The world has adopted the use of the term greenwashing to mean “green marketing [that] is deceptively used to promote the perception that an organization’s products, aims or policies are environmentally friendly.” The term can be traced all the way back to 1986 when Jay Westervelt wrote an essay on the reuse of towels by the hotel industry.
In that vein, I’d like to promote the use of the word datawashing to mean, “the biased use of data or statistics to justify a position that you really want to be true, or the use of complicated methodology to justify your role as expert in the analysis of data.” I’m not the first person to coin the term (e.g.) but its use is not common and most prior uses seem to have a slightly different definition.
In my field - urban data science - this might happen in regard to transportation options, when we often use data science to justify rail investments, or pedestrian and bicycling infrastructure, with the belief that people will switch modes from automobile if the infrastructure is only present. We justify this because we know that the automobile industry does the same with their billions of dollars of political influence, and has for a hundred years. Fortunately, academia and journalism alike are an open marketplace of ideas, and counter-arguments like this one are numerous on both sides of the argument. To be clear, I think this specific argument varies by situation and both sides can be right.
This is not to say that there are no insights in data - clearly there are. Nor do I think that the prior state of the world - let’s call it the Robert Moses state - in which political power was equivalent to right-ness was better. It was not. But in today’s data obsessed world I wish you could get a paper published that says “I looked really hard at this data and found nothing” or even “I ran really complicated models on this but it turns out what I did with Excel worked just as well.”
I don’t see datawashing going away any time soon, but I will stand on my soapbox for simplicity and transparency as the key to good analysis. Insights that are not accessible to advanced high school students are not as common as we make them out to be, and we should include as many people in the conversation as we can.