back next next
Table of Contents
Index
Glossary














BR>














































































Table of Contents
Index
Glossary
back next next

Next: m5 INTERPRETATION Up: The characterisation by the Previous: m3 METHODS

  
m4 RESULTS

In our domain, the methods are used to generate results, which are then interpreted. This process is reflected in the prototypical sections IMRDC, in which the Methods are followed by the Results and then by the Discussion of those results. Analogously, we distinguish the modules Methods, Results and Interpretation.

The module RESULTS contains the results (the raw data and the treated results) generated with the different methods described and discussed in the Methods module.

The reason why the modular model strictly separates the presentation of the results from the presentation of their interpretation, is that this separation allows the reader to can consult and use the results as such, independently of the author's original interpretation, and interpret them in his own way.

Nevertheless, it will often be difficult to distinguish a separate Results module. Although the results, and in particular the raw data, are supposed to be `objective' and uninterpreted, they do depend on the context in which they were obtained.4.13 Also, it is often difficult to make a sharp distinction between the raw data and the interpreted results; there may be a grey area of results in the various stages of interpretation (see also [Buxton and Meadows, 1978] and [Dillon, 1991]). When experimental data  are generated in an attempt to test a theory, that theory determines, at least to some extent, how the experiment is set up and executed. The resulting raw data must be analysed, i.e. filtered, synthesised, and otherwise interpreted. For example, in an experiment where a lot of noise is generated, the theory is used to distinguish the ``real" events from that noise. In order to be suitable for a clear presentation, the analysed data are then converted into other units, e.g. multiplied with functions or plotted on a logarithmic scale. These various presentation forms also depend on the expectations of the readers and thereby on the context of the discussion.4.14 The interdependency of experiments, theories and results is inherent to the nature of scientific research. Hence, the results always must be linked explicitly to the method and tools that were used to get them, and they must be presented with their restrictions.

As a rule, the Results module is a complex module.  It can be a cluster of modules with similar, more specific results, which are distinguished by means of the domain-oriented characterisation. It can also be a compound module consisting of complementary components: in experimental research, there may be raw data, in addition to treated results that are presented tabularly or graphically. In that case, the Results module contains the constituent modules Raw data and Treated results. Each of these constituent modules may in its turn be a cluster module containing constituent modules distinguished using the domain-oriented characterisation of the information.



m4a Raw data    

The module RAW DATA contains:

In order to allow the reader to manipulate the data himself, the raw data should be presented in a manipulable (in practice, machine-readable) format. Intermediary data can be generated in the course of some measurements or calculation. These intermediary data then can be copied or extracted from their original module (Methods or Interpretation) and included in this module.

The restrictions of the data are based on the restrictions of the methods used to generate these data. For example, when the raw data represent the output of a detector for some scattering reaction, the reliability of the data depends on the resolution of the detector. In an error discussion, the accuracy of the data is therefore connected to the precision of the methods. Restrictions can also pertain to the range of the data, resulting from the range of the methods.

In some domains, the raw data are already nowadays published in a database (for a brief discussion, see [Reichhardt, 1999]). Examples are the data on genome [Cameron, 1998], astronomical data [Boyce, 1998] and crystallographic data [Kennard, 1998]. In those fields, the module Raw data could simply contain a link to the appropriate address in the database.



m4b Treated results  

The module \m{TREATED RESULTS} contains:

The treated results can be presented in a format facilitating reader's comprehension, such as in a figure or in a concise table. They can also be presented in a manipulable (in practice, machine-readable) format.

Intermediary results can be generated in the course of some measurements or calculation. If they are to be retrievable as results, they must be characterised as such. That can be achieved by copying them from their original module (Methods or Interpretation) to include them in this module

This module contains a report on the restrictions of the results in terms of an error discussion pertaining to the range and the precision of the results. These restrictions are based on the restrictions on the raw data (represented in module m4a), and thereby on the restrictions on the methods used to generate the data, as well as on the restrictions on the methods to treat the data. The discussion can include comparison with results obtained in other work.



Next: m5 INTERPRETATION Up: The characterisation by the Previous: m3 METHODS