Data only has meaning when it is interpreted. Interpreting is the process of looking at what the data tells you, and exploring the “why” behind the numbers or stories. This step is often taken by researchers or evaluators, but when you want to use data for decision-making, it is much more useful for those who are involved with making the decisions to be the ones interpreting the data. They will most likely surface much deeper insights than an outside researcher can articulate and have a sense of ownership over those interpretations, making it easier to use the data to help make decisions.
At Spark, we know that when we interpret the results alone, the value of the information decreases: our partners who are embedded in the problem and its solutions understand what they are seeing in ways we simply do not.
During the interpretation stage, data is often not formally presented or documented in a polished manner. Bringing skillfully assembled reports, PowerPoints and charts to a group can decrease the sense of ownership and willingness to ask what else we can learn from the data. Instead, quick and dirty visuals showing the main points along with talking points presented by the people most involved with the data can be more than enough to trigger a productive interpretation conversation.
A few rules of thumb about what to bring to the dialogue:
- A clear understanding of what questions the data can and cannot answer. It is highly likely the group will surface some additional things they want to know from the data as they process it, and it is therefore helpful to know upfront what is and isn’t possible.
- A clear understanding of the facts in the data, but openness to what the facts mean. The person presenting the data needs to be comfortable and well versed when talking through what has been found as well as equally comfortable letting go of the “why” behind the findings. The less time you have spent personally trying to decrypt why the data is telling the story it is telling, the less defensive you may feel as others do the same exploration.
- A readiness to accept multiple interpretations. Rarely is there one underlying reason for why the data has generated a given set of results. There may be multiple reasons, some of which are highly relevant to the problem you’re trying to solve and the solution you’re applying, while others are less important or useful to consider. Don’t seek to find consensus, rather, just seek to explore.
Once the information is analyzed, it is not a heavy lift to take the time to interpret the results. Plan a small window of time (roughly 30 to 45 minutes) that can be part of existing meetings to review the results of the analysis as a small group including key leaders, staff, and/or partners. During this meeting, some of the questions to ask include:
- What does our data tell us in relation to the question we asked? For example, did we implement the way we expected? Did we achieve our desired outcomes? Did we surface insights about the population in need?
- How does the data align with our intuitive and experiential knowledge? How does it conflict? What are the implications of the ways it might conflict with what we already know?
- What else do we know that helps us to understand what the data is telling us? E.g. is there historical context, anecdotal stories, data from other sources, or other insights to explain some of the patterns in the data?
Someone should be responsible for taking notes during this meeting to document the “why” behind what the data is telling you. This feeds directly into the process of making decisions based on the information.