Forum Replies Created
-
Posts
-
I rarely use box-and-whisker plots and histograms in the day-to-day of my job but I have encountered them in presentations of clinical trial data. These data presentations are normally developed by other members of my team so I have not yet had experience in creating one other than in this course.
I do find them easy to understand and especially beneficial to show the statistical results of a clinical trial rather than solely reading a statistical analysis. Having a visual representation of the distribution of the data certainly makes it much easier to draw conclusions and present these conclusions to others who are not specifically involved in the clinical trial.
The best type of graph to show ranking would be a bar graph so the reader is able to quickly view the size of each category in reference to one another.
For the class dataset, the service category with the most frequent claims is the same in both the Northern and Eastern Zones – Curative Care visits. In the Eastern zone, the service categories with the least frequent claims are Inpatient Medical and Inpatient Obstetrics. In the Northern Zone, the service category with the least frequent number of claims is Speech Therapy.
To demonstrate the Flores del Mundo membership trends over the course of the year, I would use a line graph to show the overall trend. The reader would be able to get a clear sense of how the data has changed through time. If the total membership was split into two datasets to look at the Eastern and Northern membership over the course of a year, I would still use a line graph to clearly summarize the membership trend and to allow for easy comparison between the two regions.
To graph the total membership vs. employee/spouse/dependent, I would chose a bar graph. When re-reading ‘The Right Graph’ article, the author shows an example of two sets of categorical data which could be used to easily display the total membership in addition to a bar for each subgroup (employee, spouse, or dependent). I would suggest dividing the time into quarters as the author did to ensure the graph is easy to read with the multiple bars at each timepoint.
Although the ‘WHO Maternity Dashboard’ does include some color coding to help the reader draw conclusions, this appears to be more of a report rather than aligning with Few’s definition of a dashboard. This ‘dashboard’ does not clearly allow the user to monitor information at a glance and may be better presented in graphs focusing on each performance measure to clearly identify trends. The dashboard also includes blank columns where either historically information or upcoming months of data is not included. This should be left out of the monthly until the data is available to present to the reader.
