# Why this matters Pamela Davis-Kean, APS president, frames a plenary strand for the 2027 APS Annual Convention around a practical problem: psychological science often studies a narrow slice of people and then assumes results apply broadly. That narrow slice is summarized by the WEIRD acronym (Western, Educated, Industrialized, Rich, and Democratic), introduced in 2010. Davis-Kean argues the field's default use of convenience samples—commonly undergraduates or local volunteers—builds bias into data and weakens claims of universality.
# What she proposes
# Practical advantages Using existing, population-representative datasets offers concrete benefits:
- Broader samples that better capture heterogeneity in experiences and outcomes.
- Larger sample sizes that provide statistical power for complex models and subgroup analyses.
- Cost and time savings compared with collecting new, large-scale data.
Davis-Kean uses her own work on how parents' educational attainment shapes child development to show why representative samples matter: effects can vary across children and contexts, and a college-town sample can miss that variation.
# Why psychologists don't do this more often Davis-Kean identifies several obstacles:
- Many large datasets weren't designed with typical psychology questions in mind, creating measurement or design mismatches.
- Psychology training often omits deep instruction on sampling, representativeness, and generalizability, which makes researchers default to convenience samples.
- Underpowered studies remain common, reducing rigor.
# Existing community responses There are emerging solutions within the field. Collaborative replication projects such as ManyLabs and shared archives like CHILDES demonstrate ways to pool data across teams and studies. These efforts expand the range of samples studied and allow testing of which findings generalize beyond WEIRD populations.
# Opportunities for researchers and attendees Davis-Kean notes that exhibitors at the 2027 APS convention will present data resources available to researchers. She encourages psychologists to consider whether existing datasets can answer or replicate important questions with broader samples, rather than assuming new data collection is the only path.
# Takeaway for practice If your research question depends on claims about broad populations or developmental trends, evaluate whether convenience sampling is adequate. Review available population-representative studies and archives for comparable measures, and consider collaborations or pooled-data approaches to gain power and representativeness. Strengthening training on sampling and generalizability will change what questions the field can answer.
# Invitation