Using Facebook Advertising to Recruit Representative Samples: Feasibility Assessment of a Cross-Sectional Survey.

Area of Science:

• Public Health
• Health Services Research
• Digital Health

Background:

• Facebook advertising shows potential for economical participant recruitment in health research.
• Few Canadian studies have explored this method, especially for older adults in Newfoundland and Labrador.

Purpose of the Study:

• To evaluate Facebook advertising for recruiting a representative sample of adults aged 35-74 in Newfoundland and Labrador for a health survey.
• To assess the cost-effectiveness of Facebook advertising compared to traditional methods.

Main Methods:

• Web-based survey on cancer awareness and prevention using Facebook ads (April-May 2018).
• Targeted ads to improve demographic representation.
• Comparison of sample characteristics with population data using Cramer V and chi-square tests.
• Cost analysis versus random digit dialing (RDD).

Main Results:

• Facebook recruitment yielded 1048 participants at Can $2.18 per recruit, significantly cheaper than RDD's quoted Can $35.52 per recruit.
• The sample was representative or practically representative of geography, age, income, healthcare provider access, diabetes, and smoking status.
• The sample was over-representative of women and higher education levels, and not representative of arthritis prevalence or perceived health.

Conclusions:

• Facebook advertising is a feasible, rapid, and economical tool for recruiting middle-aged and older adults for health surveys.
• The method achieved partial representativeness across key health and demographic characteristics.
• Caution is advised for research requiring strictly random samples due to Facebook's targeting algorithm.