9 _______________ 10______________ 11______________ 12_______________ 13_______________ In Sentences 1 and 2 'Data obtained in previous studies1,2 using a fi xed on-site monitor indicated that travel by car resulted in lower CO exposure than travel on foot. According to Figo et al. (1999), the median exposure of car passengers was 11% lower than for those walking.2' the writer refers to the fi ndings and conclusions obtained by other researchers. 3.3.2 Key
114 Science Research Writing Why not start by describing my results? If you begin by describing individual results, the reader will need to build an overall scheme or pattern of your results by putting those individual results together. Th is is diffi cult for the reader to do; it is your job as a writer to arrange the information so that it is easy for a reader to process it. As with all subsections, therefore, it is more 'reader-friendly' to start with some introductory material. When you start any new section or subsection in your work the fi rst sentence(s) should provide a smooth transition for the reader between the new (sub)section and the previous one. Th ere are two good ways to do this: 1. You can begin by offering an overview of the current section. Th is is a description of the overall pattern or trend of the results. If you start with individual results and your reader puts them together 'bottom- up' to create an overall picture of what happened, there is a risk that each reader may end up with a diff erent picture of your results. It is preferable therefore to begin with general statements about what was found (in most cases, generally speaking, overall). Providing an overview enables you to show your reader the 'wall' before you begin to describe the 'bricks'. It is useful to remember that this type of general overview may need to be repeated when you move from one set of results to another. 2. You can begin by referring back to something from the previous section(s). For example, you can refer back to: • the general aims of research in this area (which you mentioned in the Introduction) • the specifi c research problem you are focusing on, or the aim of your project (which you mentioned in the Introduction) • the methodology • the original prediction or assumption to be tested (which you mentioned in the Introduction) • the fi ndings of other research in this area (which you mentioned in the Introduction)
Results — Writing Task 115 Why should I re-state my specifi c research problem or the aim of my project here? Eventually, in the Discussion, you will need to say to what extent your study (and in particular your fi ndings) solves the problem or fulfi ls the aim(s) you set out in the Introduction. Your results should therefore be very closely related to your aim(s); in fact, when you examine your results carefully you may even decide to go back to the Introduction and redefi ne the original aim(s) in relation to the results you obtained. Why should I re-state the original prediction or the fi ndings of other research here? Your results support, modify or contradict the original prediction, and they may support, modify or contradict the fi ndings of other researchers. By repeating the original prediction or the fi ndings of other researchers at the start of this section, your readers can see more clearly how your results relate to that prediction or those fi ndings. Your readers will not remember the earlier parts of your paper as clearly as you do. In Sentences 3 and 4 'In our study, modelled emission rates were obtained using the Traffi c Emission Model (TEM), a CO-exposure modelling framework developed by Ka.3 Modelled results were compared with actual roadside CO concentrations measured hourly at a fi xed monitor.' the writer refers back to his/her own methodology and adds more information about it. You may decide to refer to or summarise your methodology in your opening sentences. One reason for doing this is to highlight the important aspects of the materials, equipment or methodology you used to obtain your results. Another reason is to remind your readers of the methodology. You of course, remember it well — aft er all, it's your own research — but your readers don't share that familiarity. Also, extended details of the methodology are oft en given here rather than in the previous section,
116 Science Research Writing which may have included only the basic framework of the method. Th e more specifi c and complex the method, the more likely this is. Background information is as common and as necessary here as elsewhere. In this case, information is provided about the instrument(s) or equipment used to obtain the results (a CO-exposure modelling framework developed by Ka3). Later on in the Results section you may need to provide more factual information in order to explain why a specifi c result occurred. For example, perhaps a result was obtained because of a particular property or characteristic of the materials used, in which case it would be appropriate for you to off er information about that property or characteristic to the reader. As always, it is better to off er slightly too much background factual information than too little. Th e wider the topic — and therefore the wider the readership — the more background information you should provide, so that all readers can understand why the results occurred as they did. In Sentence 5 'Figure 1 shows the results obtained using TEM.' the writer invites the reader to look at a graph/fi gure/table etc. Why do I need to invite the reader to look? Surely they will see the fi gure if they continue reading down the page. What do you do when you are reading and you come to a sentence like this? You stop reading and take a look at the fi gure; you try to understand it or interpret the data you see in it; then you return to the text and keep that interpretation in mind when you carry on reading. If the data in the fi gure is very clear and has only one possible interpretation, it doesn't matter when you invite your reader to look at it. In this case, the results are very clear and easily interpreted, and so it is safe to let the reader view them before you comment on them. However, the data in many fi gures, tables and photographs can be interpreted in more than one way, in which case you should comment on the results in that fi gure before you invite the reader to take a look. If not, the reader may interpret it diff erently from you.
