A shows high energy barrier / lots of energy 0.8 2.8 4.0 required to start reaction

T3c A / C Bond breaking => energy is given out - - 0.4

T3 Total 6.8 4.8 5.6

T4 B Uncodeable / No explanation 10.8 3.2 4.8

T Total 20.8 10.8 11.2

U1 Uncodeable 0.4 0.4 0.4

U2 No response 21.6 6.0 5.6

U Total 22.0 6.4 6.0

Overall total 100.0 100.0 100.0

n ˆ 250


energy release with a difference in arrow lengths in energy level diagrams, even if this is at best on an intuitive basis. In contrast to ‘Methane’, few students respond that bond breaking gives out energy.

Covalent bonding

Covalent bonds

The question on covalent bonds was part of ‘Chemical bonds’ shown in figure 2.

The chemical idea being tested is that a covalent bond comprises two electrons,

Figure 2. Questions ‘Chemical bonds’ and ‘Methane molecules’.


one from each atom, sharing the same electron orbital. If two electrons are involved from each atom, a ‘double’ bond is formed. Covalent bond formation confers stability on the atoms. The expected answer, coded P, is that line 1 repre- sents a single covalent bond formed when a carbon and a hydrogen atom each donate one electron to form an electron pair between the two nuclei. Line 2 represents a double covalent bond in which two electron pairs are shared between two carbon atoms.

Table 3 shows an increase in frequency of P-coded answers over the three surveys significant at the 0.01 level (À2ˆ 114:2). By the third survey a majority of students described single/double covalent bonds in terms of the numbers of elec- trons involved.

The less precise Q-coded responses ‘single/double bond’ or ‘chemical bond’

decreased in frequency, although this was still given by 25% by the third survey.

Coding this response separately permitted monitoring of the use of more accurate chemical language. Around 17% changed from the ‘single/double bond’ answer (Q1) to a more detailed response between the first and second surveys. At inter- view, several students said their first answers used language learned at GCSE.

A relatively small proportion gave T-coded responses. The most frequent are shown in table 3. By the third survey, no student thought that an ionic bond was represented (T3a) and few who used the term ‘electrons’ made errors in the num- bers involved.

Examination of the response code database suggests that most of the correct second and third survey responses arise from students learning new information.

Table 3. Changes in SAC students’ responses to ‘Covalent bonds’.

RC Description 1st % 2nd % 3rd %

P1 2 / 4 electrons shared 8.8 43.2 51.6

P2 Single / double covalent 9.6 10.0 14.0

P Total 18.4 53.2 65.6

Q1 Single / double bond 43.6 33.2 24.8

Q2 Chemical bond 1.2 0.8 -

Q Total 44.8 34.0 24.8

T1 1 / 2 electrons shared 14.8 7.6 7.2

T1 Total 14.8 7.6 7.2

T2a Saturated bond 2.8 0.4 0.4

T2b Simple bond - - 0.4

T2c Weak / strong bond 1.2 0.4 -

T2 Total 4.0 0.8 0.8

T3a Ionic bond 4.4 1.2 -

T 3b Link between two compounds 2.0 3.2 0.8

T3 Total 6.4 4.4 0.8

T Total 25.2 12.8 8.8

U1 Uncodeable 9.6 - 0.4

U2 No response 2.0 - 0.4

U Total 11.6 - 0.8

Overall total 100.0 100.00 100.0

n ˆ 250


Although students’ responses changed markedly, at interview most did not attri- bute this to particular course materials or specific units, perhaps reflecting the fact that SAC features covalent bonds so frequently that students do not recall learning about them at any specific point. Also, the wide variety of contexts in which covalent bonds are met means no one unit acts as a memory aid. The drip-feed strategy seems to be the key factor prompting change.

Methane molecules

The question on methane molecules is shown in figure 2. The chemical idea being tested is that stability is associated with the formation of covalent bonds by which electron orbitals are filled by sharing a pair of electrons between two atoms. The expected answer, coded P, is that CH4 is the most stable of the formulae listed.

