psych 350 exam 1 essays

sources of new knowledge

Intuition → instinctive feelings as knowledge

• Authority → knowledge from trusted experts

• Rational inductive → proof-like reasoning from axioms

• Empirical → scientific method, data-driven knowledge

• Only empiricism = accepted new knowledge in psychology

proof vs evidence

• Proof = absolute certainty (logic/math, needs axioms)

• Evidence = probabilistic, from empirical research

• No axioms in behavioral science → rely on evidence

• Best evidence = replication & programmatic studies

the research loop

Steps: library research → hypothesis → design → collect data → analyze → conclude

Novel hypothesis testing

Replication research = confirm results

Programmatic/convergent research = variations for consistency

types of validity

• Measurement validity → data reflect what’s studied

• External validity → participants, setting, tasks match target

• Internal validity → confounds controlled

• Statistical conclusion validity → correct results, depends on chance & other validities

random samples

• True random = full sampling frame + 100% return/no attrition

• Rarely achieved

• “Random sample” often means purposive frame + good return rate (15–75%)

IVs VS Confounds

• IV = causal variable studied (IV → DV)

• Confound = alternative cause interfering with IV–DV relationship

variables in a study

• Before study: DV, IV, potential confounds

• After study: confounds → constants, controls, or actual confounds

• Researcher’s resources/expertise decide control level

participant selection vs assignment

• Selection (sampling) → external validity (population)

• Assignment → internal validity (initial equivalence)

• Random assignment = required for causal inference

• Random selection = desirable, not required

precursor vs trade off models

• Precursor: internal validity first → external validity later

• Trade-off: every design choice affects both internal & external validity

• Preference depends on valuing causal vs. associative hypotheses

limits of testing casual hypothesis

• True experiment needed (random assignment, manipulation, no confounds)

• Sometimes impossible:

  • Technical limits

  • Ethical reasons

  • Resource limits

causal interpretability statements

• True: causal inference requires true experiment with no confounds

• False: random assignment ≠ guaranteed equivalence; confounds can arise during study

associative vs causal knowledge

• Colleague believes only causal matters → wrong

• Associative knowledge = crucial for prediction

• Predictions guide science, medicine, law, daily life

research process steps and validity

• Hypothesis: associative or causal

• Selection → external validity (population)

• Assignment → internal validity (initial equivalence)

• IV manipulation → measurement validity + external/internal

• DV measurement → same as IV

• Data analysis → statistical conclusion validity

attributes affecting causal interpretability

• Directly influence: random assignment, IV manipulation

• Indirect (don’t directly influence): participant selection, setting, data collection, analysis

• Harder with: field settings, long studies (ongoing equivalence issues)