Results — Writing Task 117 In Sentence 6 'As can be seen, during morning peak-time journeys the CO concentrations for car passengers were signifi cantly lower than for pedestrians, which is consistent with results obtained in previous studies.2' the writer refers to specifi c results and compares them with those obtained in another study, using subjective, evaluative language (consistent with). Do I need to compare my results with those of other researchers? One of the aims of this book is to make you aware of the diff erence between the kind of writing you produced before you began to do your own research and the kind of writing you want to produce now. Until now, you have probably written reports for people like your teachers or professors, who know more about your research topic than you. You have performed experiments or simulations that have already been performed by other researchers and the results were therefore predictable in most cases. Your only task was to describe the methods you used and the results you obtained to readers who already knew what methods you should use and what results you would obtain. Now, however, things have changed, and in addition to reporting your results you should locate them on the 'research map' in your fi eld. Th is means that you need to show your reader how and where your results fi t in with the existing research picture, so you need to compare your results with those in the literature. You will develop this 'mapping' of your work more extensively in the Discussion, but in order to do it eff ectively, you need to fi rst set your results against existing results. What order should I present my results in? Th e order in which you present your results to the reader is very important. You may be impatient to present your most important results, but it may be necessary to start by describing the results which underlie or lead to the more important ones. Why do I have to use evaluative language — why not simply describe the results which are in the fi gure or table? As stated earlier, results do not speak for themselves. You do not have to use evaluative language in every case; sometimes results can be given
118 Science Research Writing objectively, either numerically or in non-evaluative language. However, if you simply describe what is in the fi gure or table, you have not added anything to what the reader can see for themselves — so why bother? Th e comments you make on your results infl uence the way readers perceive them. As we noted earlier (in Section 3.2.3), if you write As can be seen in Fig. 1, the two curves are very similar, the reader will focus on the similarity between the curves; however, if you write As can be seen in Fig. 1, the two curves are noticeably diff erent, the reader will notice the diff erence between them. In Sentence 7 'However, the modelled data were not consistent with parallel FOM measurements for aft ernoon journeys.' the writer off ers a general statement about his/her results to begin a new paragraph. In this sentence, the writer is moving on to more interesting, controversial results, and communicates this to the reader by starting a new paragraph and using a signal (However) at the start of the sentence. As a writer, you always know which results are interesting or signifi cant, but unless you communicate this to the reader by using a signal like this, all results will be perceived as having the same function and importance. In Sentence 8 'Although the mean CO concentrations modelled by TEM for aft ernoon journeys on foot were in line with those of Figo et al., a striking diff erence was noted when each of the three peak hours was considered individually (Fig. 2).' the writer refers to specifi c results and compares them to those obtained in another study, using language that comments on the result(s) (a striking diff erence). At some stage, you need to describe individual results in some detail, selecting results which are important, typical, or especially interesting. Isn't a word like 'striking' considered too informal? Defi nitely not. In science research writing, you do not normally use exclamation marks (!), even though you may feel that you want to if your
Results — Writing Task 119 results are very exciting. Instead, science writing uses a variety of words and phrases to achieve that 'wow!!' feeling, including striking. A full list of these can be found in the vocabulary section for the Discussion/Conclusion (Section 4.4). Notice also that by starting the sentence with Although, the writer helps the reader to predict the function of the sentence correctly. Th e sentence could also have been written as follows: Th e mean CO concentrations modelled by TEM for aft ernoon journeys on foot were in line with the FOM data but a striking diff erence was noted when each of the three peak hours was considered singly. but the reader would have had to wait until the middle of the sentence (but) to discover the function