This arrangement confers the greatest stability on both atoms as their outer elec- tron orbitals are filled by sharing electrons. Data relating to this question are given in table 4.

Table 4 indicates that progression among students towards using ideas about molecular stability appears to be relatively limited, with 30% giving this response by the third survey. However, the increase in P-coded responses is significant at the 0.01 level (À2ˆ 22:8). Inspection of the response code database shows that about 6% change to the correct response between first and second surveys. This suggests that the SAC approach in which molecular stability features in the fourth unit, The ` Atmosphere’ , in the context of atmospheric chemistry, has had some effect. About 12 %change to P-coded responses between second and third surveys.

This corresponds with the revisiting of molecular stability in unit 8, ` Engineering

Table 4. Changes in SAC students’ responses to ‘Methane molecules’.

RC Description 1st % 2nd % 3rd %

P1a CH4is energetically most stable - 0.4 1.6

P1b C and H are more stable as CH4 6.4 10.0 16.4

P2 C & H need 4 & 1 more electrons for noble gas 6.0 8.4 11.6 configuration

P Total 12.4 18.8 29.6

Q1 C needs four bonds / 4 more electrons 46.8 60.4 51.6

Q2 C has valency / oxidation no. of 4 / 4 links 8.8 10.8 9.6

Q Total 55.6 71.2 61.2

T1 C has 4 bonding pairs / lone pairs 2/ 6 / 8 / 10 / 18 1.6 2.4 0.4 electrons

T2 C / H is saturated - - -

T3 More C in air / H travels in pairs /1:4 is equal mass 2.0 - - T4 Because there are 4 Hs and 1 C / other statement 9.6 1.6 2.0

T Total 13.2 4.0 2.4

U1 Uncodeable 6.8 1.6 3.6

U2 No response 12.0 4.4 3.2

U Total 18.8 6.0 6.8

Overall total 100.0 100.0 100.0

n ˆ 250


Proteins’ , where students learn about the shapes of simple molecules. This section features methane specifically.

Despite this increase, Q-coded responses type remain the most popular at all three surveys. The database shows that about 29% give a valency or carbon-only - type answers (Q1 or 2) at all three surveys, suggesting that these students knew about the number of bonds made by a carbon atom prior to starting the course and that SAC material did not prompt any change. Use of the term ‘valency’ increased during the study, and as this term does not feature in SAC, this must arise from teacher (or other teaching material) use.

Few responses were coded T. The response T4 is the most frequent of these, suggesting that students giving this answer think ‘because it just is’. The frequency of this answer decreased markedly by the third survey. The persistent popularity of Q-coded answers implies that for many, the limited answer ‘Carbon needs four bonds’ is sufficient.

If the levels of P-coded response to ‘Covalent bonds’ and ‘Methane molecules’

are considered jointly, we see that many more students can describe single and double covalent bonds in discrete molecules than can explain why four bonds should form in methane. The idea probed by ‘Methane molecules’ is an extension of that in ‘Covalent bonds’. The relatively low level of P-coded responses to the extension question suggests that the notion of molecular stability seems difficult to grasp.

Ions and ionic bonds

Sodium and chlorine

‘Sodium and chlorine’ is shown in figure 3. The chemical idea being tested is that an ionic bond may form between atoms when one or more electrons are transferred to make charged particles called ions. The ions bond together by electrostatic attraction, releasing energy and forming an ionic lattice. The chemical event is the reaction between sodium and chlorine. The P-coded answer is that in the reaction, one electron is transferred per sodium atom to each chlorine atom, result- ing in the formation of ions. Energy is released when the ions interact forming the sodium chloride lattice.

The frequency of the P1 response, that energy is released in bond formation, increases by 11% over the three surveys. The change is significant at the 0.01 level (À2ˆ 12:4). The relatively low level of this answer implies that most students may find it difficult to relate the observable events occurring in a chemical reaction to the energetics of bond formation. This is despite the formation of sodium chloride receiving explicit treatment in SAC in unit 1, ` The Elements of Life’ , and unit 3,

` From Minerals to Elements’ . The response code database indicates that the pic- ture is more complex than first inspection of table 5 suggests.

