Table 2

Summary of study characteristics and head-to-head comparisons testing behavioural economics decision biases.

Author (year)	Country	n	Effect	Study length	PA measure	Incentive size	Behavioural economics concepts	Head-to-head comparisons (relevant behavioural economics decision biases)
Acland (2015)44	USA	120	+/−	12,24	Gym	1.00–3.57	PB, SAL, COM, FS	None
Adams (2017)15	USA	96	+/NA	16,0	Fitbit Zip	1.00	PB, SAL	Immediate outperformed delayed incentives (PB) Adaptive goals outperformed static goals (SAL)
Andrade (2014)45	USA	61	±	15,24	Pedometer	0.91–7.00	PB, LA, OO, EE	None
Babcock (2015)46	USA	365	+/NA	2,0	Gym	2.50	PB, SAL, HB	Team incentives produced larger increases in gym visits than individual-level ones (HB)
Carrera (2018)47	USA	690	−/−	6,8	Gym	0.71–1.42	PB, LA, EE	Pre-selected item-based incentives outperformed cash incentives (LA) $60 for nine gym visits no more effective than $30 (SAL)
Condliffe (2017)48	USA	181	+/+	3,2	Gym	0.29	PB, SAL, HB, EE, FS	Team incentives produced larger increases in gym visits than individual-level ones (HB) Information on peers’ gym attendance promoted more frequent gym use (HB)
Finkelstein (2016)	Singapore	800	+/−	24,24	Fitbit Zip	3.19	PB, LA, COM	None
Finkelstein (2017)49	Singapore	161	+/NA	16,0	Pedometer	1.48	PB, LA, OO, COM	Incentives received by participants themselves increased physical activity, but incentives in the form of charitable donations did not (HB)
Harkins (2017)	USA	94	+/−	16,4	Fitbit Alta	2.86	PB, LA	Charitable donations were effective for older participants (HB)
Kullgren (2014)	USA	92	−/−	16,8	Pedometer	3.00	PB, LA, OO, HB	Participants with access to an online peer network walked no more than participants without access (HB)
Leahey (2015)	USA	268	+/NA	12,36	Self-monitor	0.53	PB, SAL	None
Losina (2018)53	USA	202	+/NA	24,0	Fitbit Zip	1.82	PB, LA	None
Patel (2016a)	USA	281	+/−	13,13	Smartphone	1.40	PB, LA, OO, SAL	Loss-framed incentives outperformed lottery and gain-framed incentives (LA, OO, PB)
Patel (2016b)	USA	304	+/−	13,13	Smartphone	1.25	PB, OO, SAL, HB	Incentives for individual- combined with team-based goals outperformed either approach alone (HB)
Patel (2016c)	USA	286	+/−	13,13	Smartphone	1.40	PB, LA, OO, SAL, HB	Incentives for performance compared with group median outperformed comparison to 75th percentile (HB))
Patel (2018)57	USA	209	+/−	13,13	Smartphone	1.25–1.40	PB, OO, SAL	Daily small lottery incentives combined with jackpot lotteries outperformed either condition alone (PB, SAL, OO)
Petry (2013)58	USA	45	+/+	12,24	Pedometer	5.57	PB, LA, OO	None
Pope (2014)	USA	117	+/−	12,20	Gym	4.67	PB, LA, SAL	None
Rohde (2016)	The Netherlands	415	+/+	26,26	Gym	0.09–0.15	PB, SAL, FS	Participants receiving quarterly feedback on the exact number of weeks they met their gym visit goal attended the gym more often than participants who did not (SAL)
Royer (2015)61	USA	1000	+/+	12,36	Gym	4.80	PB, LA, SAL, COM, FS	None
Shin (2017)	South Korea	105	+/NA	12,0	Fitmeter	1.24	PB, LA	None
van der Swaluw (2018)	The Netherlands	163	+/−	13/26, 13/0	Gym	0.89	PB, LA, OO	None
Washington (2016)64	USA	19	NA/NA	3,1	Fitbit One	2.38	PB, LA, SAL, EE	No difference between $25 deposit contract (participant money is ‘on the line’) and no-deposit group (LA)

COM, commitment; EE, endowment effect; Effect, evidence of positive (+) or no (−) effect when incentives in place/after incentives withdrawn; FS, fresh start; Fitmeter, combined accelerometer + app;Gym, number of gym visits; HB, herd behaviour; LA, loss aversion;NA, not applicable due to no post-intervention follow-up period; OO, over-optimism;PA, physical activity; PB, present bias; SAL, salience; Self-monitor, self-monitoring of physical activity behaviour; Smartphone, built-in smartphone accelerometer + app;incentive size, incentive magnitude (per person per day in USD); n, sample size;study length, duration in weeks of intervention.