En el trabajo de esta semana nos han vuelto a pedir que analicemos una visualización e intentemos mejorarla con nuestras aportaciones. Este vez he decidido hacer los sketchs a boli.
The purpose of the chart is to show the relationship between the various scientific publications and the citation to make between each other. At first glance the visualization is very cool. You can select the data at a journal level and at a category level. The journals are divided in four categories levels, the first problem that arise is that we don’t know what are the name of this categories and what are the reasons for this grouping.
The display shows two types of data for each publication:
- EigenFactor of each journal. How this value is calculated isn’t disclosed neither the visualization page neither a link with information about the EigenFactor. We understand that is a value that must be known by the scientific community. If we want to reach a bigger audience we should explain exactly how this value is calculated. In the visualization the value of this factor is represented by the surface of the circle. With this type of representation is difficult to calculate the relations with the rest of the journals. Without doubt it will be more clear you use a simple bar char to compare this value.
- IN/OUT citation: This factor make a relation between the number of citations that comes from a journal and go to another. In the visualization don’t appear exactly how is calculated this value so it’s difficult to extract information about this point.
The interaction with the display occurs at two levels: You can select the sector of a journal and it appear highlighted all the journals of that sector and the relations with the journals that send / receive a citation with that journals. The other option is to select a journal and the appears all the journals that send / receive a citation with it. As the size of the journal in the visualization depends of the Eigenfactor it’s hard to select journals with a small Eigenfactor. The citation relation between the journals is made with the thickness of the lines. This is another point that make the comparison very difficult. It’s practically impossible to look for correlations and patterns in the visualization.
To improve this part of the visualization I propose to include a scatter plot chart. The idea is to select first a journal to analyze. The new chart will include the IN / OUT data in the X / Y Axis. The size of the bubble associated will be represented by the EigenFactor. I believe that this kind of graphic will show more easily the patterns that we’re looking for.
I’m not a user related to the world of scientific publications. With the visualization shown I can’t draw any conclusion or story. As a reader the questions that I ask me are: What are the most prestigious journals? Are the same journals cited between them? are there more important factors that the EigenFactor, maybe the numbers of articles, the age of the publication or the country where the journal is published?