Health Psychology lecture 2 Research Methods I Operational Definitions A B C Definition observable behaviors representing concepts that researchers can measure The scientific method 1 Specify hypothesis 2 Form operational definition 3 Collect data 4 Analyze data 5 Form theory seeks to explain why Correlation vs Causation 1 Correlation association between two variables a Smoking lung cancer b c Vitamins fewer colds Anxiety heart disease 2 Causation only if experiment has been conducted a Comparison between treatment group and control group D Randomized trials 1 True experiments a Possible in health psychology 1 Depends on the question a Does having a hostile personality cause heart disease i No you cannot randomly assign personalities b When it s not possible Quasi experimental correlational design 1 Characteristics a b c Compare pre existing groups i Non manipulating Temporal priority i What occurs first in time Dose to response relationship i How much heavy smokers are more likely to develop cancer Consistent with existing knowledge Consistent in different populations d e f Animal analogs g Intervention has an effect i Does quitting smoking reduce risk of lung cancer 2 Artifacts a b Findings due to research design but attributed to reality Examples i Sample selection bias where participants are from Depression among widowed men Found in nursing homes already depressed in nursing 1 2 3 homes Come to wrong conclusion ii Selective attrition 1 Leaving the study 2 Manipulate independent variable measure how it influences dependent variable 3 Must have at least 2 groups with different levels of IV 4 Random assignment to eliminate bias 5 Problems a Sample size small difficult to generalize b Identifying moderating factors 1 Moderator relation between IV and DV is different for different groups of people E Survey method 1 Asking people to self report 2 Advantages a Efficient b 3 Disadvantages a Leading questions 1 b Response options c Cannot infer causality Response bias d 1 2 F Case study Self presentational Recall bias Can measure constructs that are not directly observable Indicating what you re looking for by the way you ask 1 How HIV was discovered 2 In depth analysis of one subject or more a Both cases usually rare 3 Advantages a More complete and unique information b Generates new hypotheses 4 Disadvantages Cannot infer causation a Cannot generalize b c Observer bias Observational methods 1 Describe and measure people s behavior systematically 2 Advantage G a Rich description of behaviors 3 Disadvantage a Cannot infer causation b c Observe bias People behave differently H Archival methods 1 Use of existing data sources 2 Advantage a Fast access to large amount of data 3 Disadvantage a Cannot ask questions researchers want to ask I Prospective vs Retrospective research 1 Retrospective participants report something in the past 2 Prospective predictor measure assessed well in advance of outcome 1 Cross sectional assessing samples of different ages at one point of time a Disadvantages 1 2 Outcome already known correlational a Disadvantage 1 Correlational J Longitudinal vs Cross sectional design a Disadvantages 1 2 Cohort effect Correlational a Disadvantages 1 Correlational 2 Longitudinal conducting study over time K Field vs Lab research 1 Lab conducting study in lab setting a Can control conditions b Problem 1 Not realistic 2 Field conducting study in natural setting a More real life like b Problem 1 Lack control L TRADE OFF 1 You can gain internal validity but it decreases external validity less like real life 2 INTERNAL VALIDITY a Likelihood that changes in DV were caused by IV 3 EXTERNAL VALIDITY a Likelihood that same results would be obtained using the same study with other people and other situations b Generalizability does what you did in your study apply to other samples