JPSM MPSDS Seminar Series: Social Media and Hard-to-Reach Audiences
The Use of Advanced Social Media Targeting Methodology During Recruitment of Hard-to-Reach Audiences
Ipek Bilgen and Amelia Burke-Garcia, NORC at the University of Chicago
Wednesday, March 9
12:00pm to 1:00pm
Join via Zoom:
Meeting ID: 97702415176
One of the major benefits of social media ad-based survey recruitment is the use of various types of data to target ads to users of these platforms. To target users of social media, researchers can use the basic demographic and geographic that social media platforms currently provide, or they can use enhanced data that can be embedded within the social media platforms supplied by third party providers based on external data sources, e.g., historical purchase data. We will examine whether and how much this enhanced data can impact ad based social media recruitment capabilities to reach niche and hard-to-reach audiences.
To investigate the targeting efficiency, quality, and cost differences among these two approaches that can be used to target audiences within social media platforms, NORC piloted a strategic initiative research study in 2020. A web survey was constructed using existing items from national surveys on individual’s health and online habits, as well as new items related to life changes during the pandemic. Two main audience groups that are generally hard to recruit through probability-based studies were targeted – young adults, ages 18-24, and people with low education (defined as anyone who has completed high school as the highest level of education or lower). Five sets of tailored ads with unique URLs that linked to a web-based survey were designed and launched via Facebook and Instagram. Two sets used basic targeting to recruit the sample and the other three used the enhanced targeting. This brown bag will present the design of the study, our approach to the ads and targeting, and what we learned through our examination of the differences between the samples obtained from basic and advanced targeting on the dimensions of recruited sample composition, survey estimates, and recruitment costs.