Friday April 25 y Research Methods Exam 1 Material The Science Game The Cycle of Scientific Progress 3 non scientific modes of understanding their limits experience what seems to have been true for from from the past intuition what feels like the true answer tradition authority what does my culture believe What do trusted people claim empirical evidence objective observation or measurement that contributes to either verifying or falsifying a claim about what s true empirical testing any situation or procedure that creates empirical evidence which allows a claim on true to be verified or falsified golden rule scientific claims must be subject to empirical test that produce empirical evidence golden assumption if an object of study exists in nature it is knowable we could perfectly predict and control the outcome The role of If Then reasoning in Theory Data cycles if then reasoning enables us to test refine theories Basic Applied Cycle if then reasoning also enables us to take basic findings and apply them to more realistic scenarios external validity generalizability how widely does the claim generalize Is it true for people other than the sample True for all students All americans All people everywhere Operationalizing Measures psychological construct any explanatory variable that is not directly observable or tangible ex happiness intelligence addiction operationalized how to turn abstract constructs into specific measurable instances that are both reliable and valid Ways to Operationalize psychological constructs Self report recording people s answers to verbal questions about themselves in a questionnaire or interview PROS easy low cost large anonymous samples 1 Friday April 25 y CONS open to biases memory distortions laziness Observational recording observable behavior or physical traces of behavior PROS recorded with less interference more shielded from respondent bias CONS complicated to collect subject to experimenter bias discrepancies over behavior Physiological recording biological data believed to be associated with a a construct PROS hard to control fake precise credible CONS expensive time consuming may require technical expertise machinery most sensitive to uncontrollable sources of error Validity vs Reliability content validity the operationalized measure must capture all parts of a defined construct predictive validity operationalization should predict future outcomes related to happiness internal validity third variable problem there must not be a third variable connected to both A and B independently Reliability inter rater reliability measures the agreement between two independent raters in scoring behavioral or observational measures test retest subjects should score similarly to items meant to assess the same construct Population vs Sample vs Representative Sample representative sample sample that is like the populations with respect to the measurement being made increases generalizability of results sample subset of the population of interest measured by survey population the entire set of people to interest measured by census simple random sampling every person has a specified probability of being included stratified random sampling subpopulations are identified and proportions are assessed final sample reflects the proportions of subgroups represented in population 2 connivance sampling using whatever sample is accessible but randomly assigned to condition in experiments not appropriate when generalizing broadly Friday April 25 y Survery items clarity core vocabulary unbiased items balanced items double barreled asking about two things at one time and negative phrasing often requires work to figure out what agree or disagree means Response Biases social desirability response acquiescence when similar responses to a few initial questions establish a pattern of responding Descriptive Stats central tendency mean medium mode mean average medium more representative in skewed distribution Mode most repeated standard deviation average amount that each score differs from the mean Shape of Distributions unimodal normal one mode or one dominant central tendency ex IQ age of first word positively skewed tail pulled high by high outliers ex Age of people at a club negatively skewed tail pulled by low outliers ex scores on exam bimodal two modes or values clearly occurring most frequently ex height in class age of people at chuckie cheeses Correlational Research positive correlation high z scores on the x variable match with high z scores on the y variable negative correlation high z scores on x match with low scores on y strong r 0 70 or higher 3 Friday April 25 y moderate r 0 30 to 0 70 weak r 0 30 or lower Four uses of correlational research 1 Identifying or exploring potential causal relationships that we were unaware of esp in exploratory studies 2 When providing converging evidence for or against a theoretical prediction 3 When establishing the reliability of a measure Test retest reliability and inter rater reliability 4 When experimental tests are unfeasible Multiple Regression Two uses of multiple regression 1 rule out variable 2 to predict unknown and unknowable future outcomes predictor variable X s all the variables we measure that may predict or explain life satisfaction criterion variable Y what we are trying to explain or predict standardized beta tells the predicted amount of change in the z score of Y for every 1 unit change in the z score of X R squared coefficient of determination tells us the total amount of variability in Y that can be accounted for by all the variables in the model written as a percent Material from Exam 2 independent variable dependent variable Internal Validity Threats selection bias two groups may be predetermined by a characteristic other than the IV maturation effects practice fatigue boredom effects sometimes participants change overtime and it has nothing to do w your manipulation counterbalance varying the order of the item completion and tracking it in order to rule out order effects attrition losing subjects overtime in a way that may be systematically related to the IV or DV Experimenter Effects 4 Friday April 25 y randomized random assignment of participants double blind neither the participants nor the researchers who evaluate them know who is in the treatment group and who is in the comparison group placebo control experiment is exposed to initial treatment EX a sugar pill between groups all participants experience
View Full Document
Unlocking...