Behavior Change Tactics
These are behavior change tactics in successful products and interventions.
Mindfulness refers to applying focused attention to certain stimuli, typically in the present moment and in a non-judgmental way. Mindfulness has become incorporated into many evidence-based therapeutic approaches, including mindfulness-based cognitive therapy (MBCT) and mindfulness-based stress reduction (MBSR). It has been also been leveraged to influence behaviors outside of traditional mental health areaas, like dietary behaviors and athletic performance.
Peer mentoring refers to having individuals with a certain lived experience guide or train others facing similar ones. It is typically considered in contrast to expert coaching, formal teaching, or management guided by a clinician or doctor. Compare, for example, a weight loss program delivered by a registered dietician vs. group sessions led by program 'graduates' like WW (formerly Weight Watchers). Note that these two approaches are not mutually exclusive.
Implementation intentions are specific details for when and how a behavior should or will be performed. These are often formulated as ""if-then"" rules, such as:- "if I crave something sweet, I'll have fruit instead of candy"- "if I am in the mood for a cigarette, I'll wait 5 minutes—then, if I still want it, I can have one"Other examples include studies where flu vaccination uptake was higher in groups of people nudged to make more specific plans (i.e. picking a specific time and date, along with a mode of transport to a specific clinic). The same general effect was observed with voting behaviors. These are a generally low-cost tool to slightly improve the gap between intention and performance of a behavior.
ideas42 (1),Yoeli (2015),Rogers (2010),Fiorillo et al. (2014),ideas42 (17),ideas42 (19),Barker (2016),Giles (2014),Zhou (2015a),Arora et al. (2014),Nickerson & Rogers (2010),Greer (2019),Liu (2018),Roepke (2015),Direito (2015),Bakker (2018),Duflo et al. (2010),Tudor-Sfetea et al. (2018),Lin et al. (2018),Martin et al (2012),Letzler & Tasoff (2013),Milkman et al. (2011)
Clawback incentives refer to a framing effect applied to rewards where participants are intended to experience losing the reward via noncompliance rather than accruing it for successful performance of the behavior.For example, a hypertension management program may credit its participants $200 at the beginning of the month, and reduce or "claw back" the amount by $3 each time the patient does not take their medication. The alternative would be starting the month at zero or the previous ballance and adding $3 each time the patient takes the medication.
Active choice, sometimes referred to as enhanced active choice or forced choice, refers to removing default options and often increasing the salience of potential decisions through emphasizing the consequences of one or more of the options. Coined by Punam Anand Keller and colleagues in 2011, it was originally intended to address concerns around paternalistic nudging for use in situations where forcing the default option may be considered unethical. In one of the original studies, CVS customers were given the choice to enroll in automatic refills of medications via delivery. The choices they were presented were ""Enroll in refills at home"" vs “I Prefer to Order my Own Refills.”
ideas42 (1),ideas42 (4),Volpp et al. (2011),Chetty et al. (2012),Keller et al. (2011)
Personalization refers to taking specific data from the individual in a behavioral intervention into account in offering a different experience vs. that given to others. An experience may be personalized based on demographic data, psychographic data, behavioral performance, or other factors.
Dobson et al. (2018),von Lengerke (2017),de Vries (2018),Direito (2015),Kunreuther et al. (2012),Garner (2005),Behavioural Insights Team (2013),Kling et al (2012)
Feedback on outcomes/consequences
Feedback on outcomes or consequences entails information about the short- or long-term likelihood of various events given current behaviors or alternatives. Examples of this kind of feedback may include information about relative cancer risk based on current lifestyle factors or calculated net worth in 20 years based on the person's current savings rate and investment returns.
Piette et al. (2012),Piette et al. (2016),Rubinstein et al. (2016)
Micro-incentives refers to small rewards, typically frequent and cash-based, given out on a per-behavior basis. A prominent example is Wellth, a program for people with chronic illness delivered via app. Some participants are given around $2 each time they take a medication or measure their blood pressure and submit a photo. In related studies, e.g. Petry et al. (2015), participants' compliance with these behaviors was significantly higher than those who did not receive the incentives, and the behaviors persisted several months after incentives were removed. Micro-incentives can be layered with other reward approaches such as lotteries and non-financial incentives.
Sen (2014),ideas42 (2),Loewenstein (2015),Kullgren,Neckermann,Bronchetti (2015),Wilson,ideas42 (10),McCoy (2017),Fafchamps,Appelt,Shyamal,Banerjee et al. (2010),Homonoff (2012),Volpp et al. (2012),Thorton (2008),Karlan et al. (2010),Haisley et al. (2012)
Non-financial incentives refer to any reward given to change the probability of a behavior which isn't money.These can include social status (e.g. an 'employee of the month' award circulated via email or hung upon a wall), admittance to an exclusive event, or simply praise.
