PSY 325 3rd Edition Lecture 3 Outline of Current Lecture I. Experimental Designsa. Self-selection bias, Independent Variable and Dependent Variable II. Quasi Experimental Designsa. Variables in These DesignsIII. Research Methods in EpidemiologyIV. Additional Issues in Health Related IssuesV. Determining CausalityVI. Role of Placebos in Health ResearchVII. Placebos and Research DesignsVIII.Health Carea. Seeking and Receiving Medical Attentionb. Segall’s Three Rights and Dutiesc. Factors influencing Seeking Medical AttentionIX. Seeking Medical Information From Non-medical Sources X. Receiving Medical Carea. Receiving Medical Care in the Hospitalb. Children and Hospitalizationc. Adherence: What is It?d. How is Adherence Measurede. What are the barriers to Adherence?Current LectureI. Experimental DesignsAssist researchers in determining if one variable directly causes another variable.In true experimental designs, the researcher - Must systematically manipulate or vary at least 1 independent variable- Must assign people to groups in a way that ensures their initial equivalence.- Must control for extraneous variables that may affect the results (i.e. experimental control)That way, you can be pretty sure the ONE THING you changed was responsible for the outcome of the experiment. These notes represent a detailed interpretation of the professor’s lecture. GradeBuddy is best used as a supplement to your own notes, not as a substitute.Compare at least two groups to be able to draw cause and effect conclusions- The experimental group receives treatment- The control group does not receive treatment- Participants are randomly assigned to either the control or the experimental groupWhat does this help us ensure?*Self-selection bias—those looking for choosing the experimental rather than control group is biased. Randomly assigning will give you some of both who want and don’t want change. The manipulative variable is the one that causes the effect.*Independent variable—Condition of interest; is manipulated to see its effect on the dependent variable (what you measure).*Dependent variable—The outcome, the behavior or response that is being studied, what you’re interested in.Main Point:The effect of X on Y:X causing variable (IV)Yeffected variable (DV)Questions to Ponder:Does public speaking cause an increased heart rate?Does smoking cause lung cancer?How would you do an experiment to test this?We cannot ethically manipulate certain variablesII. Quasi-experimental designs (ex post facto designs)Used to approximate cause and effect relationships when the variable of interest cannot be manipulated due to practical limitations. These are considered acceptable alternatives.Ex post facto designs- Similar to experimental designs except does not involve manipulation of IV- Instead choose participants who already differ on variable of interestA. Variables in these designs*Subject Variable—equivalent of the independent variable in experimentsExample—gender, sexual orientationDependent variable—measured outcome (same)Ex-post facto designsQuite common in health related research, probably more so than true experiments.Limitations of such designsDepends upon how many “confounds” the researcher is controlling forConfound is a variable that essentially ruins your results. Way we get around it is to control for these things. In other words, these designs are only as good as the number of alternative explanations one controls forThis is typically accomplished by “matching” participants in each condition on theoretically important variables. Or measuring important constructs and controlling for them in analyses (less precisebut sometimes is the only option).Measure confounding variable then control fro them later. Less precise but its acceptable if it’s the only option. Sometimes you don’t expect it. Can’t go back and re-run the study.III. Research Methods in Epidemiology*Epidemiology—Branch of medicine that investigates factors contributing to increased health or the occurrence of disease in a particular popEtiology—the study of causation or originThese are two of the main goals of health psychology:Epidemiologists identify risk factors of diseases…health psychology example: demographic and behavioral factors that increase the risk for heart disease?Basic TerminologyTwo primary ways to describe the rates of diseases*Prevalence: represents the proportion of the population that has a disease or condition at a specific timeLifetimePoints (often 1 year)*Incidence: represents the number of new cases of a disease or condition during a specific period. Epidemiology ResearchTypes of methods: - Randomized controlled trials, clinical trials- Natural experiments (like quasi experiments)- Meta-analysis- Observational MethodsBasically the same as correlational methods we discussed earlier... however it cannot be used to demonstrate causation.Two types of observational methods:1. Prospective studies2. Retrospective studies (begin with participants who have a disease now and ask them to report on behaviors from the past) Problem is that they don’t allow us to draw inferences about causation. Randomized Controlled Trials (RCT)RCTs are basically experimental studiesParticipants are randomly assigned to either a study group or control group.Often considered the “gold standard” of research design.Clinical trials or Randomized Clinical Trials Same as RCTs but investigates the effects of a new drug or treatmentAgain why are RCTs and experimental designs preferred? Helps eliminate potential confounds like self-selection biases. Natural Experiments and Meta-Analysis- *Natural Experiments—similar to ex-post facto design involve the study of natural conditions (public health and public policy studies)- *Meta-analysis—evaluation of many research studies on the same topic to draw an overall conclusion on that topic; a composite of many studies.These studies are good for summarizing the literature particularly those with mixed findings. It gives us an estimate of the strength for a particular relationship in research studies. IV. Additional issues in health related ResearchDetermining causality in health research and placebosIf we cannot conduct true experiments due to ethical and practical concerns, how do we fully determine causality?Risk Factor ApproachTwo primary ways to convey risk:Absolute riskRelative riskV. Determining Causality1. A