Lecture 1 Research Methods PDF

Lecture 1 Thursday, September 19, 2024 Confirmation bias – tendency to want to believe we’re accurate about our views about the worlds ○ Keeps us psychologically happy and sane ○ Confirm...

Lecture 1 Thursday, September 19, 2024 Confirmation bias – tendency to want to believe we’re accurate about our views about the worlds ○ Keeps us psychologically happy and sane ○ Confirming our beliefs ○ Can feel threatened when our beliefs are threatened What is science ○ Process or method for generating a body of knowledge ○ Psychology relies on formal, systematic observation to answer questions about behaviour Objective: use control and manipulation ○ White swan analogy Anything can be disproven Just because you haven't observed others ○ Not about certainty; about confidence ○ Use probability system; can never be 100% certain about something Goals: ○ Description “What is it” ○ Explanation “Why is it” ○ Prediction “What if…” Doing a study, testing the hypothesis ○ Control “So then, could we…” Putting into action Applied science/psychology What is theory ○ Contains interrelated constructs (concepts), definitions, and proposition ○ Presents a systematic view of a phenomenon by specifying relations among variables ○ Explains and predicts the phenomenon ○ The questions that arise come from theories What makes a good theory? ○ want to be specific and precise Basic terminology and concepts ○ Induction: Data > Theory Most of the time, we have a theory. We’re going to use that theory to derive some hypotheses. What happens when we just have some observations but we don’t know why ○ Deduction: Theory > Data Testing that theory over and over again ○ Causal inference – can be made when data indicate that a causal relationship between two variables is likely Variables Independent variable (IV) Dependant variable (DV) Extraneous variable ○ We need to establish experimental research to establish a causal relationship Has to be based on an experiment ○ Control – Important to ensure that a causal inference can be made ○ Ways to control Hold extraneous variable constant Systematically manipulate extraneous variables (make part of experimental design) Use statistical control  Sometimes, instead we need to include it as variable ○ Validity - do you have what you think you have ○ Internal validity – extent to which causal inferences can be drawn about variables ○ External validity – extent to which results generalize to other people, settings, time Outside of your sample Is that what actually happens in the real world? We seek to rule out alternative explanation by controlling more elements of the study to increase out internal validity ○ What happens to the external validity of our result Research designs [insert process diagram] – stage model of the research process Types of research designs ○ Experimental methods Lab experiments Field experiments Quasi-experiments ○ Observational methods Naturalistic observation Archival research Survey research Types of research designs Characteristics of experimental methods: ○ Random assignment Each participant has an equal chance of benign assigned to each condition ○ Manipulation of one or more IV Systematic control, variation, or application of one or more IVs ○ Lab experiments Conducted in aithgly- controlled setting ○ Field experiments Conducted in a naturally occurring, real-world setting ○ Quasi-experiments Field experiments without random assignment E.g. using pre-existing groups No one right way to do it ○ A question of what works best for your purposes Observational methods ○ Correlation designs, description research ○ No random assignment or manipulation of IVs Can observe relationships but NOT causality ○ Can be useful in predicting behaviour ○ Common in field settings Data Collection techniques ○ Naturalistic observation Observation of someone or something in its natural environment ○ Participant observation Observer tries to “blend in” with subjects ○ Unobtrusive naturalistic observation Researchers observes from the sideline without interfering Does not try to blend in ○ Case studies Examination of single individuals, groups, companies, or societies Main purpose is typically description Not typically used to test hypothesis Provide details about a typical or exceptional firm, group, or individual ○ Archival research Answering a research question using existing, secondary data sets  Very efficient form of data collection but Lack control over quality of data Cross-sectional – one point in time Longitudinal – multiple time periods ○ Surveys Most frequently used method of data collection in IO Pros Cheap and easy Large samples Anonymity Cons Low response rates Limited range of data Respondents cannot ask questions ○ Interviews Face to face or over the phone/internet Can be very time consuming and expensive Benefits over questionnaires Higher response rates Ambiguity about questions can be resolved Measurement issues: reliability ○ Consistency or stability of a measurement ○ No measure has perfect reliability Measurement error renders measurement inaccurate or unreliable Outcomes cannot be accurately predicted with variables that are not measured well ○ Test-retest reliability Coefficient of stability Participants are given a test at time 1 and then given the same at time 2 ○ Interrater reliability Consistency with which multiple carters view/rate the same behaviour or person Measured by examining the correlation between ratings of multiple judges Especially relevant in performance appraisals ○ Parallel-forms reliability Extent to which two independent forms of a test are equivalent measures of the same construct Important to ensure that the two tests are measuring the same thing Often referred to coefficient of equivalence or equivalent-forms reliability Measurement issues: validity ○ About accuracy ○ Construct validity – Extent to which a test measures the underlying construct it was intended to measure ○ Ways to assess construct validity Content validity Ensures test covers all relevant aspects of the construct Criterion-related validity Predictive validity – extent to which test scores obtained at one point in time predict criteria obtained at some later time Concurrent validity – extent to which a test predicts a criterion that is measured at the same time the test is administered STATISTICS Central tendency – ○ Help to characterize a "typical" or central value in a set of data Mean: arithmetic average of a group of scores Median: score in the idle of a distribution Mode: most frequency single score in a distribution  Shape of distributions ○ Normal distribution – symmetrical, bell-shaped distribution where most of the data is clustered around the mean ○ Mean, median, and mode are the same Correlation coefficient (r): ○ strength of the relationship between two variables Indicate direction and magnitude ○ Correlation coefficient – range from -1 to 1 Regression ○ Used to predict the value of one variable based on another ○ Coefficient of determination (r2) Percentage of variance accounted for by the predictor Allows prediction of one variable from another Q: A variable that is systematically manipulated by the experimenter as an antecedent to other variables is called an… a. Dependent variable b. Independent variable a. Causal inference b. Extraneous variable Q: the systematic control, variation, or application of independent variables to different groups of participants is called: a. Random assignment b. Performance enhancement c. Manipulation d. Methodology