At the first survey, a majority of students state simply that sodium and chlor- ine are ‘reacting’ or ‘forming a compound’ (R1). About 17% maintained this response for all three surveys, suggesting that this group has not learned new material, or that these students cannot perceive a link between material studied and the question. About 36% changed from R1 to a different response at the second survey. Some moved towards a P-coded answer, including this student:


The chlorine and sodium atoms are rushing violently to get to each other to release or gain electrons and bond with each other. (1st survey, R1)

Na is reacting with the Cl2forming bonds releasing energy which keeps the reaction going on vigorously. (2nd survey, P1)

Others gave an S-coded answer on the second occasion, for example:

There is a displacement reaction occurring where the sodium is reacting with the chlorine. (1st survey, R1)

Figure 3. Questions ‘Sodium and chlorine’ and ‘Hydrogen chloride’.


[Drawing of Na and Cl forming a covalent bond] Cl needs one more electron to fill its outer shell and Na has that electron. (2nd survey, S1)

At interview, this student maintained his belief that the bond was covalent rather than ionic, and used the same idea to explain his response to another question,

‘Solution’, which involved the dissolution of an ionic lattice (reported in Barker and Millar 1999). The respondent believed that sodium and chlorine form discrete molecules. The increase in P-coded responses at the third survey was contributed mainly by students coded S or U at the second stage, rather than any further changing from the R category.

At both stages, the response code database indicates that some students move away from the P-coded response. About 7% changed from P-coded to R-type answers, as they did not use energy ideas at the second survey. A further 7%

changed in the same way between second and third surveys. These students may have given P-coded responses while relevant teaching was in their minds.

By the next survey, this was no longer immediate, so the answer ‘sodium and chlorine are reacting’ became the next most appropriate response.

The low level of S and T-type responses at all three surveys indicates the reaction between sodium and chlorine is well-known and unproblematic. It may also suggest that more explicit probing of students’ thinking is needed to explore misunderstandings in greater depth.

Hydrogen chloride

Students’ understanding of the formation of ions in solution was probed by

‘Hydrogen chloride’, shown in figure 3. Several chemical ideas are probed here:

Table 5. Changes in SAC students’ responses to ‘Sodium and chlorine’.

RC Description 1st % 2nd % 3rd %

P1 Energy is liberated in bond / ionic lattice formation 2.4 2.4 12.8 reaction

P2 e-transfer Na to Cl, stable compound forms / redox 13.6 15.6 10.8

P3 Ionic bond forms between Na and Cl 4.8 8.4 10.4

P Total 19.8 29.2 34.0

R1 Na and Cl are reacting / form a compound 52.8 45.2 46.8 R2 The element(s) is / are reactive hence violent reaction 2.8 0.8 1.2

R Total 53.6 46.0 48.0

S1 Hot Na has EA required / reaction is quick / endothermic 3.2 6.8 4.8

S2 Covalent bond forms 2.8 4.4 1.6

S Total 6.0 11.2 6.4

T1 Heat breaks bonds / sodium is burning 0.8 0.8 0.4

T2 Particles expand / contract / collide / break / split 2.8 1.6 -

T3 Heat energy used to make bonds - - 0.4

T Total 3.6 2.4 0.8

U1 Uncodeable 2.0 3.6 0.8

U2 No response 12.0 7.6 10.0

U Total 14.0 11.2 10.8

Overall total 100.0 100.0 100.0

n ˆ 250


acids contain hydrogen ions; in hydrochloric acid these form when hydrogen chloride molecules split into hydrogen ions and chloride ions (by heterolytic fis- sion) when the gas dissolves in water; the hydrogen ions are displaced as hydrogen gas when magnesium metal is added.