Behavioural Insights Team (2013),Appelt
Psychadelic-assisted therapy refers to forms of counseling or psychotherapy with the addition of a psychadelic or empathogenic substance, such as MDMA, psilocybin, or ketamine.Some evidence suggests this approach may outperform counseling alone or other forms of standard care for certain populations in certain contexts.
Reduce Cognitive Load
Reducing cognitive load refers to simply making it easier to do something, or lowering the amount of thinking required to do it. People are cognitive misers, and seemingly trivial amounts of thinking may provide disproportionate amounts of deterrence to performing a behavior (even when someone intends to do it).Checklists and mnemonic devices are both forms of reducing cognitive load. Similarly, automating parts of a person's behavior (e.g. rules-based savings plans) limit the cognitive resources required to reach a goal. Other examples include showing health risk data visually via infographic rather than in verbal descriptions, or limiting the set of options a person has in choosing a retirement plan (e.g. by parameters like "aggressive," "medium," or "conservative" rather than describing the asset class mixture).
Kling et al (2012),Fischer (2008),Choi et al (2006),Kunreuther et al. (2012),Arriaga et al. (2013),Haynes et al (2009),Bettinger et al. (2009),Bhargava & Manoli (2013),Hastings & Tejeda-Ashton (2008),Haynes et al (2011),Fiorillo et al. (2014),Schoar et al. (2014)
Behavioral economics is the exploration of how people make consequential decisions where psychological and sociological factors may influence the outcome or process. It is often considered the fusion of economics and psychology (which itself was an interdisciplinary field entailing medicine and philosophy). The exploration of psychological factors in economic decision-making, including deviation from rationality, traces well back to classical and neoclassical economics (i.e. Gabriel Tarde, Wilfredo Pareto, and John Maynard Keynes) and prior to psychology becoming a formal discipline. Behavioral economics is often associated with behavior change tactics like smart defaults, reducing friction or barriers, increasing salience, incentives, active choice, and commitment devices.
Skill coaching simply refers to guiding a person to acquire a new behavior or set of behaviors, often by scaling the information and challenge level with their growing ability. Examples of this might include teaching someone to cook healthy foods or training them in various self-regulation techniques from cognitive behavioral therapy. This coaching may be done by an individual or automated through digital experiences.
Johnson et al. (2017),Piette et al. (2011),Shetty et al. (2011),Kiernan (2013),Wang (2011)
Acceptance and Commitment Therapy (ACT)
ACT is a therapeutic approach originalled developed by Steven Hayes. It borrows from previous concepts like cognitive behavioral therapy and Morita therapy. The principles of ACT are fairly systematic and lend themselves well to program design, finding empirical support in adaptations like 2morrow's smoking cessation and pain management interventions.
Bricker (2014),Bricker (2016),Hefner (2019),Sridharan (2019)
Random screening refers to unannounced checks of whether someone has been compliant with a given behavior. These are frequently used via biomarkers, e.g. testing if someone has been taking recreational drugs by delivering a urine test.
Behavioral Activation (BA)
Behavioral activation is a therapeutic approach that typically pairs activity scheduling with either monitoring tools or goal-setting. For example, someone might aim to balance activities they "should" do but underperform, like self-care behaviors, with activities they enjoy. Users of this technique may also track which activities cause certain cognitions or affective states, like those associated with depression.
Lotteries are any form of assigning an award where there is an element of randomness or chance.One example is a prize-linked savings account (PLSAs). One of the earliest of these was the Million a Month Account (MAMA) in South Africa, where First National Bank offered account-holders with qualifying deposits a chance to win up to one million rand each month (along with other smaller prizes given out at random). Lotteries may be used with non-financial rewards as well, e.g. offering tickets to a sold-out play or sporting event for employees reaching certain performance benchmarks.
Volpp et al. (2008b),Volpp et al. (2008a),Sen (2014),Habyarimana & Jack (2011),Asch (2015),ideas42 (10),Volpp et al. (2016),Volpp et al. (2009),Volpp et al. (2012),Haisley et al. (2012)
Priming in behavioral interventions typically refers to attempting to influence someone's behavior by brief exposures to certain words, images, sounds, or other stimuli beforehand. In cognitive psychology, priming effects have been found to improve the speed of processing related information (for example, naming an image when previously shown a related image vs. an unrelated one). Similarly, reminding someone of how to do a behavior or its consequences can be a way to influence them a brief while later by making this information more available in memory. Notably, several well-known priming studies have failed to replicate and it may be a less potent technique than previously believed. At the very least, priming should be considered to most likely impart effects only in the short term.
