# Psychology of Intelligence Analysis Author: [[Richards Heuer]] ## Review The author is a former CIA intelligence analyst who observed the pitfalls of doing analysis and came up with some ideas to overcome them. Analysis can be improved in a number of obvious ways: collecting more and better information, more concise and clear writing, asking better questions or streamlining the analysis process. His book does not discuss any of those instead addressing what was an underappreciated factor of analysis at the time of writing. He investigates how our mind's constraints and biases influence the outcome of our analysis. In recent years this has become a popular topic with books like "thinking fast and slow", or "superforecasting" but they are light on practical advice. This book (and others from the intelligence community) are more oriented towards actionable advice. The author begins by examining mental processes and how they put constraints on our ability to analyze complex and uncertain situations. These constraints create cognitive biases. He describes studies and research to give examples of how cognitive biases factor into analysis. Unfortunately awareness of these cognitive biases does not help us to mitigate them. To overcome our limitations we need a process to guide our analysis. He emphasizes the need to come up with perspectives that are outside of our mental model of the world, region or situation. - To bypass the constraints on mind and memory, we need to externalize complex pieces of analysis. - Unless we are unfamiliar with the topic, gathering more data is not a good strategy. It increases our confidence in our prevailing hypothesis but adds no signal. It is more valuable to create a number of hypotheses first and let them guide our search for information. In a study with medical students evaluating patients, those who formed hypothesis and tested them along the way performed better than those who tried to gather as much information as possible beforehand. - Start by generating hypotheses without judgment. It is difficult for people to see alternative perspective so consult others in this process. Every analyst has their own set of mental models through which they observe and analyze the world. Challenging our assumptions and beliefs through a process of evaluating alternative beliefs is his most important idea. - After coming up with hypotheses, try to disprove them rather than looking for confirmatory evidence. The most likely hypothesis is that for which there is the least amount of disconfirming evidence. - When communicating results, be clear about uncertainty, confidence and process of analysis. Vague language means different things to different people and they tend to fill it in with their assumptions. Analysts also need to be clear about what assumptions are being made. The only way of getting better is to have precise language and transparent analysis allowing for feedback on what was right and wrong. - Identify milestones that should be tracked to evaluate the performance of the analysis. There are two particularly important types of new information that we need to react to: changes in variables that are used in our model of the situation; and information that tells us that our mental model is not right for the problem being analyzed. There are many awesome insights and recommendations for organizations and analysts. I can see myself reading through this again at a later time to brush up on ways to improve. I strongly believe that analysis and decision-making skills will become one of the most sought after skills in the coming years. Many fields are starting to measure decision-making performance but there are no general frameworks for best practice. Great ideas and insights are spread across a wide range of books and fields (intelligence, investing, medicine, science, history & law) that still need to be incorporated into an overarching framework. Heuer's insights into the process of analysis align with what I have learned so far, and help me to develop my framework further. I found his process of analysis with an emphasis on alternative perspectives very compelling and will try to incorporate them into my process going forward. ## Key Ideas ## Related - [[Superforecasting - The Art and Science of Prediction]] - [[Thinking in Bets]] - [[Framework for analysis and decision-making]]