This is a companion to a presentation given to the UK Tableau User group on 5 July 2011 of the same title.
We visualise data for a variety of reasons, from data exploration to regular reporting. Both have their place when looking at HR data. However, what is common between purposes is a need to realise value from the visualisation, typically to enable users to understand, monitor and ultimately act upon the data.
HR and HR data possesses unique challenges. Firstly one must assume that the audience are less likely to be scientifically or statistically trained than some other groups. This results in two possible constraints (1) there can be a lower familiarity of visualisation techniques that would be common to other groups. Our experience shows that box plots are unfamiliar as an example. (2) there can be a lower understanding and vision of the possibilities to create personal value, to make the main job better or easier, through powerful visualisation of data.
One result that we see occurring from this situation we describe as the downward spiral of poor information design. As an individual who owns or contributes to the data is removed from the benefit that it’s careful creation can provide their personal interest in ensuring accuracy declines. Poorer data quality reduces the potential value of that data to the individual and the organization. Good information tools can serve to increase data quality and reverse the spiral.
There are three key considerations that need to be taken into account when designing information tools that add value to individuals’ work:
1) we must understand the activities and needs on the individual at a deep level. This must then be translated to understanding of what information is needed to inform decision making
2) we must enable interaction that anticipates likely follow-on questions and presents the answers in an intuitive flow
3) we must understand the principles of effective data visualisation. We must present data in a way to reduce the gap between data presentation and how to act.
The first two of these points imply that we must start not with the data, and how to communicate it, but with a deep understanding of the tasks and especially decisions that the users need to make. We use the terms ‘information design’ and ‘information tools’ as what we’re creating are tools that use historical information to guide decisions, giving confidence to facilitate effective actions. To get there requires data, statistical manipulation and an understanding of how users make decisions.
At the presentation we discussed our process for building these information designs. As noted, the most important stage is gaining a deep understanding of the users needs. We then user wireframes to design how the information could be communicated, use these to guide data requirements (often lots of calculations to transform the existing data to information that can trigger actions) and finally use usability testing techniques with the users to refine the designs.
When data can be transformed into stories or scenarios that can be communicated to help the users in their work, the enthusiasm for using data increases. Dry data on it’s own requires a certain type of hard-wiring and education to provide inspiration for action. Translating this to guiding information should be the goal of every information design.