The context of this question mirrors that used in unit 3 of SAC, ‘From Minerals to Elements’, to explain acid formation. The notion of ions in solution is met again in unit 7 ` Using Sunlight’ , where students study electrochemical cells and the reactivity series in detail. Data relating to hydrogen chloride are shown in table 6.

Table 6 shows that the proportion of P-coded responses increases to 25% by the third survey, a value significant at the 0.01 level (À2 ˆ 25:7). We also find that the level of T-coded responses remains almost unchanged.

The positive influence of SAC in changing students’ thinking is confirmed by interview data. This student, for example, explained his change to a P-coded answer at the second survey as follows: ‘It was after we did all that stuff on electrochemical cells. We did all about ions in aqueous solution’.

Table 6. Changes in SAC sample responses to ‘Hydrogen chloride’.

RC Description 1st % 2nd % 3rd %

P1a H3O+ions | displacement reaction / ionic equation - 3.2 5.2 P1b H+and Cl-ions | displacement reaction / equation 6.0 10.4 11.6

P2 Hydrogen ions present | No response 2.0 2.0 8.0

P Total 8.0 15.6 24.8

R1 No response | displacement reaction / equation 0.4 0.8 1.2

R Total 0.4 0.8 1.2

S1a H3O+| Mg reacts with or displaces Cl2/ Cl-/ O2/ H2O - 2.0 2.8 S1b H+and Cl-ions | Mg reacts with / displaces Cl2/ Cl- 4.0 5.6 9.6

S Total 4.0 7.6 12.4

T1a HCl molecules | displacement reaction / equation 8.4 13.6 11.6

T1b HCl molecules | Mg reacts with Cl2 7.2 11.6 11.6

T1c HCl molecules | other explanation / no response 12.0 12.8 16.4

T1 Total 27.6 38.0 39.6

T2a HCl / water complex | any explanation 13.2 11.2 3.2

T2b HCl & H2O shown separately / H2O alone | any 1.6 2.8 2.0 explanation

T2 Total 14.8 14.0 5.2

T3 Cl+ & H-ions | incorrect equation / explanation 1.6 2.0 0.8 T4a No response | Mg reacts with acid / Cl-/ O2 10.4 3.6 1.6

T4b No response | H2released 6.8 2.0 2.4

T4 Total 17.2 5.6 4.0

T Total 61.2 59.6 49.2

U1 Uncodeable responses to either part 4.0 2.0 1.6

U2 No response 22.4 14.4 10.8

U Total 26.4 16.4 12.4

Overall total 100.0 100.0 100.0

n ˆ 250

Key: The symbol | denotes divisions between answers to parts of the question.


By adding P and S-coded answers, we see that 37% use ions in their third survey answers, compared with 12% initially. Nevertheless, given the explicit and detailed treatment the topic receives in SAC, this is a low figure compared to that seen for other questions in the survey and indicates that the ideas probed here may be problematic for students. The request for a diagram may have made the ques- tion seem unusual and difficult, dissuading students from giving their best possible answer. Also, the question was located last in the test paper, so students may have concentrated less closely on this than other questions. They may have instinctively copied the molecule diagrams shown in the first gas jar, rather than drawing ions.

Nevertheless, these data show that about 40% drew molecular hydrogen chlor- ide at the third stage (T1), while inspection of the response code database indicates that about 29% did so in all three surveys. One explanation is that students adopt molecular hydrogen chloride as a model for acid behaviour, which may be too persuasive to relinquish. By this model, displacement reactions are explained by the willingness of the metal to ‘swap partners’ with the chlorine, making molecular magnesium chloride and releasing hydrogen gas. Students add new information such as Standard Electrode Potential values to reinforce their model. The model works without any need to use ions and supports the finding from ‘Sodium and chlorine’ that students may perceive ionic compounds as discrete molecules.

Response S1 merits discussion. These respondents appear to understand the principle that acids contain oxonium ions, but think the reaction displaces a gas other than hydrogen. Table 6 shows that 24% (S and T1b) give the incorrect gas at the third survey. The notion is also persuasive, as chlorine and magnesium are widely known as two reactive elements, so students reason that this occurs because

‘magnesium is more reactive than chlorine’ instead of considering hydrogen.

