Implicit bias

It is fundamental to the proper administration of justice and to public confidence in the judicial system that all parties are, and believe they are, treated fairly and without discrimination. Judicial officers must ensure that any personal biases or prejudices, or tendencies to stereotype or make false assumptions about people do not influence their decision making. This duty extends to ensuring transparency and particularity in their judicial reasoning.

The legal system provides mechanisms for enabling challenges to impartiality by reasons of apprehended bias or conflict of interest. Where there is proven or even the appearance of bias, a judicial decision may be set aside. Decisions made by judicial officers in circumstances where a fair-minded lay observer might reasonably apprehend that the judicial officer might not bring an impartial mind to the resolution of the question, may be set aside, unless waiver or necessity is established.

In recent years, there has been significant research into how mental short cuts, known as heuristics, affect the way decisions are made. Researchers have considered how implicit bias (sometimes called unconscious or unintended bias) can affect decision-making. Implicit bias may affect an individual’s understanding, actions and decisions in an unconscious manner, and operate without the individual’s awareness or intentional control. Seminal research by psychologists, Daniel Kahneman and Amos Tversky, identify two systems of thinking (intuitive or fast thinking; and conscious or slow thinking) and examine the biases (or systematic errors) of intuitive thinking. Intuitive thinking suppresses doubt and ambiguity in order to quickly assemble coherent interpretations. Conscious thinking, on the other hand, accommodates doubt and incompatible possibilities in the process of deliberation.

Set out below are some of the key implicit biases identified in the literature; they can apply across all facets of decision-making. Judicial decision-making in domestic and family violence related cases frequently involves identifying factors affecting risk, assessing risk, and exercising a heightened awareness of the cautions associated with typologising, stereotyping and making false assumptions about parties, their circumstances and their motivations. In understanding the potential for bias in the context of judicial decision-making, and learning, through self-examination and reflection of any particular personal biases, judicial officers can develop strategies for eliminating bias from their decision-making.

Representativeness v base-rate information

When making a judgment about facts, there is a tendency to ignore base rates and doubts about the veracity of a description and favour a judgment of representativeness. Representativeness results from a focus on aligning a description with a stereotype. Base rates are relevant statistical facts. In many cases, there is some truth to the stereotypes that govern judgments of representativeness, and the predictions that follow may be accurate. In other situations the stereotypes are false and the representativeness misleading, especially if the base-rate information points in a different direction.


When estimating the likelihood of an event or outcome or the frequency or prevalence of a category (eg people, behaviours), it is likely that the estimate will be based on the instances most easily retrievable from memory; if retrieval is easy, the estimate is likely to be high. What is more obvious or more vivid will appear to be more probable.


When making a decision there is a common tendency to rely too heavily on the first piece of information offered. For example, information may be offered about a numerical value. This information tends to have the effect of anchoring the decision to that numerical value, even where that value may bear no reasonable relation to what the value ought to be. Once an anchor is set, there is a bias toward interpreting later information around the anchor.


Logically equivalent statements—posed or framed differently—can evoke different meanings and responses. For example, risk preferences depend on whether outcomes are framed in terms of gains or losses vis-à-vis a particular reference point, which could be the status quo or current position or, alternatively, an expectation or aspiration regarding a particular position (also known as prospect theory). There is a tendency to take more risks with losses than with gains.


When an unpredicted event occurs, the view of the past and future is adjusted immediately to accommodate the surprise, and there is a tendency to lose the ability to recall what was previously believed to be probable. Hindsight bias may result in the quality of a decision being assessed not by whether the beliefs on which it was based were sound and reasonable, but by whether the outcome was good or bad. Further, a propensity to make sense of the past may lead to an illusion that the future can be predicted or controlled.


Subjective confidence in personal opinions or judgments reflects a personal feeling about the coherence of a self-constructed story and the cognitive ease of processing it. The amount and quality of the evidence have little or no bearing in this context.

Small numbers

There is a tendency to exaggerate the consistency and coherence of what is witnessed, and to believe that small samples are representative of the much larger population from which they are drawn. With little evidence, conclusions are accelerated, and a bigger story is constructed based on that evidence; this is also known as narrative fallacy.

False consensus

There is a tendency to view personal behaviour and responses as typical and appropriate while viewing alternatives as odd and inappropriate.