Module 2 - Research Strategies in Psychology PDF

Preview

Summary

Instructor: Sidni A. Justus, M.S....

Description

Module 2 - Research Strategies in Psychology Methods of Psychology Research methods == empirical approach Scientific method -> seeking empirical evidence Scientific observation Systemic Intersubjective The Scientific Method 6 basic elements: Making observations Defining a problem Proposing a hypothesis (operationally defining variables) Gathering evidence/ test the hypothesis Building a theory (summarise, explain, predict) Publishing results Operational Definitions How we decide to measure our variables -> There are usually hundreds of ways to measure a variable e.g.

Research can be cyclical Replication: repeating a study and getting the same (or similar) result Strengthening support for some theories & weeding out the weaker ones

What makes a good theory? Should be falsifiable Possible to test hypotheses generated by the theory that prove the theory is incorrect Should produce testable hypotheses Should be simple Occam’s Razor/ Law of parsimony - if two competing theories exist, we prefer the simpler of the two Descriptive Research Involves observing behaviour in order to describe that behaviour objectively and systematically 4 goals of psychology: Describe Understand Predict Control Identifying “what” without necessarily understanding “why" Naturalistic Observation Observing a person or an animal in the natural environmental context Often qualitative, but can be quantitative as well Lack of manipulation Ecological validity: generalizable to real life Concerns with Reactivity: The Hawthorne Effect: changes in behaviour that occur when people know that others are observing them >> perform better Observer bias: Systematic errors in observation that occur because of an observer’s expectations >> cultural norms Experimenter expectancy effect: When observer expectations change the behaviour being observed Solutions: “Blind” observers Inter-rater reliability Participant Observation When the researcher becomes involved in the situation Pro’s: more enriching, greater access Con’s: even greater risk for bias, low reliability Case studies Extensive examinations of an unusual person or organisation Goal: to describe events or experiences that lead up to or result from the exceptional/ unique aspect

Con: not generalizable Surveys Self-reports, interviews, questionnaires Ranges from demographics (e.g. ethnicity, age, religious affiliation) to past behaviours, attitudes, etc. Pro’s: interactive, lots of data in little time, easy to administer, cost-effective Con’s: potential for bias: Sampling error or bias Response bias: socially desirable bias Correlational Research != Causation Examines degree of naturally existing relationship between two (or more) variables Correlation coefficient: Range: [-1.0, +1.0] Size indicates strength of relationship: Closer statistic is to + -1.0, the stronger the relationship Correlation of 0.00 >> no relationship between variables Sign indicates the direction of the relationship, not whether it is “good” or “bad" Positive correlation (+): variables move in the SAME direction, up or down tgt Negative correlation (-): variables move in OPPOSITE directions

Potential Problems: Directionality Problem: hard to know direction of relationship e.g. less sleep cause more stress OR more stress cause less sleep Third Variable Problem: instead of A causing B, it might be a third variable, C, that causes both A & B Experimental Research Used to identify the cause-and-effect relationship Manipulate one variable to measure effect on a second variable Basic steps: Directly vary a condition believed to affect behaviour Create 2 or more groups of subjects

Record whether varying the condition has any effect on behaviour Independent variable (IV): manipulated by experimenter >> Experimental group Dependent variable (DV): measures the results of the experiment >> Control group Extraneous variable (confound): conditions a researcher wants to prevent from affecting outcomes of the experiment Placebo effect Control: Random sampling: each member of population of interest has equal chance at being chosen to participate Random assignment: equal chance of being in either experimental or control group Balance personal differences in the two groups Theory vs. Hypothesis??? Meta-analysis??? Research Ethics Do not harm; Accurately describe risks to potential participants; Ensure that participation is voluntary; Minimise any discomfort to participants; Maintain confidentiality; Do not unnecessarily invade privacy; Remove any misconceptions caused by deception (debrief); Provide results and interpretations to participants; Treat participants with dignity and respect....