of the information in the fi rst part of the sentence, and may have needed to 'loop back' through the fi rst part of the sentence again in order to understand it. Signals are more useful when they occur early in the sentence. In Sentences 9 and 10 'It can be observed that during the fi rst hour (H1) of the peak period, journeys on foot resulted in a considerably lower level of CO exposure. Although levels generally exceeded those modelled for car journeys during H2, during the last hour (H3) the levels for journeys on foot were again frequently far lower than for car journeys.' the writer selects specifi c results to describe in more detail, using language that comments on the results (considerably lower, generally, frequently far lower). Should I explain my results as well as present them? Th at depends on the complexity of your results and the type of paper you are writing. Explanations can be given by providing background factual information to explain why a particular result occurred, for example, information about the properties of the material you are studying or the type of method you used. Make sure that you understand the diff erence between the explanation of a result (why it occurred as it did), the evaluation of a result (what the numbers mean) and the implication of a result
120 Science Research Writing (what the result suggests or implies). At this stage your explanations should be limited to fairly direct comments about your results; you will move on to broader explanations and implications in the Discussion/Conclusion. How do I know which results to describe in detail? Why not describe all of them in detail? If you describe all your results in equal detail they will seem to have the same level of importance. Th is is unlikely to be the case: some of your results are probably more signifi cant than others, some are typical, and some are key results whereas others may be of more peripheral interest. However, your sentences are, in the end, simply black lines on a white page — the reader cannot hear your voice and so cannot hear you emphasising the importance of a particular result. You cannot print it in red and, as we have seen, you cannot even use an exclamation mark. So choosing to describe a specifi c result in detail communicates to your reader that you consider that particular result to be signifi cant, worth highlighting or emphasising. It is interesting to note that the best results are oft en described in such a way as to give the impression that they are typical results — look out for this in the papers you read. Th is is commonly done by stating a generalisation followed by for example and then the result: '...the SFS results are in very good agreement with their FE counterparts; for example, at midspan the values are almost identical.' Don't be ashamed of the need to persuade; if you proudly or shyly stick to simple descriptions of your results using 'naked numbers', your reader may be surprised by your conclusions because you have not said what those numbers mean. Your reader may not agree with you, but s/he needs to know what you think about your results. In Sentence 11 'A quantitative analysis to determine modelling uncertainties was applied, based on the maximum deviation of the measured and calculated levels within the considered period.' the writer refers to the method used to analyse the results. Why wasn't this included in the Methodology? If you look at the Results sections in your target journals, you will be
surprised by the amount of methodology included in this section.
Results — Writing Task 121 Th e Methodology oft en only deals with the basic structure and compo- nents of the materials and methods. In such cases, most of the details are incorporated into the Results. Th is way of presenting information is becoming quite common in science journals. In Sentence 12 'Based on this approach, the average uncertainty of the model prediction for this study slightly exceeds the 50% acceptability limit defi ned by Jiang.7' the writer mentions a problem in the results and uses quantity language (slightly) to minimise its signifi cance. Do I need to mention problems in the results? Won't it make the reader doubt my results? As discussed in the Methodology, the opposite is true. Don't ignore problems in your results unless you are certain that the problems are insignifi cant and invisible. If your results are incomplete or some of them don't 'fi t', you should mention this, minimise its importance if you can, and suggest possible reasons for the problem/off er a solution. Failing to mention a problem suggests that you aren't suffi ciently expert to be aware of it, and this has a negative eff ect on your professional authority. By contrast, including a discussion of a problem in your work does exactly the opposite: it shows you to be fully in control of your research and able to evaluate it clearly. Furthermore, it provides you with an essential element for