Category Archives: Hypothesis 3

Academic Impact – Comparison between all subjects investigated

Figure 1: A Comparative Visualisation of the Academic Impact for all the subjects chosen for investigation (Isabelle Blackmore)
Figure 1:  A Comparative Visualisation of the Academic Impact for all the subjects chosen for investigation – achieved using Excel
(Isabelle Blackmore)
Comments

This visualisation indicates (quite nicely) that Biology and Psychology have the highest Academic influence out of the subjects investigated and the metric indexes used. These conclusions are strengthened further by the parallel between the trends shown in both the JCR’s Impact Factors and the h5 Index. Such findings would be in line with the notion that scientific subjects such as Biology and Medicine are primarily experimental-based entities. Experimental methodologies are prime academic patents that require consistent and accurate citation, thereby naturally predisposing the subjects built upon these methodologies to a higher citation count.

However, it is this predisposition to higher citation counts held by some subjects over others, that limits the validity of the conclusions one can draw from this visualisation. These are standardised measures that don’t take these predispositions into account. One citation in a discipline prone to many citations (such as Biology) is worth comparatively less than one citation in a subject that is not predisposed to them (Philosophy). Citation scores have relative meanings depending on which academic context they occur in. In order for our visualisation to be more informative, it must also take this relativity into account.

One surprising find was the low representation Physics had in the world of Academia. One might assume that – given its history as a majorly influential academic subject on humankind (the atom bomb and space exploration to name but a few) it would be more influential. On the one hand, its low representation may well be due to data collection biases: failing to pick journals that were truly representative of the discipline. On the other hand, this may provide a genuine insight into the true academic prestige of Physics. Owing to the nature of Physics as a highly esoteric subject, it might be that this academic isolation is also reflected in the impact factors. Furthermore, one characteristic I noticed when researching and analysing these articles is that the majority of research in physics is of a theoretical and not an experimental nature. Without this experimental bedrock, there is less of a predisposition to cite other work, something that is firmly embedded within other science subjects.

As a final comment, although the data and the visualisation is far from a perfect representation of reality, it was certainly more objective than the linguistic datasets we compiled. Whereas we had to use our own judgement in selecting ‘technical’ words for our investigation, these citation metrics were indisputable numerical values and could therefore be reproducible – a characteristic that remained questionable for our linguistic datasets.

Written by: Isabelle Blackmore

Network Visualisation

Hypothesis 3, Images, Visualisations
NXFINALlayout2
Figure 1: The Network of Knowledge. Each node embodies a different academic subject, the size of which is proportional to that subject’s average h5 Index. The edges connecting the nodes represent the connectedness of each subject, the width being proportional to extent of shared vocabulary between each subject.

Comments:

Perhaps the most technically challenging aspect to this project, this visualisation is by far the most informative. Throughout this blog we have been arguing that knowledge is structured by virtue of a rhizome network of vast, non-hierarchical connectivity.  It was therefore paramount that we attempted to embody this in our own attempts to investigate the relationship and interconnectivity between academic subjects.

The conclusions we can draw from this visualisation are that, out of all subjects investigated, Biology, Medicine and – to a lesser extent – Psychology had the highest influence in the world of published academia as represented by their large h5 Indexes. This conclusion is relatively unsurprising given that these were scientific, experimentally led subjects. Given that experimental procedures are effectively knowledge patent’s that always require proper citation in order to avoid the acquisition of plagiarism, it is unsurprising that these should be the most cited subjects.

Perhaps the most important asset to this visualisation is its demonstration of linguistic connection between disciplines – one of the main aims of this project. One surprising conclusion suggested in the visualisation is the large degree of language shared between Political Science and History, larger than that of Biology and Medicine, which seem like the obvious pair. However, on closer reflection, History is a largely political subject, making not that surprising at all. One might go as far as to argue that – be it the mirco-level events deciding the governance of one country or the macro-level drivers of world wars – History is fundamentally shaped by Politics.

With regards to Hypothesis 3: that the relationship between a subject’s shared vocabulary and its impact in the published world of Academic exists and it is a directly proportional one, our network doesn’t seem to provide any clear evidence supporting this notion. The width of the edges stemming from both Biology and Medicine, the two most highly ranked subjects in terms of academic influence, appears to be very similar to that of the edges connecting other, less highly ranked, subjects. Furthermore, the thickest edge – thereby the greatest level of shared vocabulary – was between History and Politics, neither of which had a significantly high level of influence. Therefore, one might conclude that it is necessary to reject this hypothesis and look for alternative explanations for the extent of connectivity, hence interdisciplinarity, between academic subjects. Alternatives may include fundamental overlaps of content – as is the case with Biology and Medicine – or intrinsic structural links, as already mentioned with regard to History and Politics. Needless to say that, before any definitive conclusions or rejections of the hypothesis can be made it is necessary to conduct more in depth research in order to verify our findings.

Although a great technical achievement, there are still some aspects of the visualisation that might be improved, namely the size and the spread of the network. There is a lot of information encoded within the network (the size of the edges for example) that are somewhat obscured by a slightly compressed overall presentation. Furthermore, colour might also be used to further emphasise the distinction between different inter-subject relations. However, as was the case with most elements of the project, the process of compiling and coordinating the data required to create this visualisation was incredibly labour intensive, requiring effort from every team member. This therefore limited the size of the network that can be observed. Ideally this visualisation can be improved by including a wider range of academic disciplines within our investigation. Yet, given the practicalities of this endeavour, it was simply unfeasible at this point in time. Furthermore, for the sake of simplicity it was decided to use just one of the many available Citation Metric scores (the h5 Index) as an indicator of a subject’s academic influence. Not only is this a very one-sided approach, it also raises more philosophical questions about the way in which we measure influence and its impact in the knowledge economy.

Visualisation: Rain Soo Jamin, Comments: Isabelle Blackmore