Final Study Guide Research Design
Experimental Research
Researchers manipulate independent variable - 2 levels And measure the other (dependent variable) Give treatment to participants and observe if it causes changes in behavior Compare experimental group (w/ treatment) with a control group (no treatment) Can say IV caused change in the DV
Independent Variable The variable whose impact you want to know ‘Stimulus’ ‘Input’ Variable The variable you manipulate in experimental research
Dependent Variable The variable whose changes you want to know You measure it ‘Outcome’ ‘Response’ variable
Random Selection –A way to choose your sample of study –Any member of population has equal chance of being selected Random Assignment –A way to assign participants in sample to the various treatment conditions (groups will receive different level of IV) –Any member of your sample has equal chance of being assigned in any treatment group
Internal Validity Ability of your research design to adequately test your hypothesis Showing that variation in I.V. CAUSED the variation in the D.V. in experiment In correlational study, Showing that changes in value of criterion variable relate solely to changes in value of predictor variable
Confounding Whenever 2 or more variables combine in a way that their effects cannot be separated = confounding. Thus, the teaching method study as designed lacks internal validity. You don’t know if the change in the DV is from the IV or from confounding variable
Quasi-experimental research Naturally occurring conditions (IV change) No control over variables influencing behavior (confounding variables) –Another variable that changed along with the variable of interest may have caused the observed effect –(NO random assignment)
Non-Experimental Research
Non-experimental Correlational research Determine whether 2 or more variables are associated, If so, to establish direction and strength of relationships Observe variables as they are, –can’t manipulate them
Research design Manipulate IV Random Assignment Experimental (Causal) x x Quasi-experimental x Non-experimental / –Correlational Predictive Descriptive
Causal - (Experimental) one variable directly or indirectly influences another. Correlational - (Non-experimental) Changes in one variable accompany changes in another. –A relationship exists. Don’t know if either variable actually influences the other.
TERMS Population Universe/entire set of people you want to draw conclusions about Sample Subset of the population People actually in your study Sampling error Differences between sample & population
Sampling Drawing a subgroup from a population (vs. Census)
Probability vs. Non-probability Simple random Systematic random Stratified random Cluster Convenience Snowball Quota Purposive Probability SamplingNon-probability Sampling Population info Available Population info Not available
Representativenss & Generalizability Representativeness = Resemblance to the population characteristics Generalizability = An ability to generalize the results of your study to the whole population High representativeness = High generalizability Probability sampling allows higher representativeness than non-probability
External Validity Degree that results can be extended beyond the limited research setting –Generalizable –Based on sample ( rats, college students, whites, males, lab setting)
Non-Probability Sampling
Convenience Sampling Get available people in the population Low representativeness / generalizability
Quota Sampling Predetermine the proportion of groups in the sample (e.g., male 50%, female 50%)
Conceptualization & Operationalization Idea Conceptualization Operationalization Clarification
Operationalization From complex variable to series of simpler variables Redefining a variable in terms of steps to measure Conceptual definition Operational definition What the researcher must do to MEASURE it
Types of Measurement Validity Face validity Content validity Predictive Concurrent Convergent Discriminant Judgmental Empirical (Criterion- related)
O bserved score = T rue score + E error Observed = measured score, result True = “true”, actual, exact state Error = measurement error “O = T + E” rule
Reliability of a Measure Degree to which a measure (score, observation) is affected by error A reliable measure has little or no error
Types of Reliability Interobserver (interrater) reliability Test-Retest reliability Parallel-forms reliability Split- half
Inter-rater Agreement Consistency between measurements by two or more observers Different observers watch the same sample of behavior Compute proportion of time both observers recorded the same behavior as happening # agreements # agreements + # disagreements (# of observations) Training needed for observers
Increasing reliability Increase number of items on your questionnaire (no 1 or 2 item measures) Write clear, well-written items on survey Standardize administration procedures –Treat all participants alike –Timing, procedures, instructions alike Score survey carefully -- avoid errors
Valid and Reliable A good measurement Measures what it should measure in a consistent way
Reliable but Invalid Your measurement is consistent, but not measuring what it is supposed to measure
Research Report Structure Abstract Introduction Method Results Discussion Reference