Notes

On integrity of data, 06/2019


I find working with data to be an interesting dilemma.

What we might perceive as a set metric varies depending on how we measure it, and with what instrument. Instruments which are fabricated to produce a sought-after result. A result, and therefore a data point, which is intrinsically anthropocentric.

By specifying a threshold for data visualization, we are further manipulating data to fit within imposed artificial limits.

How far can we attempt to maintain integrity in data display when there is an underlying issue of facts being produced by humans? How relevant is integrity with data when there are no moral consequences?

Subtleties sometimes lie beyond sensory perception. But that which is not perceived shapes a narrative, much like the choices that weren't chosen. An event not perceived is still an event. I do not believe in discarding silent data.

I feel that representation must be true-to-reality, but what does this actually mean when this same reality is fabricated by humans, as it is in the case of data?
Is being consequent suffice? Is it that in this context it is nothing more than a direct simulation?