Shu et al. (2012),Zwane et al. (2011),Stango & Zinman (2011)
Group incentives refer to structure where an individual's likelihood or size of reward is influenced by others. The intention is to leverage positive peer pressure by causing compliant participants to influence less compliant participants to improve their behavior.For example, sales teams may be offered a bonus based on an office's collective revenue generation or provided all individuals meet a baseline level of performance. Similarly a multi-site franchise may offer an incentive for whichever location improves their performance the most over the prior month.
Kullgren,Asch (2015),Haisley et al. (2012)
Goal setting simply refers to a person choosing a specific result to aim at achieving. This might include an outcome (e.g. a goal weight) or a behavior (e.g. exercise 90 minutes 3 times a week).
Choi et al. (2017),Fiorillo et al. (2014),Ashraf et al. (2010),Cheema & Soman (2011),Block et al. (2015),Coups et al. (2015),Presseau (2015),Greer (2019),Roepke (2015),Direito (2015),Bakker (2018),Laing (2014),Arnardóttir (2019),El-Hilly et al. (2016),Fukuoka (2015),Karlan et al. (2010)
Commitment devices are tools that attempt to bridge the gap between a person's initial motivation to perfrom the behavior and the typical pattern of noncompliance as time goes on.One prominent example is the "Ulysses Pact," where Filipino banking customers were offered the option to enroll in an account where their ability to make withdrawals would be limited. In a study by Ashraf and Karlan (2005), participants with the commitment account saved 81% more than those with typical accounts. There are many other examples of commitment devices. Temptation bundling is a form of commitment device where people only engage in an enjoyable activity when it's simultaneous with an activity they intend to do more (for example, only listening to a certain podcast or audiobook while walking on a treadmill). Pre-paying for a service is a basic form of commitment device, and one used by Dan Ariely when he intended to increase his fruit and vegetable consumption. He paid for a year of biweekly deliveries from a local CSA program up-front.
Behavioural Insights Team (2013),Milkman (2013),Meeker (2014),Long (2012),Ghani (2015),Ashraf et al. (2010),Cheema & Soman (2011),ideas42 (14),Karlan & Linden (2014),Volpp et al. (2008b),Giné et al. (2008),Brune et al. (2011),Karlan et al. (2010)
Reminders, Cues, or Prompts
Reminders, cues, and prompts are simply methods to cause someone to perform a behavior by calling their attention to it with a timely message. People have limited attention and memory, so these types of influences can be very effective when done skillfully. The cue need not consist of written or spoken language; for example, it could be a certain melody, symbol, or pattern of lights on a connected home device. It might also be a bracelet or pattern of vibrations from a wearable device. Provided the cue or prompt is associated with the behavior, almost any sensory stimuli that is reliably perceived and interpreted may be used. That said, verbal reminders can be effective since they may be personalized with additional semantic information related to the person's context or leverage other effects (e.g. identity priming or framing effects).
Levy et al. (2015),Thomsen (2016),Casarotto,Schoar (2012),Yoeli (2015),Raifman,Scartascini,Valenzuela et al. (2011),Kuna (2015),Haynes et al. (2013),Habyarimana & Jack (2011),ideas42 (3),Humpage (2015),ideas42 (5),ideas42 (7),ideas42 (8),ideas42 (9),ideas42 (10),Fiorillo et al. (2014),ideas42 (11),ideas42 (13),Cheema & Soman (2011),ideas42 (15),ideas42 (16),Zaliauskas et al. (2015),ideas42 (17),ideas42 (18),Zinman (2018a),Fortmann et al. (2017),Mohammed et al. (2016),Kliner et al. (2013),Pop-Eleches et al. (2011),Rodrigues et al. (2012),Finkelstein (2015),Volpp et al. (2016),Barker (2016),Debono (2017),Mira (2014),Reyzelman (2018),Wang (2015),Turner-McGrievy (2011b),Allen (2013),Bond (2014),Gregory (2016),Greer (2019),Liu (2018),Roepke (2015),Charpentier (2011),Direito (2015),Bakker (2018),Laing (2014),Sussman (2016),Wharton (2014),Arnardóttir (2019),Kaplan (2017),Schwartz et al. (2012),Duflo et al. (2010),Kiernan (2013),Cadena & Schoar (2011),Letzler & Tasoff (2013),Kunreuther et al. (2012),Fellner et al. (2011),Castleman (2013),Turner-McGrievy (2013),Karlan et al. (2010),Kalbekken (2011),Milkman et al. (2011)
Self-monitoring or tracking simply refers to a person measuring their behavior over time.Often, merely tracking a behavior can influence the likelihood or frequency with which a person performs the behavior or related ones. For example, many pedometer studies increase walking activity merely by improving awareness, and many interventions that merely consist of rewarding someone for weighing themselves result in weight loss. That said, people often find tracking behaviors tedious and lose interest after a short period, so behavior designers should seek to reduce the burden of self-monitoring by collecting information automatically or doing so in a low-cost way.
