In our previous blog, we discussed a definition for Data Literacy. We summarized it as the combination of knowledge, skills and mindsets that allows individuals to find insights and meaning within their data to enable effective, informed decision-making. That’s all very well, but what does this mean when applied in a business context?
To find the answers, we need to look deeper at the process behind turning raw data into insights. For people looking to improve Data Literacy skills, this means aligning their development with specific competencies to make that process more tangible and accessible, as well as ensuring that people have the opportunity to apply their skills.
You’ll notice that we differentiate between competencies and skills. Most people correctly interpret the term skill as something specific you learn. But just because a skill has been learned does not mean it can be applied properly. In contrast, competencies are a broad set of knowledge, skills, and mindsets – being competent combines the acquisition of a skill with the mindset and behaviors to successfully execute it. In the context of Data Literacy, there are many situations where a person may have specific skills, but are unable to apply them because of their mindset or the culture of their organization. For this reason, Data Literacy should be discussed in terms of competencies, so that people are enabled to apply their skills appropriately – much of this is about changing mindsets.
These distinctions aside, Data Literacy is not achieved by mastering a uniform set of competencies that applies to everyone. Those that are relevant to each individual can vary significantly depending on how they interact with data and what part of the data process they are involved in. For example, the competencies required by someone responsible for cleaning and preparing data for analysis are different to those required by someone responsible for interpreting the results of the analysis. To establish the key competencies associated with Data Literacy, it is helpful to explore them through the lens of the full data analysis process. Below, we’ve outlined the six key steps that turn data into decisions – and the competencies linked to each of them:
The first step defines the purpose of the decision being made. It starts by establishing the various use cases and expectations that organizations have with data, before putting processes in place to ensure the right questions are being asked that align with business goals. It’s important to lead with human thought and creativity here – after all, you know what your organization is trying to achieve by making this decision. Data should then be used to refine, validate and inform these thoughts. Although many businesses follow the data as a first port of call, this often prevents the application of critical, creative and collaborative thinking.
For this step to be successful, it’s crucial to identify the question, problem, or goal being addressed and whether it can actually be solved using data. If it can, then set appropriate key performance indicators as a foundation and turn business questions into corresponding analytics questions. It’s also important to translate these business requirements into a structured hypothesis.
Next, there are specific skills and mindsets required to understand the concepts related to the question at hand, acquire the relevant data and prepare it so it is in the appropriate format. Once you know the issue you are trying to solve, it’s much easier to filter out irrelevant data. After the problem or question is identified, relevant concepts and data must be identified and prepared in a way that makes it accessible and suitable for further use.
This means understanding which data points are appropriate for the task at hand, using a systemic perspective. It also means developing an overview of how to retrieve data, as well as how to join, clean, transform and create new data. This enables you to use the raw materials at your disposal to more deeply explore the challenge at hand.
This step is where the real analysis comes in. The skills and mindsets required to do this effectively will allow you to infer insights like causality and correlation, but given the huge volumes of data used by today’s businesses, analytics tools are a crucial ally to achieve this. These come in many forms – whether simple descriptive analytics, diagnostic tools to enable root cause analysis or advanced predictive analytics.
From the perspective of the data practitioner, competency at this step means knowing the right analytics and statistics to use, learning how to assess patterns and trends across diverse data, performing comparative analytics, visualizing data, probing for causality and ultimately, testing hypotheses and drawing inferences.
So now, your analysis is complete and your insights compiled. But the data you have been processing only means so much in isolation. Before confidently presenting the insights in a way that answers the original challenge or question, it’s crucial to consider the skills and mindsets required to interpret what you have found and work towards drawing a conclusion. This starts by challenging your assumptions, mitigating fallacies and bias, understanding the implications and consequences of your insights and examining additional perspectives. It’s important for organizations to understand how the insights they have uncovered could be misleading if taken out of context before considering the final course of action.
To carry at this step effectively, it’s important to be confident when interpreting data – whether visualized or verbalized –to identify a course of action. Issues of data provenance, such as understanding ethical issues and possible bias, are also crucial to ensure any assumptions are challenged.
Despite all the work and time invested so far, it means nothing if you can’t create the arguments needed to communicate data-based decisions. All stakeholders impacted by the decision need to understand and support the results that led to the conclusion that was made. There are a range of skills and mindsets required to be successful here, including how to identify the right stakeholders, how to verbalize the decisions and actions and how to effectively use visualizations to share the data insights in a tangible and accessible way.
Finally, no good process is complete without an evaluation phase and that is no different for data analysis, to consider the impact and effectiveness of the decision after it has been acted on. This includes ensuring there were no unintended consequences and that the outcomes were positive in addressing the question or problem at hand. If the results were unexpected, it’s important to reassess the information used and whether any of it could have been missing or misleading that prevented a more informed decision.
As we previously discussed, the competencies required by each individual vary depending on their role in the data analysis process. However, all skills and mindsets are underpinned by a core understanding of Data Literacy. No matter what step in the process, everyone interacting with data should have this baseline understanding – including which properties and classifications the data represents (such as quantitative or qualitative), as well as the various levels of measurement of data (nominal, ordinal, interval and ratio). With these skills, anyone can apply themselves to any step of the broader data analysis process.
In the next blog in this series, we will examine the specific competencies within each step in more detail and explore the knowledge, skills and mindsets related to each of them.
For more information on how to upskill across each of these competencies, visit Qlik’s Continuous Classroom to check out the courses and resources available.