Cherry Picking Data

Decide what you want to prove, then choose the evidence to support that  position.

   It's a common practice, well known to prosecutors and defense attorneys. It's not limited to them, however. Investigators of all types, as well as proponents and opponents of issues, including advertisers, politicians, academics and journalists, have been guilty of this for generations.
   Politicians roam through academia recruiting experts to support their preconceived notions of what they believe is the correct side of an issue. Academics, meanwhile, rummage through the data trough, trying to make sense of the information mish-mash bubbling up to the surface. And with some regularity, they pick and choose the bits of information that would most help endorse their views.
    And, of course, journalists with an agenda use similar tactics. In an ideal world, journalists would be the last to cherry-pick through the information orchard. Unfortunately, many do, and it's up the readers and TV viewers to know who has an agenda and who's attempting to be objective. (True objectivity is probably not possible in humans, but they can try their best to be neutral.)
   So what's the best way to reach a truthful conclusion? Do you gather all the information you can and look for a pattern that establishes a principle, or do you start with a principle that you know (or assume) to be true and than add things that fit the pattern? In short, which do you start with, the principle or the pattern?
   In formal logic, the rules of which were set down thousands of years ago in ancient Greece, one is called Deductive Reasoning, where you start with a known principle, also known as a premise, and deduce -- that is, work down -- from that to establish a pattern.
   The other method is Inductive Reasoning, where you gather data and look for a pattern that establishes a principle.
   Of course, with the first you will encounter examples that don't fit the pattern suggested by the premise. Too often, these are simply ignored. This happens with what's called the Prosecutor's Dilemma. Police find a suspect and then look for evidence to support and prove their suspicions.
   A better method may be Inductive, where information about the crime is collected and examined for a pattern that fits only one suspect.
   So it is also with economics and politics. Gather the data, and a pattern will show itself. If you do not yet see a pattern, you may not yet have gathered enough data. And when you find something that doesn't fit what you discern as an evolving pattern, be cautious. You may have inadvertently and subconsciously outlined an erroneous pattern.
   Or you may be about to discover a totally new pattern. That's how scientific breakthroughs are made.
   On the other hand, you may be just collecting a mess of junk.

   Pick up a thread and follow it. Along the way, you can pick up other threads and weave them together. Eventually, either you will have a beautiful tapestry, or just a ball of yarn that the cat will ignore.