Using discrete choice experiments to inform the design of complex interventions

Table 3 Discrete choice experiment attributes and levels

Attribute	Levels
Time of service	Normal working hours and days
Time of service	Extended hours and weekend services
Service separation	Standard service with all clients together
	Separate waiting areas for younger and older men
	Separate services for younger and older men
HIV testing	Opt-out
	Opt-in
	Not available
Gender of service provider	Male and female
	All male
	All female
Female partner counselling	Partner counselling offered in the community
Female partner counselling	No partner counselling offered
Incentives	None
	Transport voucher for 500 Tanzanian shillings (TzS)
	Transport voucher for TzS 1500^a
	Transport voucher for TzS 4500
	Lottery low (1/10 chance of winning TzS 5000)
	Lottery medium (1/10 chance of winning TzS 15,000)
	Lottery high (1/10 chance of winning TzS 45,000)^b

^aTzS 500, TzS 1500 and TzS 4500 are roughly equivalent to US$0.30, US$0.90 and US$2.70 in 2015
^bThe expected values of the low and high incentives are equal across the transport voucher and lottery

ISSN: 1745-6215