the Discussion/Conclusion: directions or suggestions for future research. As with problems in the methodology, if you delay writing up until your results are all perfect, you may never get to publish it. So write it up as soon as your results are worth communicating; don't wait for perfection. Mention and acknowledge the problems or diffi culties you encountered with your results while you are writing the Results section; it isn't appropriate to mention them for the fi rst time when you are discussing suggestions for future work in the Discussion/ Conclusion. So how can I talk about problems in the Results? Use vocabulary that minimises the problem, suggests possible reasons for it and/or off ers a solution or a way forward. In the example above, the writer acknowledged that there was a problem and minimised its eff ects
122 Science Research Writing (slightly). You can fi nd examples of the language you will need to refer to imperfect or problematic results in the vocabulary list in Section 3.4. In Sentence 13 'Nevertheless, these results suggest that data obtained using TEM to simulate CO exposures may provide more sensitive information for assessing the impact of traffi c management strategies than traditional on-site measurement.' the writer makes a reference to the implications and applications of the work s/he has done. Shouldn't that wait until the Discussion? An examination of implications and applications is certainly one of the central areas of the Discussion, but most writers give some indication of what their results mean, i.e. the implications of their results, towards the end of the Results section. Once individual results have been described and discussed, the focus of the paper or thesis begins to open out and move away from the central 'reporting' section towards the conclusion. A sentence of this type is very common at this point, using verbs like suggest or indicate. 3.3.3 Th e model Here are the sentence descriptions we have collected: In Sentences 1 and 2
the writer refers to the fi ndings and conclusions obtained by other researchers. In Sentences 3 and 4
the writer refers back to his/her own metho- dology and adds more information about it. In Sentence 5
the writer invites the reader to look at a graph/ fi gure/table etc. In Sentence 6
the writer refers to specifi c results and compares them with those obtained in another study, using subjective, evaluative language. In Sentence 7
the writer off ers a general statement about his/ her results to begin a new paragraph. In Sentence 8
the writer refers to specifi c results and compares them to those obtained in another study, using language that comments on the result(s).
Results — Writing Task 123 In Sentences 9 and 10 the writer selects specifi c results to describe in more detail, using language that comments on the results. In Sentence 11
the writer refers to the method used to analyse the results. In Sentence 12
the writer mentions a problem in the results and uses quantity language to minimise its signifi cance. In Sentence 13
the writer makes a reference to the implications and applications of the work s/he has done. We can streamline these so that our model has FOUR basic components. Like the Methodology model, this is a 'menu' from which you select those items appropriate to your research topic and the journal you are submitting to. 1 REVISITING THE RESEARCH AIM/EXISTING RESEARCH REVISITING/EXPANDING METHODOLOGY GENERAL OVERVIEW OF RESULTS 2 INVITATION TO VIEW RESULTS SPECIFIC/KEY RESULTS IN DETAIL, WITH OR WITHOUT EXPLANATIONS COMPARISONS WITH RESULTS IN OTHER RESEARCH COMPARISON/S WITH MODEL PREDICTIONS 3 PROBLEMS WITH RESULTS 4 POSSIBLE IMPLICATIONS OF RESULTS
124 Science Research Writing 3.3.4 Testing the model Th e next step is to look at the way this model works in a real Results section, and in that section (remember it may be called 'Analysis' or 'Data Analysis' instead) in the target articles you have selected. Here are some full-length Results sections from real research articles. Read them through, and mark the model components (1, 2, 3 or 4) wherever you think you see them. For example, if you think the fi rst sentence corresponds to number 1 in the model, write 1 next to it etc. Finite element modelling of sewer linings 4. NUMERICAL MODELLING 4.1. Th e fi nite element mesh Th e cross-sectional geometries of the three egg-shaped linings are defi ned by the joining of two circles of diff ering diameters by slightly curved segments tangential to the circles. In the case of the St and Ch linings, the circles were osculating and with (275, 140 mm) and (330, 110 mm) radii, respectively; the Ce lining consisted of nonosculating circles of (250, 115 mm) radii, their nearest points from one another separated by a distance of 105 mm. Measurements were carried out on the cross-sections of the three lining types so as to determine the radii of curvature of the somewhat fl at mid-section, a diffi cult task because of this fl atness. Th e thickness of the linings was found to vary along the cross-section; hence the mid-fl at section tends to become slightly thicker than the rest of the cross-section. However, this is not accounted for in the model because it was observed that the thickness also varies along the length of the lining, and accurate measurements are not practically viable. So, a thickness of 6 mm is adopted for the St lining, 8 mm for the Ch, and a 10 mm thickness for the Ce lining. Th ese thickness values are doubled at the hoop joints so as to simulate the actual junctions. Due to symmetry about the vertical axis (i.e. the y-axis) of both loading and geometry, only half of the cross-section is