Boyd (2015),Piette et al. (2012),Piette et al. (2016),Rubinstein et al. (2016),Tudor-Sfetea et al. (2018),Turner-McGrievy (2013)
Rules of Thumb
Rules of thumb refer to simplifation heuristics used in dealing with uncertainty, situations where tracking behaviors can be onerous, or areas where one-size-fits-all approaches may not be successful. They can be a useful tool to reduce the cognitive load of complying with a new behavior.For example, a person may find it easier to "eat out at restaurants only 4 times per month" rather than "limit monthly restaurant spending to $200." Similarly, avoiding eating certain types of foods, e.g. fried foods or high-calorie drinks, may be easier to recall and comply with than hitting a daily calorie goal.
Schoar et al. (2014),Datta et al. (2015b),Chang,Datta et al. (2015a),Rogers (2014)
Social norms are shared expectations on how people within a certain group will or should behave. They are often considered as unwritten rules that govern behavior and tend to be very influential.Influencing behavior using social norms can take a variety of forms. For example, some studies aim to correct misunderstandings around descriptive norms (what people in a group actually do). One trial involving the UK Behavioural Insights Team increased tax compliance by emphasizing that the vast majority of people pay their taxes on time, which influenced non-compliers to become more like the majority. People generally do not like to deviate from the norm, which may explain the success of this tactic.Other approaches involve attempting to change social norms or create new social norms, which is substantially harder. One prominent example was the promotion of the ""designated driver"" (DDs) in the US during a period of high automobile fatalities. Public health officials influenced Hollywood producers to include the designated driver in film and television scenes, which caused viewers to: 1) likely believe the use of DDs was much more common than it actually was, and 2) likely consider using a DD was what they ""should"" do (i.e. the injunctive norm). Following the public health campaign, awareness and compliance with the DD protocol rose substantially, and auto fatalities dropped precipitously.
Behavioural Insights Team (2011),Walsh (2012),ideas42 (2),Doctor (2018),Rodriguez,ideas42 (4),ideas42 (6),ideas42 (8),ideas42 (9),ideas42 (11),Hallsworth (2016),Campbell (2018),Cimini (2009),Emmons (2003),"Gerber & Rogers, (2009)",OECD (2017c),Schultz et al. (2007a),Thombs (2007),Debono (2017),Elouafkaoui (2016),Sussman (2016),"Slaunwhite, J. (2008)",Jessoe & Rapson (2012),Blumenthal et al. (2001),Kunreuther et al. (2012),Wood et al. (2012),Bond et al. (2012),Cialdini (2003),Fellner et al. (2011),Schultz et al (2007b),Behavioural Insights Team (2013),Habyarimana & Jack (2011),Kalbekken (2011)
Moral suasion refers to appealing to a person's sense of morality or ethics in order to influence their behavior.In one example, the UK Behavioural Insights Team compared reminders to pay taxes. One letter featured a moralistic message ("Paying your taxes is the right thing to do") whereas another empahsized social norms ("9 out of 10 people pay their taxes on time"). The moral suasion message performed worse in this case.
Social benchmarking refers to comparing a person's behavior, trends, or status to others. Often, merely providing data on others can change behavior by leveraging social norms. For example, letters comparing homeowners' use of electricity with peers were found to significantly reduce the amount of energy used by high-consumption households compared to non-comparison messages.
Meeker (2016),Datta et al. (2015a),Rogers (2014),Appelt,Coups et al. (2015),Heldt (2005),Ayres et al. (2012),Croson & Shang (2008)
Education or Information
Education refers to empowering a person with more knowledge or training than they had previously. While providing information alone is often a suboptimal way to drive meaningful behavior change or long-term interventions, the right message at the right time can be a powerful part of a behavior change strategy.