Intermolecular bonds

Students’ understanding of intermolecular bonds was explored using three ques- tions; ‘Hydrogen bonds’, ‘Boiling’ and ‘Chlorides’.

Hydrogen bonds

This is the companion question to ‘Covalent bonds’, described earlier (figure 2).

The chemical idea being probed is that hydrogen bonds form between molecules in which hydrogen is covalently bonded to a highly electronegative atom, namely nitrogen, oxygen or fluorine. Hydrogen bonds increase the boiling and melting points of substances to higher values than expected from relative molecular mass values. The P-coded answers are that line 3 represents a hydrogen bond which arises because oxygen is highly electronegative and so a dipole exists in water molecules. The negative and positive regions of different water molecules form a bond. Differences between lines 1 and 3 include: line 1 is an intramolecular bond and line 3 is intermolecular; line 1 is longer and weaker than line 3. Data relating to this question are given in table 7.

This question shows a very large increase in the proportion of correct responses at the second survey which is retained at the third survey. The change is significant at the 0.01 level (À2 ˆ 133:5) and arises because students change their responses from all other categories. These data suggest that most students learn about hydrogen bonds in the first few months of SAC.


Initially, a relatively small proportion give a P-coded response, while a high proportion offer no response. This indicates that few students met hydrogen bonds at GCSE. This is to be expected, as GCSE does not normally feature hydrogen bonds. About one-quarter of first survey responses showed some understanding (category Q), probably based on guesswork; the description of the bond as ‘weak’,

‘liquid’, or involving ‘cohesion’ is approaching correct chemical idea. These responses are not completely correct, nor do they express a misunderstanding, so are coded separately. A small proportion gave T-coded answers at this point.

Table 7 indicates that up to 24% of students at the third survey suggest that the bond ‘is an attraction, not a bond’ (T1a). This is illustrated by one student who gave these responses:

The weak bond between molecules which signifies liquid or solid substances. (1st survey, Q2)

Hydrogen bonding. A small ionic force Line 1 is covalent, line 3 is a bit ionic. (2nd survey, P2)

A hydrogen bond. [The line] is an attraction, not a bond. (3rd survey, T1a) At interview this student explained her second survey response: ‘ I think I meant that it’s not really a bond, it’s a bit ionic but it does attract each other but it’s not really a bond’ and continued to explain that a bond is ‘when they’re joined together

Table 7. Changes in SAC students’ responses to ‘Hydrogen bonds’.

RC Description 1st % 2nd % 3rd %

P1 Hydrogen bond + explanation 6.0 34.8 40.4

P2 H-bond only - no explanation 4.0 23.6 28.0

P3 Intermolecular bond / polar attraction 7.6 6.0 0.4

P Total 17.6 64.4 68.8

Q1 ‘‘Liquid’’ bond + explanation 9.2 0.4 -

Q2 Weak bond between molecules + explanation 11.2 0.8 0.4

Q3 van der Waals’ / dipole-dipole bond 1.6 3.2 1.6

Q4 Cohesion / magnetic attraction / semi-permanent bond / 4.4 1.2 0.4 attraction

Q Total 26.4 5.6 2.4

T1a Line 3 is an attraction force not a bond / not a real bond 7.2 19.2 24.0 T1b Doesn’ t exist / is imaginary / temporary / repelling force 1.2 0.8 ±

T1 Total 8.4 20.0 24.0

T2 Triple / covalent / ionic / dative / delocalised / double / 6.0 2.8 2.0 1.5 bond

T2 Total 6.0 2.8 2.0

T3a Shows water moves around / is in suspension 2.0 0.4 - T3b Shows water is rigid / is stronger than a covalent bond - 1.6 0.8

T3 Total 2.0 2.0 0.8

T Total 16.4 24.8 26.8

U1 Uncodeable 7.2 1.6 0.4

U2 No response / Don’t know 32.4 3.6 1.6

U Total 39.6 5.2 2.0

Overall total 100.0 100.0 100.0

n ˆ 250


and wouldn’t want to come apart unless there was some chemical reaction or other’. This student appears to mean that hydrogen bonds cannot be considered to be bonds in the same sense as covalent and ionic bonds and that they should really be classified as ‘attractions’. The increase in frequency of this response is significant. The words ‘attractive force’ are used on only one occasion where SAC discusses hydrogen bonding, so the course does not emphasise this terminology.