Results — Writing Task 125 analysed (see Fig. 2). Moreover, because of symmetry about the x-y plane, only half of the total lining length is considered (see Fig. 3). Th e cross-section of the lining which is under study is situated at the x-y plane of symmetry, located at a distance equal to half the total length of the lining. Th e reason for restricting the calculation of the stresses and defl ections to this cross-section, is that full experimental data was only obtained at this lining location [8]. Th e element used in the analysis is an eight-noded isoparametric thin-shell element [6] with six degrees of freedom (i.e. three displacements and three rotations) at each node. Bending and membrane stresses are calculated at nodal points of individual elements and are then averaged for nodes which are common to more than one element. Finally, the mesh adopted consists of 180 elements subdivided into nine elements in the longitudinal direction and 20 elements in the hoop direction (see Figs. 2 and 3). Th e subdivision of elements in the hoop direction consists of six elements at the invert of the lining, six elements at the mid-fl at section, and eight elements at the top section of the lining (Fig. 2). Th e same mesh was used in the analysis of all three linings. In addition to imposing the relevant displacement and rotation constraints along the two planes of symmetry that allow only one-quarter of the lining assembly to be analysed, and full fi xity at the end of the pipe (i.e. all displacements and rotations are set to zero there), the boundary conditions corresponding to the fi ve restraint set-ups were readily simulated by reference to the nodes situated at the joint section. Th us, whereas both displacements in the plane of the cross-section and the rotation about the longitudinal axis were suppressed at all mid-section nodes for BCI , no constraint for any of the degrees of freedom at the joint was imposed in the case of BC5 (see Fig. 1). For the three intermediate restraint set-ups, displacements of those nodes in contact with the wooden segment(s) were suppressed. All analyses were carried out by reference to a value of (uniform) suction pressure equal to 1 kN m−2, and subsequent results, therefore, should be viewed with this value in mind.
126 Science Research Writing 4.2. Stanton and Staveley lining As mentioned earlier, a thickness value of 6 mm was adopted in the analysis of the St pipe, this becoming equal to 12 mm for the elements located at the joint section. Th e mid-length of the pipe encompassed by the FE mesh was 1200 mm, as measured from the end of the lining assembly to the central cross-section monitored during testing and presently under study. Figures 4 and 5 show a comparison between the experi- mental values of inner (hoop) strains and (transverse) defl ections [8], and the ones stemming from the FE analysis for all fi ve boundary cases. (Th ese and subsequent results are plotted against the vertical distance from the crown right up to the invert.) Here, the material properties used in the numerical model were selected from the lower range of values listed earlier, as this gave rise to a better correlation between the experiments and the FE analysis. Th ese material properties for the lining were chosen as follows: Eh = 11.5 kN mm−2, El = 5.5 kN mm−2, vh = 0.29 and vl = 0.14, where, in order to be consistent with Betti's condition, vl∙Eh = vh∙El, it was decided to adjust the value of Poisson's ratio by taking the lower range of vl (vl being considered, in general, a more reliable test value than vh) at 0.14, and then working out vh at 0.29, a sensible approximation, as argued elsewhere [8]. It is noticed from the results of the analysis that the hoop strains and defl ections resulting from the FE model follow similar patterns to the ones recorded from the vacuum rig tests. Moreover, it can be seen from Table 1 (which includes the maximum percentage error between the experimental results and the analytical predictions) that the FE model predicts reasonably well the critical values of defl ection at the mid-fl at section of the lining for the fi ve test cases. Th is was also true for the critical values of inner strain (occurring at the invert of the lining) for test cases