We like to talk about data. Big data, crowd sourced data, etc. Big data likes more data to better interpret or smooth the data statistically. But what happens when the data is wrong?
Simple example, all of the nutrient tracking apps, think MyFitnessPal and similar, are a mix of crowd sourced, vendor provided, and authoritative data. They generally do a good job on macros like protein, carbs and fat. A significant percentage of items are crowd sourced and I’ve found that they can be terrible counting at lesser nutrients. Potassium and Iron are good examples.
Authoritative sources I’ve seen show that a 100g of beef provides about 7% of a person’s daily potassium requirement and between 10 and 24% of a person’s daily iron requirement, depending on the cut. In many apps, both of these are regularly zero in nutrient apps.
So what happens when big data relies on inputs that are mostly wrong? I suspect you get a lot more wrong and the wrong becomes entrenched.