It is imperative to be aware of the influence of statistics and how they can be utilised to alter the message of a dataset. The adage "lies, damned lies, and statistics Opens a new window" illustrates the potential for statistics to be employed to deceive and misinform. To avoid this, it is essential to be vigilant of the potential biases that can be introduced by selective omission of data and to ensure that charts and other forms of data representation are lucid and easy to comprehend.
For instance, a graph depicting five years of data can be used to present a process as being easily manageable. However, including an additional nine months of data, immediately preceding the original chart may reveal that the five years of exemplary performance has only been achieved through rigorous control. This alters the interpretation of the initial graph and implies that the performance is almost unpredictable, fluctuating between meeting the Service Level Agreement (SLA) and perfection.
It is also crucial to consider the audience when presenting data: not everybody can interpret numerical data or charts with equal ease. Your managers are dealing with the reality of the situation, while your peers are only viewing the reports once a quarter or even less frequently. They are heavily dependent on someone interpreting the graph for them, so it is important to ensure that the meaning is clear and easily understood. A blue arrow with the text "Better" eliminates any doubt about whether the performance line being above the SLA line is positive or negative.