Assumptions are inherently necessary in the production of intelligence. They prime the pump and get the process rolling. They help fill in gaps and make sense of raw data when information is incomplete or ambiguous. However, by definition, an assumption is a belief or conclusion that is taken as true without proof. Assumptions can be based on inadequate or outdated information, or even be guesses. They can be dangerous in the world of intelligence, leading to bad decisions, missed opportunities and security breaches.
The problem with assumptions is that they can become self-fulfilling prophecies. If you assume that an adversary is going to do something, you may be more likely to interpret ambiguous information as confirmation of your assumption. This can lead you to make decisions based on faulty conclusions.
For example, if you assume an adversary is planning a terrorist attack, you may subconsciously seek out evidence that supports your assumption and overlook or misinterpret information that contradicts it. This could lead you to focus on the wrong information, in the wrong places. The result is that you may concentrate on reports of suspicious activity in areas where you think the attack might happen but miss real threats or take counterproductive actions.
Avoiding the drawbacks of assumptions
The best way to avoid the dangers of assumptions is to be aware of them. When you’re making an intelligence assessment, ask yourself what your assumptions are, where you got your information and how accurate and complete it is. To avoid the hazards of assumptions in intelligence:
- Challenge your assumptions – don’t just accept them at face value. Ask yourself if there are other possible explanations for the information you have.
- Be open to new information that might challenge your assumptions. The world is constantly changing, and so should your understanding of it.
- Finally, be aware of your own biases and how they might be affecting your analysis. We all have biases, and these can lead us to make assumptions that are not supported by the evidence.
‘Bias’ is the first cousin of assumption, and most intelligence failures come from false assumptions that are a direct result of individual bias. It can cause analysts to ignore alternative explanations and disregard information that does not fit ‘the script’. Information may be discounted because it does not suit the analyst’s bias towards a favoured solution, or overlooked because the analyst’s bias has resulted in too narrow a focus.
The risks of bias
According to Psychology Today, bias is a “tendency, inclination, or prejudice toward or against something or someone”. Biases, especially unconscious ones, have a negative connotation. They are part of the human condition, but can lead to distortion in thinking and perception, and ultimately to false assumptions and flawed analysis. They can be introduced in the way an analyst phrases their questions and be influenced by factors like their gender, ethnicity, or experience.
There are many different cognitive biases, and they all have the potential to affect belief formation, behaviour, decisions and research, and hinder accurate intelligence gathering. Bias can be introduced by an individual or the organisation, so there are both internal and external sources of bias.
Internal bias is generally a feature of the individual. Typical sources of internal bias include:
Evoked set reasoning. This is when the analyst relies on a predominant thinking approach, which is based on previous experience without exploring other possible explanations or new information that might contradict past experience.
Premature conclusions. Usually results from a desire for simple solutions and stability and a reluctance to be controversial, which can lead to premature closure of a problem.
Mirror-imaging. Defined as “the judging of unfamiliar situations on the basis of familiar ones”, this is the projection of the analyst’s values, beliefs and expectations onto the adversary based on the assumption that their imperatives are the same.
Authority Principle. This is where the analyst develops a particular preference for one source at the expense of others, which are generally deemed to be of lesser reliability. Relying on one authority avoids having to consider alternatives.
Lack of empathy. Lack of empathy for an adversary’s point of view results in an inability to appreciate what they are actually capable of doing and may lead to underestimating them.
Anchoring. The analyst’s thinking is dominated and over-influenced by another ‘expert’.
Ruling Theory. Internally, this is where the analyst shapes information to fit a pre-conceived theory or disregards data that disproves that theory or supports an alternate explanation. Ruling Theory leads to subjectivity and prevents any meaningful intelligence from happening, except by chance, because alternative explanations are not considered.
Sources of bias external to the individual can be found within the organisation and outside of it.
Parochialism. This happens when someone is overly loyal to the organisation’s rules or norms, causing them to focus only on certain things or stubbornly adhere to their previous judgments. It can lead to Group Think.
Ruling Theory. This is a dangerous external source of customer bias and may be ‘politically’ motivated. When intelligence doesn’t support the customer’s Ruling Theory, it may be ignored or discounted. However, it’s an analyst’s duty to share crucial “bad news” with the customer, even if it takes some courage.
Worst-case analysis. A symptom of excessive scepticism, pessimism, and overcaution, this can be a source of internal bias as well. It often results from previous experience and may reflect personal or organisational prejudices.
Confirmation bias. This is a type of cognitive bias that favours information that supports what we already believe, and it’s one of the most common biases that can impact how different sources of information are analysed.
Confirmation bias is selective thinking where information that confirms assumptions and preconceptions is:
- Automatically noticed
- Actively sought
- Accepted without reservation
And anything that contradicts assumptions and preconceptions is:
- Automatically ignored
- Not sought
- Rejected out of hand
In the collection planning phase, confirmation bias can lead the analyst to assign more weight to information that confirms the favoured hypothesis, whilst information that could disprove it is ignored or discounted. To compensate for this, analysts should conduct an analysis of competing hypotheses to disprove rather than confirm them.
Everyone has biases, but it can be challenging to see them in ourselves. Even when we do recognise our biases, it’s not easy to do anything about them. But just knowing that biases exist can help reduce their influence.
Consequently, it’s vital to view and redefine a problem from various angles, often in a team setting. There are also a number of methods that can also be used to counter inherent bias, and these include:
Paraphrasing the question. The question can be reworded without changing its intent and can trigger new perspectives and insights.
Turning the question through 180 degrees. Taking the opposite point of view can, sometimes, be surprisingly effective.
Broadening the focus of the question. When a question is too simple or narrowly focused, it may be a symptom of customer bias, and may prevent consideration of alternative hypotheses.
Changing the focus of the question. Changing focus entirely is difficult because it requires more thought and creativity, but it can be the most productive.
Networking. Involve colleagues and customers in the analytical process, because fresh minds will often identify flaws in an argument. Use a team approach involving techniques such as brainstorming and cognitive mapping. Consider peer review or dialectics to identify and counter personal bias.
Honing assumptions and overcoming bias in intelligence
Most intelligence failures result from false assumptions and individual bias, rather than the absence of information. Left unchecked, assumptions and bias can cause an analyst to discount, miss, or worse still, ignore key information. To build solid, reliable assumptions – the foundation of successful analysis – all existing information must be given equal consideration and internal and external bias must be addressed.
But analysts rarely have the time to manually translate, deduplicate and process all the information available to them, and struggle to evaluate assumptions, overcome bias and produce quality intelligence.
Advanced tools like Silobreaker can automate the collection and processing of structured and unstructured, open-source, deep and dark web data, and finished intelligence. This saves analysts time and effort, allowing them to focus on higher-level tasks, like hypothesis testing and spotting trends. Silobreaker also enables seamless information sharing and discussion among analysts and teams. This collaboration can produce a more well-rounded and objective analysis.
Silobreaker’s capabilities empower analysts to access a comprehensive range of information, build solid assumptions, and overcome biases effectively, ultimately leading to the production of higher-quality intelligence. This, in turn, supports better decision-making for organisations and stakeholders.
Learn more about how you can overcome bias in intelligence and use tooling to produce high-quality, relevant and timely intelligence with confidence.