Dobson et al. (2018),Fortmann et al. (2017),Krahé (2005),Hill (2008),Kellar (2005),Kwasnicka (2017),Lally (2005),Larkin (2015),Lemmens (2010),Llewellyn (2012),Mosleh (2009),Newbury-Birch (),Powell (2015),Yates (2015),Zhou (2015c),Drossaert (1996),Hallsworth (2016),Hill (1982),French (2017),Barker (2016),Gardner (2017a),Phillips (2018),Volpp et al. (2009),Alcalay (1993),Morrison (2016),Miller (2017),Tombor (2016),Treweek (2014),Giles (2014),Lavelle (2017),Creedon (2006),von Lengerke (2017),Boyd (2015),Hammond (1999),Springvloet (2015),Perez (2013),Barnett (2014),Leppla (2016),Witteman (2017),Bates (2017),Aventin (2015),Beard (2015),Wardle (2003),Clarkesmith (2016),Evans (2016),Kiene (2006),Norris (2016),Carmack (2006),Paradis (1995),Zhou (2015a),de Vries (2018),Wang (2015),Van Het Reve (2014),Turner-McGrievy (2011b),Allen (2013),Bond (2014),Olano-Soler (2017),Gregory (2016),Greer (2019),Wall (2018c),Liu (2018),Roepke (2015),Charpentier (2011),Direito (2015),Bakker (2018),Laing (2014),Sussman (2016),Arnardóttir (2019),Kaplan (2017),Swartz et al (2011),Kiernan (2013),Chamberland (2017),Hastings & Tejeda-Ashton (2008),Gold (1994),Mathios (2000),"Fraser, Hite, & Sauer (1988)",Kunreuther et al. (2012),Small et al. (2007),Hershfield et al. (2011),Bhargava & Manoli (2013),Newell & Siikamäki (2013),Turner-McGrievy (2013),Fukuoka (2015),Bertrand & Morse (2011),Wang (2011),Chen et al. (2017),Merchant et al. (2018),Merchant et al. (2015)
Behavior substitution refers to attempting to eliminate a problematic behavior by replacing it with another one. Often, the substituted behaviors are intended to have similar sensory qualities (e.g. drink flavored sparkling water instead of soda). The goal is typically to disassociate the original behavior from its cue, enabling the more positive behavior to be triggered automatically.
Reduce Friction or Barriers
Reducing friction or barriers to performing a behavior is simply making it easier or removing things that may be preventing someone from doing something. This is a foundational technique in changing behavior, and part of the UK Behavioural Insights Team's 4-point approach ("Make it easy"). That said, knowing where the friction and barriers exist may not always be straightforward, and different groups of people may experience different barriers in different contexts. Note: It is possible to remove too much friction. In a well-popularized study, a travel booking site found that delays in loading the best deals or travel options actually increased conversions. Similarly, longer input forms in digital interactions sometimes outperform, as people may consider the results more personalized or experience greater cognitive dissonance after having invested so much time in exploring the service.
Casarotto,Humpage (2015),ideas42 (12),ideas42 (19),Choi et al (2006),Hastings & Tejeda-Ashton (2008),Kling et al (2012),Letzler & Tasoff (2013),Kunreuther et al. (2012),Bettinger et al. (2009)
Automation refers to having another person, group, or technology system perform part or all of the intended behavior. A prominent example is Thaler & Bernartzi's Save More Tomorrow intervention, which invested a portion of employees' earnings into retirement funds automatically and even increased the contribution level to scale with pay raises. Other examples include automatically scheduling medical appointments so the patient needn't do it themselves and mailing healthy recipe ingredients to the person's home to reduce the burden of shopping.
Ghani (2016),Sussman (2016),Chetty et al. (2012),Behavioural Insights Team (2013),Behavioural Insights Team (2013)
Increasing salience simply refers to making a choice, behavior, or set of consequences more obvious. The forms this can take are myriad—increasing the size of a button on an app, showing a picture of lungs with cancer rather than verbally describing it, or adding warning labels onto unhealthy foods are all examples of attempting to influence behavior through manipulating salience. In financial interventions, this might entail calculating the total lifetime cost of failing to utilize an employer 401k match or showing a detailed photo of a potential drream home to influence savings rates.
Bhargava & Manoli (2013),Stango & Zinman (2011),Newell & Siikamäki (2013),Kallbekken et al. (2012),Chetty et al. (2009),Chojnacki (2017)
Environmental restructuring refers to modifying the physical environment around someone in order to influence their behavior.On the less intensive end, this could be as simple as having someone leave a pill bottle in a more obvious location or switch to using a pillbox with compartments for each day. More complex examples include carpooling potential voters to election sites to improve turnout, redesigning a workplace cafeteria layout to bias toward healthier foods, or setting up booths for influenze vaccination in offices or shopping malls.
Presseau (2015),Debono (2017),Gardner (2016),Gardner (2017a),Guzmán (2016),Phillips (2018),Tombor (2016),Munir (2018),Direito (2015),Sussman (2016),North (1997),Dayan & Hillel (2011),Hershfield et al. (2011),Cialdini (2003),Kallbekken et al. (2012),Hanks et al. (2012)
Checklists are an age-old tactic for remembering to do certain tasks. Checklists are sometimes used to measure behaviors that should take place with a certain frequency, e.g. every day or X times per week, and other times, to ensure certain steps are followed every time a person does a complex behavior.For behavior designers, the challenges of checklists often entail choosing the right behaviors, breaking them down to the correct level of granularity for a given population, and serving them up in the proper context or sometimes with personalization. They are likely underutilized and consistently improve the performance of even experts, like pilots and surgeons.
