Applied Thinking for Intelligence Analysis: A Guide for Practitioners (2014) [pdf]
airpower.airforce.gov.au
airpower.airforce.gov.au
There are two things I looked for there and did not see. These are standard issues in intelligence analysis.
1. Duplicate confirmations. If you have two reports which agree, are they from different sources, or from the same source via different paths? If the latter, it's not a confirmation.
2. The difference between noise and deception. Some information is noisy, and statistical methods for filtering may help extract signal from noise. Some information is deliberately deceptive, and noise filtering will not help. The average will still be biased.
We've reached the point in politics where you need to know this just to read news.
The paper does mention the strong human tendency (all the higher mammals, really) to see patterns in noise. This is a useful survival instinct, in that fleeing before you're certain there's a threat is a benefit. It's a big problem in complex situations. It's also the basis of religion. There are so many classic examples of that in intelligence history. The paper mentions over-analysis of V-1 bomb targets, although I'm not sure that was a real thing at the level of people who had access to maps of the hit points. The inaccuracy of the thing was clear; "Greater London" was about as good as it could do. Air bases were never attacked; too small a target. If the V-1 had been accurate enough to hit air bases, it would have been used against them in the Battle of Britain.
There is, however, a classic example of that mistake. During the 1960 Cuban Missile Crisis, US analysis of aerial photos of the missile sites showed a layout more suitable to the Russian tundra than a tropical island. So there was a theory that the USSR now wanted the US to recognize them as missile sites, following a long period of deception as the equipment was moved into place. That was considered an open threat.
In 1987, there was a sort of Cuban Missile Crisis reunion, with people from both sides present. That subject came up. The officer who'd been in charge on the Soviet side said, "No, we just did it that way because that was what the field manual said to do". Everyone on both sides who'd ever worn a uniform nodded in understanding.
The late Dick Heuer (referenced a few times in this paper) wrote the Psychology of Intelligence Analysis, the gold standard: https://www.cia.gov/library/center-for-the-study-of-intellig...