Using Facebook Advertising to Recruit Representative Samples: Feasibility Assessment of a Cross-Sectional Survey
Bibliographic record
Abstract
BACKGROUND: Facebook has shown promise as an economical means of recruiting participants for health research. However, few studies have evaluated this recruitment method in Canada, fewer still targeting older adults, and, to our knowledge, none specifically in Newfoundland and Labrador (NL). OBJECTIVE: This study aimed to assess Facebook advertising as an economical means of recruiting a representative sample of adults aged 35 to 74 years in NL for a cross-sectional health survey. METHODS: Facebook advertising was used to recruit for a Web-based survey on cancer awareness and prevention during April and May 2018; during recruitment, additional advertisements were targeted to increase representation of demographics that we identified as being underrepresented in our sample. Sociodemographic and health characteristics of the study sample were compared with distributions of the underlying population to determine representativeness. Cramer V indicates the magnitude of the difference between the sample and population distributions, interpreted as small (Cramer V=0.10), medium (0.30), and large (0.50). Sample characteristics were considered representative if there was no statistically significant difference in distributions (chi-square P>.01) or if the difference was small (V≤0.10), and practically representative if 0.10<V≤0.20. The cost per recruit of Facebook advertising was compared with a quote for a random digit dialing (RDD)-recruited postal survey to determine if this method was economical. RESULTS: Facebook advertising is feasible and economical to conduct survey research, reaching 34,012 people, of which 2067 clicked on the ad, for a final sample size of 1048 people at Can $2.18 per recruit versus the quoted Can $23,316.05 for 400 recruits (Can $35.52 per recruit) via RDD. The sample was representative of rural and urban geography (P=.02; V=0.073), practically representative of age (P=.003; V=0.145) and income (P<.001; V=0.188), and over-representative of women (P<.001; V=0.507) and higher levels of education (P<.001; V=0.488). The sample was representative of the proportion of people with a regular health care provider (P=.94; V=0.025), diabetes prevalence (P=.002; V=0.096), and having had a colonoscopy or sigmoidoscopy (P=.27; V=0.034), and it was practically representative of smoking status (P<.001; V=0.14), and body mass index (P<.001; V=0.135). The sample was not representative of arthritis prevalence (P<.001; V=0.573), perceived health (P<.001; V=0.384), or time since last seasonal flu shot (P<.001; V=0.449). CONCLUSIONS: Facebook advertising offers an easy, rapid, and economical means to recruit a partially representative (representative or practically representative of 8 of the 13 characteristics studied) sample of middle-aged and older adults for health survey research. As Facebook uses a nonrandom targeting algorithm, caution is warranted in its applications for certain types of research.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.251 | 0.204 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.003 | 0.003 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.003 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".