ideas42 (2),ideas42 (7),ideas42 (11),ideas42 (12),ideas42 (16),ideas42 (18),Arriaga et al. (2013),Haynes et al (2009),Haynes et al (2011)
Self-efficacy refers to one's belief that they can succeed with a given behavior in a particular context. It is a concept largely popularized by Albert Bandura as a key component of social cognitive theory. Increasing self-efficacy can be an effective tactic for mediating a person's intention to perform a behavior and their actual performance. Typically, this entails positive feedback around elements of the behavior they've performed correctly, information about others who have succeeded (i.e. viarious experience), or other information aimed to improve a person's belief they can be successful with a behavior.A notable perspective on self-efficacy comes from the Health Action Process Approach, which distinguishes self-efficacy related to initial performance vs. maintenance and other instances of performing the behavior.
Feedback on behavior
Feedback on behavior refers to information about how a person's current performance tracks against goals or standards.Performative information might include data on how a person's current diet tracks with nutrition recommendations or how their home power consumption compares with nearby households.
Kamal et al. (2015),Piette et al. (2011)
Defaults refer to what happens if a person makes no choice or goes with a pre-selected choice. The influence of defaults is a foundational component of behavioral economics. Perhaps the most famous example of defaults is the difference between opt-in and opt-out organ donation programs. While not universal, several studies have found that the rate of organ donation consent in a population seems to be influenced by the default (i.e., what happens if a person does not check a box or change the pre-selected preference on a form). Smart defaults do not only refer to one-off events, however. In the well-known Save More Tomorrow program, participants were not only included in a savings program by default, but the amount they saved was also changed over time automatically (again by default). Similarly, other behavior change programs have default settings that include at-home medication or food delivery, rules-based reminders on different platforms, etc.
Johnson & Goldstein (2003),Benartzi & Thaler (2004),Mathias,Ghani (2016),Behavioral Insights Team (2012),Brown et al. (2013),Loewenstein et al (2013),Pichert & Katsikopoulos (2008),Madrian & Shea (2001),Goda (2010),Kalbekken (2011)
Feedback entails providing qualitative or quantitative information about a behavior's performance or consequences. Performative information might include data on how a person's current diet tracks with nutrition recommendations or how their home power consumption compares with nearby households.Feedback on outcomes may include information about relative cancer risk based on current lifestyle factors or calculated net worth in 20 years based on the person's current savings rate and investment returns.
Block et al. (2015),Meeker (2016),Kuna (2015),Doctor (2018),Datta et al. (2015b),Asch (2015),Rogers (2014),Appelt,Coups et al. (2015),Presseau (2015),Elouafkaoui (2016),Van Het Reve (2014),Turner-McGrievy (2011b),Allen (2013),Bond (2014),Wall (2018c),Direito (2015),Laing (2014),Wharton (2014),Arnardóttir (2019),Kaplan (2017),El-Hilly et al. (2016),Fischer (2008),Kunreuther et al. (2012),Hershfield et al. (2011),Ayres et al. (2012),Chen et al. (2017),Merchant et al. (2018),Merchant et al. (2015)
Financial incentives are monetary rewards given for performing a certain behavior. These come in many different varieties; for example, they may be guaranteed vs lottery-based, or group-oriented vs individually-assigned.
Volpp et al. (2008b),Volpp et al. (2008a),Behavioural Insights Team (2013),Volpp et al. (2016),Volpp et al. (2009),Curry (1991),Edwards (2016),Gastfriend (2019),Paradis (1995),Kurti (2019),Cadena & Schoar (2011),Fraser & Hite (2006),Gächter et al. (2009),Belot et al (2013),Jue et al. (2012),Thorton (2008),Fryer et al. (2012),Hossain & List (2009),Haisley et al. (2012)
Goal setting (behavior)
Goal setting for behaviors refers to a specific measurable result involving a behavior, as opposed to an outcome. It cannot, for example, specify an outcome like a target blood pressure, as there are numerous behaviors that could be undertaken to achieve this result. Instead, the goal in this case would likely be taking medication 2 times a day, walking 10,000 steps per day, measuring blood pressure every morning at 9am, etc.
Rubinstein et al. (2016)
Tracking cognitions or emotions
Tracking cognitions or emotions (often both) refers to a person recording when they have certain thoughts or feelings. The person might note every time they experience a given thought or specific feeling whenever it comes up, or alternatively simply keep a diary of any notable thoughts or feeling at pre-determined times. Often, this also includes noting what triggered or occured before or alongside these thoughts and emotions. Many therapuetic approaches utilize this tool, even if only for brief periods, to help a system, clinician, or patient better understand the patterns around their thoughts and feelings. Often, this data is integrated into additional behavior change approaches, like behavioral activation or implementation intentions.
