Glossary

Availability heuristic

The tendency to infer that things that spring to mind easily are more important or more likely to happen.

Behavioral economics

The study of how economic behavior is influenced by factors that are not fully represented in (or that conflict with) neoclassical economic thinking. Behavioral economics uses findings and methods from other disciplines, such as psychology, sociology, and anthropology.

Behavioral insights

An approach that applies evidence of the conscious and nonconscious drivers of human behavior to practical issues and evaluates the results, wherever possible.

Behavioural Insights Team (a.k.a. the Nudge Unit)

A team set up inside the UK government in 2010 with the remit to apply “a more realistic model of human behavior” to policies and services.

Big Five

A widely used taxonomy of personality traits, consisting of five dimensions: openness; conscientiousness; extraversion; agreeableness; and neuroticism.

Bounded rationality

An account of human decision making that proposes that people seek a good enough option (“satisfice”) rather than the optimal option (“maximize”) when making a choice.

Choice architecture

The way in which choices are structured and presented.

COM-B

A model that proposes three factors involved in producing a behavior: capability, opportunity, and motivation.

Complex adaptive systems (CAS)

A dynamic network of many agents, which each act according to individual strategies or routines and have many connections with each other. These agents adapt and self-organize, which means that changes are not linear or straightforward and the behavior of the whole cannot be predicted from its constituent parts.

Compromise effect

The tendency to “go with the middle option” when presented with a set of options.

Confirmation bias

The tendency to seek out, pay attention to, and remember information that confirms a preexisting belief.

Control

The group in a randomized experiment that receives no additional intervention beyond business as usual.

Dark patterns

The use of choice architecture to sway the decision maker toward an outcome that is not in the decision maker’s best interests.

Default

The preselected option that will be taken if no action is taken to override it.

Deliberative forums/citizen juries

Involving the public to help inform policy principles and goals by gathering a representative view of what is desired and acceptable.

Design thinking and human-centered design

Deploying design principles to understand people’s needs and shape processes, policies, products, and services to best meet those needs.

Dual-process theories

Theories that have the common underpinning principle that people have two main ways of thinking: one slow, reasoned, and deliberative, and the other fast, associative, and automatic. Daniel Kahneman characterizes these dual modes of thought as System 1 (fast) and System 2 (slow); Richard Thaler and Cass Sunstein call them the Automatic and Reflective systems.

EAST

A framework for organizing insights from behavioral science. Its core principle is that behavior is more likely to happen if it is easy, attractive, social, and timely.

Ecological rationality

The idea that the rationality of a decision depends on the context in which it takes place, rather than being based on fixed principles. Heuristics may therefore be the best way of making decisions if they fit well with the environment.

External validity

The extent to which the results from one study are generalizable beyond the test population.

False negative

A result from a statistical test that indicates there is no significant difference between the treatment and intervention groups where, in reality, a difference exists.

False positive

A result from a statistical test that indicates there is a significant difference between the treatment and intervention groups where, in reality, none exists.

File drawer effect

The overrepresentation of interesting, novel, or surprising results in academic publications owing to factors that drive researchers to leave their unsurprising or null results in “the file drawer,” rather than putting them forward for publication. Not only does this effect skew perceptions of how likely it is that the average research project will be novel, it also means false positive results are more likely to be published.

Framing

The way in which information is presented, especially in terms of emphasis or order. For example, the following phrases are identical in meaning but vary in frame: nine out of ten people pay their tax on time; one out of ten people do not pay their tax on time.

Heuristic

A simplified rule, or rule of thumb, that can be used to help make a decision with complex informational inputs. Heuristics are practical, producing results that are sufficient but may not be optimal.

Homo economicus

A decision agent that makes decisions solely on the basis of what will maximize value for itself. Homo economicus makes consistent decisions over time, has infinite willpower, and considers all information when making a decision.

Internal validity

The extent to which the methods used in any given experiment can reliably allow the researcher to infer causality.

Intervention

A planned action taken to achieve a particular outcome.

Loss aversion

The tendency to seek to minimize losses, even when the cost of doing so is likely to outweigh the benefits.

Mental accounting

The tendency to place money in certain mental “accounts” linked to particular purposes, and to resist moving money between these accounts. This tendency goes against standard economic theory, which treats money as fungible (i.e., can be used for any purpose).

Mere exposure effect

The phenomenon whereby familiarity (repeated exposure) to an object or person leads to people liking that thing more.

Nudge

An influential 2008 book by Richard Thaler and Cass Sunstein that introduced the concept of “nudging.”

Nudging

Designing choice architecture to make it more likely that a decision maker will choose one option over another. These designs incorporate evidence of the nonconscious drivers of behavior, preserve the right to choose, and seek to advance the interests of the decision maker.

Null hypothesis

The hypothesis that there is no difference in outcomes between the treatment(s) and control groups in an experiment. The null hypothesis is the starting assumption for an experiment, and is rejected if an appropriate statistical test shows a significant difference between the groups.

Omission bias

The tendency to favor not acting over acting when faced with a decision that requires one or the other, partly because we anticipate that not acting will attract less blame. For example, we judge people less harshly if they fail to administer an antidote than if they poison someone.

Overconfidence

The tendency for our subjective confidence in our abilities or judgments to be higher than our actual competencies.

Peak-end effect

The tendency to judge the quality of an experience based on its peak and final moments, minimizing everything else about the experience as a result.

Power calculation

A calculation that uses inputs on the key features of experiment to determine statistical power.

Predictive analytics

Applying analytic techniques to large datasets to predict future behavior using information on what people did in the past.

Randomized controlled trial/experiment/evaluation

An evaluation method that allows the experimenter to infer whether an intervention affected a measured outcome. It does this by creating a control group, randomly allocating participants to this group or a group that received an intervention, and comparing the outcomes for the groups. See chapter 4 for details.

Replication crisis

The failure of many seminal experiments in psychology and scientific fields to find the same results when rerun.

Sludge

A specific form of “dark pattern” in which friction or hassle is introduced into a process to discourage the user from completing it.

Social proof/social norms

The tendency to take our cues from others and behave in line with our perceptions of what is normal or usual.

Statistical power

The probability that an experiment rejects the null hypothesis when a difference in outcomes of the treatment and control groups truly exists. Statistical power ranges from 0 to 1. As power increases, the chance of a false negative result decreases.

Temptation bundling

Pairing a necessary but unappealing activity with one that is appealing in order to motivate completion of the unappealing activity.

Treatment group

The group in a randomized experiment where members receive the intervention.

Utility

A unit of measure that denotes the value of an activity, good, or service for the consumer or beneficiary.

WEIRD population

A study population, overrepresented in academic literature, that is primarily from a culture that is Western, Educated, Industrialized, Rich, and Democratic.