Combining Studies, and Tracking Change
When many individual studies address the same question, two related methods help synthesize the evidence. A systematic review comprehensively and methodically identifies, evaluates, and summarizes ALL available research on a specific question, following an explicit, pre-specified search and inclusion process (to avoid cherry-picking only supportive studies). A meta-analysis goes a step further, STATISTICALLY combining the numerical results (typically effect sizes) from many individual studies into one overall, more precise estimate — a study of studies, effectively pooling many smaller samples into a much larger combined one, increasing statistical power well beyond what any single original study achieved alone.
| Method | What it does |
|---|---|
| Systematic review | Comprehensively identifies and summarizes existing research |
| Meta-analysis | Statistically combines numerical results across studies into one estimate |
Developmental psychologists (and psychologists generally) also choose between two contrasting designs for studying change over time. A longitudinal design follows the SAME group of individuals over an extended period, directly tracking real change within each person — but it is time-consuming, expensive, and vulnerable to participant dropout ("attrition") over the years. A cross-sectional design instead compares DIFFERENT age groups all measured at a single point in time — much faster and cheaper, but vulnerable to cohort effects: apparent "age differences" that are actually generational differences in life experience, unrelated to aging itself (comparing today's 20-year-olds to today's 70-year-olds conflates true age effects with the very different historical contexts each group grew up in).
| Design | Tracks | Key strength | Key weakness |
|---|---|---|---|
| Longitudinal | Same people, over time | Directly measures real change | Slow, costly, attrition |
| Cross-sectional | Different age groups, one time point | Fast, cheap | Confounded by cohort effects |
Common pitfall: treating a cross-sectional age difference as automatically proving something changes WITH AGE. Because different age groups in a cross-sectional study also grew up in different historical eras, an observed "age difference" could just as easily be a cohort effect — a longitudinal design, tracking the same people over time, is needed to cleanly separate true aging effects from cohort effects.