Cognitive Behavioral Therapy (CBT)
Cognitive behavioral therapy (CBT) is a therapuetic approach to improving mental and behavioral health. The core philosophy is that behavior can be modified by noticing and correcting patterns in thought that influence the behavior. Modern CBT is typically associated with Albert Ellis and Alan Beck.The structured and rules-based nature of CBT have made it a popular candidate for digital interventions and application by lightly-trained or even untrained practitioners.
Lin et al. (2018),Tudor-Sfetea et al. (2018),Greer (2019),Fitzpatrick (2017),Watts (2013),Birney (2016),Bayley (2015),Schneider (2004),Cook (2016),Dobson (2018)
A framing effect refers to changes in people's choices within a given set of options based on how the options are presented. This are typically associated with behavioral economics, as it violates utility theory's premise that people will choose according to a rational assessment of the outcome.The most common example of this is posing a question as a loss or a gain. In several instances, people have been found to choose differently based on whether a proposition is losing lives vs saving them, an X% of infection vs. a Y% chance of immunity, etc despite the options being mathetmatically identical between the two framings.
Scartascini,Doctor (2018),Pouliquen et al. (2015),ideas42 (5),Wilson,Choi et al. (2017),Lee,Bertrand,Zaliauskas et al. (2015),Hershfield (2018),Chetty et al. (2012),Keller et al. (2011),Gächter et al. (2009),Kahneman & Tversky (1981),Slovic et al (2002),Small et al. (2007),Lamberton (2013),Slovic et al. (1997),McNeil et al. (1982),Kogut & Ritov (2005),Redelmeier et al (1993),Bertrand et al. (2010),O’Connor (1988),Breman (2011),Behavioural Insights Team (2013),Croson & Shang (2008),Hasseldine et al. (2007),Beatty et al. (2011),Viscusi et al. (1987),Fryer et al. (2012),Kalbekken (2011),Hossain & List (2009),Bertrand & Morse (2011)
Rewards & Incentives
Changing the consequence of a behavior is one way to influence it. Rewards and incentives constitute positive reinforcement—the addition of a desired stimulus following the successful performance of a behavior. These may be monetary or non-monetary.
Motivational interviewing (MI) is a therapeutic approach that aims to influence behavior by eliciting goals, motivation, insights, and specific behavioral plans through structured dialog. It's largely associated with William Miller and Stephen Rollnick, and bears some relation to the Socratic method (as does the original cognitive therapy approach). While originally developed as part of a treatment for substance abuse, the method has been generalized and found empirical support in assisting behavior change in diet, exercise, and other areas.
Bayley (2015),Golin (2010),Gourlan (2013),Krukowski (2016),Moreau (2015),Morton (2015a),Murphy (2010),Paranjothy (2017),Perry (2007),Phillips (2018),van Keulen (2008),Berdal (2018),Ahluwalia (2007),Kiene (2006),Marrazzo (2011),Pakpour (2015),Cimini (2009)
AI or Chatbot
Using a chatbot or simulated conversational interaction.
Fitzpatrick (2017),Greer (2019),Liu (2018),Bakker (2018)
Hot & Cold States
Hot and cold states refer to periods of high or low arousal. This can be caused by a perceived threat, a very large reward, or a highly desirable "super" stimulus (e.g. a vivid video of a pizza, a scantily-clad human, etc). People tend to behave differently in hot and cold states, and also underestimate the difference between their selves in each.Attempts to influence behavior using this principle tend to aim to shift someone from a hot to a cold state, or vice versa. There is some evidence that people use more automatic thinking (sometimes referred to as "System 1") during hot states, and more deliberate thinking ("System 2") during cold ones.
Habit formation refers to approaches where a person is trained to associate a behavior or chain of behaviors with a given cue, often by introducing rewards or punishment. While traditional habit formation approaches derived largely from behaviorism and operant conditioning, many modern approaches emphasize the cognitive processes that mediate the behavioral response to cues or explore methods of modifying environments (either to reduce the frequency or salience of cues to undesired behaviors, or do the reverse for desired ones). Many habit formation programs, e.g. BJ Fogg's Tiny Habits approach, emphasize reducing the scope of the desired behavior and scaling it over time with the person's growing ability.
Identity priming refers to attempting to influence someone's behavior by emphasizing their being part of to a certain group or being a certain type of person. These often leverage social norms—particularly injunctive norms—and introjected regulation.For example, voter turnout campaigns often emphasize the person's membership to the community or previous voting history in reminder letters.
