I use this phrase often in evaluation of my work. It is a subtle reminder that my mental impression of the world is unique to only me. Having learned long ago (but still probably too late) that the world does not revolve around me, considering only my point of view becomes potentially dangerous to producing quality action.
I compare data interpretation to reading a book. Through language, you can understand a narrative. Through numbers, you can do the same. When reading a story, you cast a mental impression of the narrative based on your experiences, imagination, and beliefs. When you read data, you cast the same mental impression on your understanding.
Datasets, or the collection of figures, numbers, and information, are values that represent tangible results. Simply, datasets are facts. A dataset on its own cannot tell you your next action, the task to change, or the person to follow up with. Datasets need people to provide analysis. As analysis is interpretation of facts, it allows opportunity for bias to exist. We are human, after all.
One of the more common forms of bias that exists in analysis is confirmation bias. At every exploration of the answers to questions, a hypothesis is formed. More often than not when you gather information around a hypothesis, you have an understanding of what that information will tell you before reviewing the facts.
Biases are an evolutionary trait that benefit us in analysis, to make quick deductions of what is presented to us. However, combining bias with an expectation of what information should be present often causes your attention to gravitate to the facts that support your expectations. This leaves data uninterpreted as you continue to unconsciously select the information that confirms your expectations. As a result, your narrative is formed without taking everything into consideration.
To combat confirmation bias, it is best to have your work reviewed and discussed. As your interpretations are your reality, someone else’s interpretations will be theirs. It is valuable to have a group of people with a variety of experience, beliefs, cultures, and ages on your team for this reason. Through analysis you find a solution, through collaboration you find truth.
For large decisions in your position, do not be afraid to ask for help and be transparent. Show your work, how you got to your conclusions, and be open to other’s interpretations. Just like in a book club, you are encouraged to share your interpretation among the group. Find peers that can be a part of your data club and do the same!
Perception is reality until proven otherwise. Connecting and sharing information with others broadens your perspective and encourages a continuous state of learning. While you may not introduce wine to the conversation, like some book clubs do, your data club can be intoxicatingly real and open. As they say, in vino veritas!