PS375 Lecture Notes - Lecture 8: Partial Correlation, Longitudinal Study, Fluid And Crystallized Intelligence
Document Summary
Conclusions about age differences or age changes are valid only if the measured used are equally appropriate for the age groups being compared. Even if the content is age appropriate, test taking factors (ex difficulty in filling out lengthy questionnaire) may affect responses of older adults. The bulk of what we know about personality and aging is derived from rating measures (ex self or friend reports) rather than observation and experimental approaches. Biases (evaluation apprehension, response sets, misunderstanding of instructions, fatigue) can affect not only conclusions about the elderly but also conclusions about age differences in personality. What we are interested in are cause and effect relations (ex activity as a cause of satisfaction) yet what we are working with are nonexperimental designs that do not permit establishment of causality. Use various statistical methods (ex partial correlation) to control for potential confounding factors and specify more exactly the relation between the variables of interest.