Behavioural Insights Team (2013),Yoeli (2015),ideas42 (9),Lee,Curry (1991),Debono (2017),Tombor (2016),Roepke (2015)
Self-Monitoring or Tracking
Self-monitoring or tracking simply refers to a person measuring their behavior, experiences, cognition, or other data points over time.Often, merely tracking a behavior can influence the likelihood or frequency with which a person performs the behavior or related ones. For example, many pedometer studies increase walking activity merely by improving awareness, and many interventions that merely consist of rewarding someone for weighing themselves result in weight loss. Similarly, when cognitive behavioral therapy patients track which cues or environments are associated with undesired behaviors or thoughts, they may begin to avoid them.Unfortunately, people often find tracking behaviors tedious and lose interest after a short period, so behavior designers should seek to reduce the burden of self-monitoring by collecting information automatically or doing so in a low-effort way.
Holmen et al (2014),Orsama et al (2013),Quinn et al (2011),Charpentier et al. (2011),Levy et al. (2015),Bolinder et al. (2016),Haak et al. (2017),Rossi et al. (2010),Rossi et al. (2013),Gajecki (2014),Gardner (2016),Reyzelman (2018),Wang (2015),Turner-McGrievy (2011b),Allen (2013),Bond (2014),Wall (2018c),Roepke (2015),Charpentier (2011),Direito (2015),Bakker (2018),Laing (2014),Sussman (2016),Wharton (2014),Arnardóttir (2019),Kaplan (2017),El-Hilly et al. (2016),Lin et al. (2018),Turner-McGrievy (2013),Fukuoka (2015),Chen et al. (2017),Merchant et al. (2018),Merchant et al. (2015)
Public commitments are stating one's intention to perform a behavior or achieve a certain outcome so that others become aware of it. The commitment need not be truly public (i.e. possible for any person to learn), and can be limited to family, coworkers, a small group of friends, etc.Public commitments are often layered with rewards (e.g., "I will walk 10 miles for every $100 donated by my Facebook friends") or punishments (e.g., "I will do [something embarassing] if I fail to follow-through"). Public commitments are a form of commitment device, where loss of social status or other social consequences may be implied.Some studies have found mixed results with public commitments, so this tool should be handled carefully. It is believed that in some contexts, some of of the benefit someone receives from performing a behavior or reaching a goal is actually imparted by the social status they receive in stating their intention to do so. Thus, the actual performance become less important (especially when compared with the cost, which is typically borne over many weeks or even years).
Meeker (2014),Long (2012),Hicken (2013)
Social support refers to the perception or reality that other people will provide assistance in a given context. It is a key component of several behavior models and plays an important role in mediating behavior change.
Block et al. (2015),Islam et al. (2015),Kamal et al. (2015),Kliner et al. (2013),Lester et al. (2010),Mbuagbaw et al. (2012),Mohammed et al. (2016),Rodrigues et al. (2012),Shetty et al. (2011),Curry (1991),Debono (2017),Guzmán (2016),Presseau (2015),Tombor (2016),Munir (2018),Wang (2015),Turner-McGrievy (2011b),Allen (2013),Bond (2014),Direito (2015),Bakker (2018)
Gamification refers to leveraging mechanics and other experiential elements typically associated with games in non-game contexts.These can be fairly subtle (e.g. a progress bar for filing out a health risk questionnaire), moderate (e.g. achievements given for reaching personal finance goals, contests for steps walked as a team in a workplace wellness competition), or extreme (e.g. an augmented reality experience to treat chronic pain). At the extreme end, the distinction between a gamified experience and an actual game may be considered almost academic.
Block et al. (2015),Arora et al. (2014),Oddsson (2017),Edwards (2016),Curtis (2015),Cowdery (2013),Lee (2010),Pham (2016),Wang (2015),Roepke (2015),Direito (2015),Arnardóttir (2019),El-Hilly et al. (2016),Tudor-Sfetea et al. (2018),Lin et al. (2018),Fukuoka (2015)
Coaching or Counselling
Coaching or counselling here refers to having a trained person provide guidance to someone attempting a behavior. Many mental health and lifestyle programs utilize coaching in various forms, including phone calls, video chat, text messaging, or in-person sessions. Some programs have replaced some or all of these traditionally human-delivered touchpoints with AI or rules-based interactions.
Block et al. (2015),Holmen et al (2014),Orsama et al (2013),Quinn et al (2011),Charpentier et al. (2011),Rossi et al. (2010),Rossi et al. (2013),Marler (2019),Munir (2018),Golin (2010),Liu (2018),Herghelegiu (2017),Jongstra (2016),Leonhardt (2008),Thomsen (2016),Darlow (2017),Gourlan (2013),Joost (2014),Newbury-Birch (),Kiene (2006),Skouteris (2016),Chen et al. (2017),Merchant et al. (2018),Merchant et al. (2015)
©2020 The Behavior Institute LLC. All rights reserved.