10: Longitudinal and Cross-Sectional Designs

cross-sectional designs

different cohorts of individuals tested at the same time

• cohort: group that has some characteristic in common (usually different age groups)

• between subjects

longitudinal design

one cohort (usually individuals of same age group) tested repeatedly over time

• within-subjects (same individuals at each test)

cross-sequential/time lag design

different cohorts tested at multiple points in life

• differences between cohorts in rates of change

• differences between cohorts at particular ages (holding ages constant)

• confounding variables still present

selectie survival

changes in population across time (due to deaths)

• problem for cross-sectional and longitudinal designs

• threat to external and internal validity

  • external → population no longer one that was sampled from

  • internal → weaker, less competent, less adjusted individuals typically live shorter lives

selective dropout

participants drop out of the sample

• problem for longitudinal designs

• threat to external and internal validity

  • external → if non-random, then systematic differences between stayers vs, leavers

  • internal → same reasoning as above

practice/testing effects

repeated testing makes participants task- or test-wise

• problem for longitudinal designs (and same individuals at each testing point

• threat to internal validity

• particularly issue for development of cognitive ability and skills

history effect

environment factor changes trend in scores over time

• problem for longitudinal designs

• threat internal validty

cohort effect

cohort is a confounding variable

• problem for cross-sectional designs

• threat to internal validity