Seeing data
Plotting data graphically allow you to literally see more than simply looking at values and summary statistics. The key to ‘seeing’ is being able to recognize pattern in a plot, and how plot design influences what pattern is shown and emphasized, and what is not – ‘graphic literacy’. Apply different graphic approaches to look at data from different perspectives and to focus on different ‘aspects’ (e.g. the line and area graphs used to visualize retailer stores). Pattern must be considered in relation to how data was collated and consistency with expectation. Outliers and abrupt change may be erroneous or ‘true’, but require scrutiny to determine which. The magnitude of values or unexpected relationships between variable may be indicative of incorrect scaling or other processing error, but may also be real, requiring a re-evaluation of your underpinning conceptual model.
There are many excellent books and resources on visualizing data that cover data design, software usage, and the power of graphics to accessibly communicate the pattern and hence information in large complex datasets. The need for data literacy has existed since humans first began recording harvests and other information necessary for society to operate. Data and information extracted from it has become more and more important through time culminating in our digital data-driven age that has only just begun, as AI, quantum computing, human-computer interfaces, and social media portent. In an age of mass information and disinformation data literacy is an increasingly marketable skill in environmental management and more widely.
A research blueprint
An aim of this tutorial is to inspire similar comparative research into the progress of organizations towards sustainability goals. UK companies are required to publish emissions data in their annual reports but many also publish environment, social and governance (ESG) report which report other sustainability KPIs. These KPIs are company specific, but there are problems across a sector so there is some commonality (e.g. all supermarkets generate food and packaging waste). Data availability will vary with jurisdiction. A common denominator is required to scale absolute values to intensity metrics. In the tutorial supermarket emissions were scaled by estimated store floor area using a secondary data set, but such data will often not be available and meant intensity could not be calculated for the online retailer Ocado. Gross company turnover is a universal denominator that can be used to scale any KPI.