Students therefore seem to acquire this phrase from their teachers rather than course materials. Another interviewee supports this:

I: How do you know a hydrogen bond doesn’t involve sharing of electrons?

S: Because we were told!

I: So there is a difference in your mind between a covalent bond and a hydrogen bond then?

S: Yes. A hydrogen bond is attractions between charges, different charges.

Nevertheless, table 7 suggests that a majority of students recognize and describe hydrogen bonds accurately by the third survey. One student who gave these two responses at the first and second surveys highlights this:

Ionic bond . . . (1st survey, T2)

There are hydrogen bonds when the H atom is slightly Ave (s) and O s¡. The‡ve and ¡ve charges attract one another and make intermolecular bonds. (2nd survey, P1) In explaining the change at interview she commented:

S: . . . this is the sort of thing I don’t think in my whole life I’ll ever forget after doing A level chemistry, because you just can’t help knowing so much about bonds.

The course materials had clearly created a great impact on this student, suggesting that the continued revisiting and renewed application of knowledge assisted her learning.

Figure 4. Questions ‘Boiling’ and ‘Chlorides’.


The role of hydrogen bonds in change of state was explored using ‘Boiling’ and

‘Chlorides’ (figure 4). Changes of state do not receive explicit treatment in SAC, but students learn that hydrogen bonds influence boiling points of compounds like alcohols, hydrogen fluoride and water.


The chemical idea being tested is that boiling is a change of state; involving, in the case of water, breaking hydrogen bonds between molecules. Gaseous water is called steam. The chemical event here is the change in state from liquid water to steam. The P-coded answer is that the bubbles contain steam, since the hydro- gen bonds break allowing the water molecules to separate from each other. Data relating to this question are given in table 8 and the question is shown in figure 4.

The change in the proportion of correct responses is significant at the 0.01 level (À2ˆ 17:3). Inspection of the response code database shows that about 20%

give the P-coded response on all three occasions and that about 32%did so at both the second and third surveys.

The responses coded Q do not express change of state ideas. Oxygen and air are dissolved in water, so these responses are not incorrect. However, these answers may hide misunderstandings in students’ thinking, as one interviewee illustrates. His written answers were:

Table 8. Changes in SAC students’ responses to ‘Boiling’.

RC Description 1st % 2nd % 3rd %

P1 Steam, water vapour, gaseous water 27.6 39.2 45.6

P Total 27.6 39.2 45.6

Q1a Oxygen 25.6 20.0 18.4

Q1b Dissolved or evaporating gas 2.0 2.0 2.0

Q1c Air 7.2 8.4 10.0

Q Total 34.8 30.4 30.4

T1 Heat, energy 0.4 - 0.4

T1 Total 0.4 - 0.4

T2a Hydrogen 6.4 8.0 7.6

T2b Oxygen and hydrogen 19.6 13.6 8.4

T2c Oxygen or hydrogen 2.8 2.0 2.4

T2 Total 28.8 23.6 18.4

T3a Carbon dioxide 2.8 1.2 1.6

T3b Gas 2.8 2.8 1.2

T3 Total 5.6 4.0 2.8

T4 Nothing / Vacuum - 0.4 0.4

T4 Total - 0.4 0.4

T Total 34.8 28.0 22.0

U1 Uncodeable 0.8 0.4 0.8

U2 No response 2.0 2.0 1.2

U Total 2.8 2.4 2.0

Overall total 100.0 100.0 100.0

n = 250


Hydrogen and oxygen gas. (1st survey, T2b)

AIR (Atmospheric, dissolved in the water). (2nd survey, Q1c)

At interview, the student explained: ‘hydrogen and oxygen, that’s the steam that’s coming off which is as hydrogen and oxygen rather than the bubbles’. He added that air would come out of solution as the increased temperature made it less soluble. Then he was asked:

I: So besides air, once you’ve boiled off all the air that was in it, what will the bubbles be then?

