|Table of Contents|
|Table of Contents|
The modularisation process
As we mentioned in section 4.2.2, there is no sharp distinction between the data and the conclusions, but a grey area, with the outcome of an experiment in the various stages of interpretation. In practice, it is possible to make the distinction between results and their interpretation, by focusing in the Treated results module on the graphical or tabular presentations and including in the Interpretation the discussion in terms of the theoretical model.
The idea of a Results module is based on the prototypical Results section. And indeed, the core of the Treated results module of the modularised version of A05 has been derived from that section. The original version of A08 does not include a section called Results; here the Treated results module with the theoretical results is derived mostly from the section 4.Calculations and the one with the experimental results from the section 5.Measurements.
The modules do not correspond exactly to the original sections. On the one hand, a lot of the information contained in the Results sections has been recast in the Interpretation modules. On the other hand, the Treated results modules represent a lot of information that has been derived from other sections as well. In the original version of A08, we have identified information concerning both the experimental and the theoretical results in the sections 6.Comparison of measurements and calculations and 8.Rotational coupling. The general shape of the differential cross section is presented and discussed in those sections, and the comparison of the experimental with the theoretical results includes a note discussing the reliability of the experimental results. In the modular version, the discussion pertaining to the interpretation of the experiment is given in the Interpretation module A08-m5bii, whereas the results themselves and the discussion on their reliability are also presented in the Treated results module A08-m4bii2.
In the original article A08, the data analysis techniques are described in the subsection 2.5.Data of the section 2.Experimental. In the modular version, this information has been represented in the constituent modules of the experimental Treated results module A08-m4bi, and in the mesoscopic module MESO-m3c-treat.
The problem how to analyse and present the raw data in an insightful way is addressed in an implicit `sub-problem-solving process': in the context of molecular beam experiments, the results can be obtained from the data in the way described in this module. The outcome of this process are the results presented in a figure. Many details on the analysis and treatment of the raw data are applicable to the results on differential cross sections in general. This warrants the creation of a mesoscopic module MESO-m3c-treat, which is connected to the Treated results modules containing results on differential cross sections by means of a link expressing the fact that the mesoscopic module is more general and more detailed.
The resulting module
We assume that readers are interested in `objective results' and try to interpret them using their own theories. For example, Bernstein and Levine have based an important theoretical article on Delvigne and Los's results of the differential cross section for sodium atom-iodine atom collisions that are presented in the Treated results module A08-m4b. Rather than taking into account the rotational coupling in the interpretation of the measurements, they have refined the application of the Landau-Zener coupling. Therefore, the different results are represented in different modules in order to allow the modular article to meet the communication criterion for efficiency.
The results in A05-m4bii and in A08-m4bii are copies from the interpretation module. These `intermediate results' are included in the Results module in order to allow for searching by the label `results'. These are direct copies and as such overlap completely with the interpretation modules where they came from. This overlap inconveniences neither the reader, nor the author. The author does not have to do much extra work to create these modules, and the comprehensive reader can simply avoid them, as the essay-type routes pass over the modules A05-m4bii and A08-m4bii.
In A05-m4b, we see massive overlap between the `parallel' Treated results modules, each with same type of results for different samples. These modules contain almost the same text. Information on the type of result, the methods used to obtain it and the reliability is repeated in each module dealing with a comparable results. However, the figures are very important here and those differ. The overlap between the parallels is a direct consequence of the criterion that the modules have to be self-contained: readers should be able to consult the modules separately. In order to allow the modular article to meet the communication criterion that there should be no unnecessary repetitions, those overlapping texts can be hidden from view, to accommodate those who consult more than one of the parallel modules. There is also some `hierarchical' overlap with the module summaries of complex modules which provide a `table of contents' of the constituent modules.
The core of the treated results are the differential cross sections of the scattering reaction, plotted against the scattering angle. These are represented graphically, once the raw data have been analysed and the variables have been multiplied with some function for the sake of clarity.
In the original version, comparable results are plotted in the same figure. For example, the experimental differential cross sections for and in Figure 5 of A05 (p.67). This not only saves space, but it also expresses the coherence and allows the reader to compare those results. This comparison is explicitly aimed at in Figure 11 of A08 (p.192), which presents both the experimental smoothed differential cross sections and the comparable theoretical ones.
In the modular version of A08, the comparison between the experimental result and the theoretical results is made in the Interpretation module A08-m5bii1. The experimental results themselves are also presented in A08-m4bi and the corresponding theoretical ones in A08-m4bii, in order to allow the reader to consult and manipulate these results separately. However, separating the results diminishes the cohesion. In order to restore the cohesion, we have connected the modules by means of a link that expresses the fact that these results can be compared. In addition, the reader has to be explicitly enabled to plot the results in the same figure, to facilitate that comparison. Therefore, the results should also be presented in a machine-readable format, allowing the reader to manipulate them. In the original articles, the results are not available in a machine-readable format, and therefore the modularised versions could not include such a presentation either.
Separating the raw data from the treated results and making the data-reduction methods explicit, allows the reader to obtain other results by using other data analysis techniques to treat the same data, or vice versa, by applying the same data analysis method to a different set of data. All dependency relations between the methods, the results and the interpretations are expressed by explicit links, so that the article can also meet the criteria for the clarity of the structure.
In order to be able to judge and then use the results, the reader has to be informed how reliable the results are, and to be convinced of that reliability. An account of the restrictions of the results includes an error discussion based on the restrictions of the raw data (m4a), and thereby on the restrictions of the methods, possibly on the reliability of the treatment techniques, and possibly on comparison to other people's results. In the original section 3.Results of article A05 (p.66), it is stated that ``The results for are less reliable due to apparatus effects''. According to the rules for writing a modular article, the notions of `apparatus effects' in A05-mb4i1 and A05-m4bi2 should be specified in greater detail, to allow the modular version to fulfil the reader's requirements.