click below
click below
Normal Size Small Size show me how
RMDA ch 4
| Term | Definition |
|---|---|
| constructs | intangible, complex ideas (e.g. mood, personality) |
| validity | did we measure what we intended to? accuracy |
| construct validity | ensure the test you're using is appropriate for what you're studying; are the studied constructs appropriate? |
| internal validity | how you're interpreting what you're measuring; whether conditions in your study caused the results |
| external validity | how well results match the real world; how generalizable results are; ecological validity |
| construct validity | if the measure addresses construct it was designed for |
| accuracy | how well you hit the target correctly |
| precision | how well you can hit the same spot, even if incorrect |
| variable | event, situation, behavior, or characteristic that varies; at least 2 values; with humans, only thing not variable = species |
| operational definition | statement that determines how a variable will be measured to determine its existence, duration, or quantity |
| operational definitions are | quantifiable; objectively measured; behavioral; very clear/specific |
| e.g. operational definition "wash hands thoroughly" | warm water, soap, 30 sec, scrub hands together (lack of any of those doesn't meet criteria/definition) |
| why are operational definitions important? | explains abstract constructs; reproducibility; communication; need to know what happened to understand findings |
| e.g operational definition "memory improvement" | pre and post test of words to memorize (different words for post-test) and improvement is more words memorized |
| correlational or nonexperimental method | measure main variables |
| experimental method | manipulate main variables |
| experimental control | all extraneous variables kept constant |
| SES | similar economic status |
| logic of experimenters | if you have experimental control, then difference between groups must be due to relationship between variables |
| confounds | third variables; variables you aren't measuring, that may impact results |
| experimental methods | manipulate independent variable and measure dependent variable |
| independent variable (IV) | cause; the variable being manipulated/ whose effect is being studied |
| dependent variable (DV) | effect; measurable behavior; may change in response to independent variable |
| random selection | when recruiting from large population, everyone has equal chance of being in sample; no cherry-picking |
| random assignment | randomly putting people in experimental vs control groups; reduces bias; equate experimental and control groups besides IV |
| between-subjects design | different people in each group |
| within-subjects design | same people in experiment conditions; same person is their own control |
| field experiment | type of experiment where IV is manipulated in natural setting (e.g. public places, like classrooms, parks) |
| ethical concerns of field experiment | stress; didn't sign up for it |
| e.g. students watched cartoon with others or alone and rated how funny it was (IV, DV, within or between subjects) | IV: tv watching condition, levels = alone or w/ others; DV = rating of how funny cartoon was; between-subjects bc dif groups ("or") |
| uncertainty | don't know response and population ahead of time |
| variability | differences in responses; range of responses |
| random variability | ppl's responses can vary in any random way from each other; lots of factors determine random variability |
| correlational/ nonexperimental pros | see relationship, observe in natural context, study participant variables that can't be manipulated |
| participant variables | aka subject variables, personal attributes; characteristics of individuals; e.g. age, gender; cannot be manipulated |
| correlational/ nonexperimental cons | no causality; third variable problem; difficult to control many aspects of situation |
| experimental pros | control; determine causal mechanisms; attempt to eliminate impact of confounding third variables |
| experimental cons | high control may create artificial atmosphere; can be unethical or impractical |
| general experimental research hypothesis | the experimental (values of DV) caused only by experiment itself (values of IV) |
| to determine covariation | baseline tests and control group |
| APA format | there was (not) a statistically significant correlation between variable A and B, r(df) = x, p ><= .05 or .10; such that... |
| risk-benefit analysis | compare potential risks with potential benefits; minimize risks, maximize or equalize benefits |
| correlation | the measure of the extent to which 2 factors vary together and how well either factor predicts the other |
| correlation coefficient | -1.00 to +1.00; closer to 0 is no significance; +/- .6 is moderate/strong significance; strength of variable relationship |
| positive correlation | both variables increase or decrease; e.g. as attendance increases, so does class grade |
| negative correlation | inverse relationship; as one variable increases, the other decreases; e.g. as freq of exercise increases, anxiety decreases |
| internal validity (textbook) | ability to draw conclusions abt causal relationships from results of study |
| high internal validity | strong causal inferences can be made |
| causation requires | temporal precedence, covariation of cause and effect, eliminate alternative explanations |
| positive linear relationship | as value on variable increases, so does value of second variable |
| negative linear relationship | increase in value of one variable, decrease in value of other variable |
| curvilinear relationship | direction of relationship changes at least once; variable A increases, then systematic increases and decreases of variable B |
| no relationship | graph is flat line; unrelated variables vary independent of each other |
| third variable problem | there may be a relationship between 2 variables bc a third variable causes them both; extraneuous |
| confounding variable | uncontrolled third variable operating; experimental method helps control it |
| external validity | extent to which results can be generalized to other populations/settings, or the real world |
| some variables shouldn't be manipulated, such as | child-rearing practices; types of punishment; anything that can cause harm |