S: [pauses] I don’t know! Hydrogen?

The response ‘air’ hid his thinking that water splits up on boiling. The level of T- coded responses shown in table 8 may therefore be inaccurate, the true figures being higher.

The proportion giving T-coded responses decreases to 20%by the third sur- vey. The response database suggests that these third survey responses are given by students whose thinking has changed from the Q-coded ` oxygen’ /’ air’ answer at an earlier survey, supporting the point made above. The notion that water molecules break up on boiling and reform on condensing seems to be persuasive and difficult to change.


‘Chlorides’ probes ideas about intermolecular bonds other than hydrogen bonds.

The chemical idea investigated is that small dipole-dipole attractions (van der Waals’ forces) between covalent molecules require much less energy to break than an ionic lattice. The P-coded answer is that the vapour consists only of titanium (IV) chloride molecules because the intermolecular bonds between these require relatively little energy to break them, whereas magnesium chloride has an ionic lattice structure which requires much more energy to break up. Data relating to this question are given in table 9.

Although the increase in the proportion of P-coded responses is significant at the 0.01 level (À2ˆ 7:1), these data also show a rise in the level of T-type responses over the three surveys. Before exploring reasons for this, we will look first at one student who illustrates the move towards the P-coded response. She gave these written answers:

Because ionic bonding is stronger than covalent bonding . . . it will take more heat energy to break the ionic bonds . . . (1st survey, T1)

MgCl2has a much higher boiling point because ionic bonding is stronger as the ions bond in a continuous lattice . . . in covalent compounds . . . only the weaker intermo- lecular bonds need to be broken (covalent bonds are not broken). (2nd survey, P1) At interview she explained that her first survey answer was based on recall of a table listing ‘Type of bond’ and ‘Boiling point’. Thus, she associated covalent bond rather than ‘weak intermolecular bond’ with ‘low boiling point’, saying: ‘I probably thought here that you broke up the titanium [chloride] into that, and then it reforms in the vapour form . . .’ This student’s answer illustrates two points.

First, ideas learned pre-16 seem to contribute to the high level of T-coded answers, as students associate covalent compounds in general with lower boiling point figures than ionic ones. Hence, covalent bonds break when these substances


boil, and covalent bonds are ‘weaker’ than ionic ones. Second, this student con- firms the model for evaporation/condensation suggested above in ‘Boiling’, that molecules break up on changing to the gaseous state and reform as molecules during condensation. The breaking up is an essential part of the state change.

Some students apply this model to ionic solids too, as they picture ionic substances as discrete molecules with uni-directional ionic bonds between ions. As these are

‘harder’ to break than covalent bonds, ionic compounds have higher boiling points.

They thus make a direct comparison between ionic and covalent compounds, thinking of both as simple molecular structures. Such students are likely to strug- gle with the notion of a lattice structure for ionic solids in which there are no

‘intermolecular bonds’ of the hydrogen bonding or dipole-dipole type.

Inspection of the response database suggests that a significant proportion of A level students may complete their course with this faulty state change model.

About 12% who offered Q-coded answers at the first survey changed to T- coded responses at the second survey. New material on bonding studied between surveys may have prompted a shift from chemically acceptable thinking. Around 19%give T-type answers at all three surveys and 38%do so at the second and third, suggesting that these responses remain persuasive. These data imply that students learn ` covalent bonds are weaker than ionic bonds’ rather than ` covalent bonds require more energy to break than intermolecular bonds’ . SAC makes this distinc- tion explicit in unit 12, ` Aspects of Agriculture’ , but this was taught after the third survey was carried out.

