click below
click below
Normal Size Small Size show me how
ch 8 quiz RMDA
| Term | Definition |
|---|---|
| internal validity | can we validly determine what is causing results of experiment? |
| general research hypothesis | experimental outcome (values of DV) are caused only by experiment itself (IV) |
| confound | 3rd variable; unmeasured variable other than IV actually led to results |
| confounds within experiment | procedural differences per group; biased assignment |
| confounds outside experiment | history; maturation; cultural change; can't really control (e.g. pandemic) |
| internal validity exists when | results of experiment can be confidently attributed to effect of IV; temporal precedence, rule out alt explanations, covariation |
| post-test only design | obtain 2 equivalent (at least) of participants; eliminate selection differences; introduce IV and measure effect on DV |
| pre-posttest design | pre-test given to every group prior to intro of manipulation; quickly measure changes; assures equivalent groups in beginning |
| advantages pre-posttest | mortality (dropout factor); assess equivalency of groups with small sample times; can be used to select participanta |
| disadvantages posttest | time consuming; awkward; sensitizes participants to what is being studied (demand characteristics) |
| demand characteristics | things abt study could be revealed; decrease by adding filler tasks |
| independent groups design (between subjects) | participants participate in only one group |
| repeated measures design (within subjects) | same participants in all conditions |
| advantages repeated measures/within | fewer participants; conditions identical bc participant is control |
| order effects | ordering of trial affects DV |
| practice effects | practice from first task helps lead to improvements |
| fatigue effects | deterioration bc ppl become bored, tired, distracted |
| carryover effects | previous condition produces a change that gets carried over to later conditions (e.g. taste testing) |
| complete counterbalancing | all possible trial orders presented; typically done with gender or other participant variables; used with repeated measures |
| latin squares | when you have too many possible orders; can be used for condition and trial order; each condition once per row + column |
| matched pairs design | goal is to match people on a participant characteristic; not completely random assign; e.g. age, gender, SES |
| cross-sectional | one observation per subject; same age no groups, or different age groups, or same age different ability groups |
| longitudinal | at least two observations per subject (more than one day); age, ability, training, microgenetic |
| microgenetic | closely spaced observations (days apart); as participants transition from one skill to another |
| give pretest? | no = post-test only; yes = pre-test post-test design |
| same people in all conditions? | no = independent groups/between subjects; yes = repeated measures/within subjects; sorta = matched pairs design |
| test once or more? | once = cross-sectional; more than once = longitudinal |