I understand what you're saying which is that the numbers scale up.
If you want to understand what raw data means you need to organise it conceptually. You can call it ideological or something else. What matters is whether the conclusions are appropriate and reasonable.
If there's some impropriety with the interpretation then we need to understand what that is exactly. Not clear on this.
What matters is that the conclusions are accurate. The mistake is to take data and filter it through your ideological and political framework. If you believe there is bias and are determined to find it, you will find it.
As I mentioned above, if I were a Christian and I believed that God had a plan for me I would interpret all my experiences as a sign of God interceding in my life. From the context of my world view the conclusion that god was sending me a sign would be reasonable and accurate.
It's not so much a matter of impropriety as it is an example of finding what you're looking for. This is exactly what happened when the DOJ released a report about school discipline. They assumed disparate impact was a sign of something more when that's not necessarily the case. When it was examined further it was determined that the discrepancy was a result of differences in history.