Table 4. Recruitment Strategies for Nonprobability Samples of Food Insecure Adult Populations in the Food Assistance and Health Care System Sectors.

Example Nonprobability Sampling Frame	Food Assistance Sector		Health Care System
	Private (Charity)	Public	In-Person Recruitment at Waiting Room in Saftey-Net Health Care System (eg, Federally Qualified Health Center)
	In-Person Recruitment in Waiting Room of Staging Or Delivery Locations for Food Distribution Center, Food Pantry, Meal Program	In-Person Recruitment at Government Enrollment Offices (eg, SNAP)
Subpopulations included based on receipt of food assistance
SNAP only		x	x
Charity only	x		x
Both	x	x	x
None			x
Concentration of food insecure	High		Varied
Selection bias	May be biased toward current recipients of food assistance		Underrepresents individuals with limited health care access
Advantages	Potential to link with objectively measured eligibility and food assistance data		Potential to link with objectively measured health status and health care use data
	Potential for improved recruitment because of client trust of partner organization	—	Facilitates identification of food insecure with specific health needs, conditions
Disadvantages	Program catchment area may be small (eg, single metropolitan area) resulting in small geographic scope	Significant administrative, government approval processes; additional restrictions related to data release, sharing, storage, destruction
		—	Costly because of time and effort required to screen patients to identify food insecure

Abbreviation: —, not applicable; SNAP, Supplemental Nutrition Assistance Program.