PSYC 31773 Lecture Notes - Lecture 2: Serendipity, Big Data, Informed Consent
23 Jun 2018
School
Department
Course
Professor
Research methods in IO Psychology
Why research- to answer questions
3 goals of science
oDescribe
oPredict
oExplain
Can you apply the results to a broad population?
Generalizability- the extent to which conclusions draw from on research study spread or
apply to a larger population
Empirical research process
State problem
oTheory- statement that proposes to explain relationships among phenomena of
interest
Design research study
oResearch design- plan for conducting research to learn about a phenomenon of
interest
oNaturalness of the setting- does the setting reflect the setting in which the point of
interest normally occurs
oDegree of control
oExternal validity- degree to which the research can generalize to other contexts
oInternal validity- how confident you are that study results are based on variables
Measure variables
Analyze data
Conclusions from research
Primary Research Methods
Laboratory experiment
oManipulate variables in laboratory experiment
oAssign subjects to experimental or control condition
oDetermine causality
oProblems of generalization
Quasi-experiment
oManipulate variables in natural environment
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
3 goals of science: describe, predict, explain. Generalizability- the extent to which conclusions draw from on research study spread or apply to a larger population. Empirical research process: theory- statement that proposes to explain relationships among phenomena of interest. Primary research methods: manipulate variables in laboratory experiment, assign subjects to experimental or control condition, determine causality, problems of generalization. Quasi-experiment: manipulate variables in natural environment, less control, more generalization. Questionnaire: subjects respond to written questions from an investigator, response rate can be an issue, responses may not be accurate, internet surveys can help with both. Observation: monitor employees in natural setting to develop new experiment, generate ideas for further testing, people alter behavior when they know they are being watched, tells you that things happen, not why they happen. Meta-analysis: combines many individual studies, statistical study of studies. Aggregate results: estimates true relationships, subjective decisions, file drawer effect of non-published studies, problem from different levels of analysis of original studies.
