Consider the three charts above. Though all three charts were created from the same data set using the same date range, they appear to show very different results. The reason for this difference is that filters were applied to the bottom two charts to allow the editorial office to answer specific questions.
Interpretation issues can result from producing charts without listing the parameters used to create them. Firstly, it is not possible to tell if the last data point shown on the charts above represents a full year of data or only a partial year. This can confuse any possible trends, since some data are very sensitive to influences of partial years, such as the submission volume, the number of accepted manuscripts, the reviewer response to invitation conversion rates, etc.
If any filters were applied, that should be recorded directly on the chart. Many people believe it is sufficient to note the filters used to create the chart within the written report. Unfortunately, some people do not read the full text in an editorial report; they skim sections and glance at the charts, which makes it easy to misinterpret the data. Also, when the filtered chart is used in a presentation, it may be presented without additional information, such as parameters. The filters could be mentioned by the presenter, but that can cause confusion for the audience members if they do not hear those specifics relating to the chart or don’t understand them. Additionally, there is the possibility of the chart being stored electronically and separately from the written report, and then mis-interpreted in the future by someone who did not create the chart and does not understand the history of how or why the filtered chart was created. This can happen when someone new takes on the reporting role.
Parameters to include
Time Period: The time period makes clear the start and end dates of when the data were collected. This is especially critical when your chart includes partial years of data (e.g., current year).
Filters: Filters give you the power to drill down into the data to reveal trends, but they should be clearly described on the chart for the viewer. Some common filters to show on your charts include manuscript types, revisions included, decision types, countries of submission, and section category. Which filters need to be included will depend on the type of chart you have created.
It is best to always think of any chart that you create as a stand-alone object. The chart needs to be completely understandable without any supporting documentation. Including the parameters directly on your chart is the best way to ensure that your chart will be clearly understood, now and in the future!