Equally, though, this question may have been problematic because the need to use intermolecular bonds in responses was not explicit. Respondents are given

‘ionic’ and ‘covalent’ in the test, so adopted these as clues to the correct response.

Also, titanium (IV) chloride may be unfamiliar, so students were not prompted to

Table 9. Changes in SAC students’ responses to ‘Chlorides’.

RC Description 1st % 2nd % 3rd %

P1 Intermolecular bonds between TiCl4molecules break 0.8 4.4 13.2 P2 Intermolecular bonds in ionic solids are stronger 0.4 3.6 3.2

P Total 1.2 8.0 16.4

Q1 Covalent subs have lower boiling points / more heat 21.2 14.0 13.6 required

Q2 MgCl2only melts / lattice needs to break down 0.8 2.0 4.0

Q Total 22.0 16.0 17.6

T1 Ionic bonds can’t be broken by heating 12.8 18.8 14.8 T2 MgCl2ionises / is less reactive / already vapourised 1.6 2.8 1.6 T3 Covalent bonds are weaker than ionic ones, so break 24.0 28.4 30.8 T4 Covalent bonds are stronger than ionic ones 5.6 8.4 7.6

T Total 44.0 58.4 54.8

U1 Uncodeable 12.4 8.8 4.4

U2 No response 20.4 8.8 6.8

U Total 32.8 17.6 11.2

Overall total 100.0 100.0 100.0

n ˆ 250


consider intermolecular bonds as responsible for the boiling point variation.

Choosing more familiar compounds may have affected responses.

Changes to students’ thinking about intermolecular bonds as probed by these questions shows mixed progress. Responses to ‘Hydrogen bond’ suggest that most students can at least recognize these and explain their formation. However, only about half of the sample could apply this knowledge to explain water boiling.

Further, ‘Chlorides’ suggests that the effects of any other intermolecular bonds are difficult to understand, as most students cannot explain differences between two compounds in which hydrogen bonds do not feature. These questions also reveal that students may complete SAC with a faulty model of state change which involves covalent bond fission and reformation. This represents a contradiction:

students appear to know that hydrogen bonds are responsible for high boiling points, but argue that it is covalent bonds, not hydrogen bonds, which break during state changes. They are unable to link knowledge about hydrogen bonds or other dipole-dipole bonds to boiling points of chlorides.


The evidence presented here indicates that the SAC approach to teaching basic thermodynamics has had significant positive impact on students’ learning, enabling half of the students to demonstrate an understanding about where energy comes from in fuel-oxygen systems. SAC materials were cited in very positive terms by a number of interviewees who could recall specific contexts and chemical ideas in a clear and lucid way indicative of the impact of their learning experiences.


This study probed the basic idea that energy is released in chemical bond forma- tion. Evidence indicates that about half of the cohort involved learned this effec- tively, but that about one-quarter thought that bonds or molecules ‘store’ energy which is released when the bonds are broken, perhaps in the same way that the contents of an egg are released on fracturing the shell. This idea appears to resist change. One reason for this is the persuasive earlier teaching that fuels store energy, and the implication that oxygen is not really involved in the energy release.

These findings suggest that teachers should avoid using this notion, as it appears to cause confusion among students when they are taught more sophisticated ideas at a later stage. Ross (1993) suggests that combustion reactions are presented to students as ` fuel-oxygen systems’ to help students associate energy release with bond formation.

Chemical bonding

Although basic ideas about covalent and hydrogen bonding appear to be learned by a majority of students, ions and ionic bonding continue to cause difficulties. Some students seem to imagine ionic compounds exist as discrete molecules like as covalent compounds and therefore think of ionic bonds as uni-directional and subject to the same ‘rules of behaviour’ as covalent bonds. These ideas seem to originate in pre-16 teaching, which encourages students to link the low boiling points of covalent compounds to the idea that covalent bonds are ‘weak’ compared




Relaterade ämnen :