SAP SuccessFactors People Analytics Stories Data Visualization Series: Selecting the right chart Part 3 – Trend and Distribution Charts
In Part 1 of the series, we talked about various charts supported in SAP SuccessFactors People Analytics Report Stories. In Part 2, we focused on comparison charts and their usage. In this blog post, we would try to cover the Trend and Distribution Charts and their application in details.
We will reuse the same example organization “BestRun Pvt Ltd” and various individuals in the organizations we had introduced in Part 2.
Best Run Pvt Ltd is a large global multinational headquartered in the United States. Let’s look at part of it’s HR Organization Structure which we would be used to demonstrate use cases for the charts.
Trend charts are used for demonstrating a change in a variable across a time series.
Distribution charts are used for demonstrating the distribution of values within a dataset.
Let’s look at some of the best practices while using various Trend and distribution charts
- Unlike area charts, these are charts suited for showing trends across discrete data points.
- It is ideal for comparing time-based data points and help in identifying pattern and trend across the given time frame for the respective measure.
Example – Eugene, VP Recruitment and Onboarding wish to review Internal Move In and Internal Move Out for Manufacturing Business Unit across the years and see if his multi-year initiatives taken in 2017 after witnessing mass internal move-outs have helped in reducing move out and improve internal team satisfaction.
Stacked Area chart
- Stacked area charts are suitable for dimensions that are continuous in nature or we have a very large set of discrete data points.
- Stacked Area chart sometimes can be confusing if we have 2 many measures as visually comparing areas with different baseline can be confusing for viewers.
Example – Max, VP Employee Relations is looking into understanding absenteeism in the best and worst-performing teams across 2018 to identify initiatives for reducing the same.
- Box Plot is commonly used to show the dispersion of data across 5 major aspects.
- Minimum value
- Maximum value
- Quarter 1 percentile
- Quarter 2 percentile
- Quarter 3 percentile
- Outliers are displayed as separate data points.
- This chart should be used cautiously as it expects readers to understand the concepts of dataset value dispersion and hence should not be used for the audience not well versed in it.
Example – Jenny, HR Head America, wishes to analyze the performance of all the Recruiters reporting to Eugene understanding minimal, maximum and quartile-based performance in terms of job requisitions that have been opened in the last one year. The following graph shows 10 out of 129 are outliers for filled status.
- Heat Maps help in understanding the distribution of values across more than one dimension with color scale used for showing changes across the spectrum.
- The aim is to discover patterns across the scale of dimensions used by analyzing the color spectrums observed.
- The maximum and minimum of the values act as 2 ends of the color spectrum.
Example – Max, VP Employee Relations is trying to analyze terminations in an organization in a given the Fiscal year 2019-20 w.r.t Separation Reason and Gender.
- This chart is often used to display multivariate quantitative data on a two-dimensional plane. For each observation values across the dimensions are joined to give the star-like appearance.
- It also helps in the analysis if we are witnessing clusters on values across a specific dimension.
- One benefit of the chart is outliers are visible predominantly which also skews the chart towards the outlier reading.
Example – Steve, CHRO is analyzing section performance ratings of individuals reporting to him w.r.t those who have completed and those who are in modify stage.
- Treemap shows the distribution of hierarchical data as a set of nested rectangles.
- The difference between a Treemap and a heat map is a treemap shows the variation in measure value by changing the size and color for the specific dimension value.
- Treemaps though require a single measure and single dimension.
Example – Steve, CHRO is looking at how his workforce is distributed across countries
The following example even though maybe best supported on a Geo Map but for the lack of current support Geo Map a good alternative we have chosen is TreeMap which shows distribution with size varying across size and color.
In the next article, we will try to cover the remaining charts and best practices associated with them.
You can also refer to the official SAP Analytics Cloud Reference chart type guide.
Note: All images are SAP Copyright and created in Report Stories in test company instance for this blog post