Week 2 - Research Methods in Psychopathology PDF

Week 2 Research Methods in Psychopathology PSY 2234 Fall 2024 Overview Approaches Nomothetic Idiographic Methods Case Study Correlational Experimental Quasi-Experimental Natural Experiments Analogue Experiments Longitudinal Studies Res...

Week 2 Research Methods in Psychopathology PSY 2234 Fall 2024 Overview Approaches Nomothetic Idiographic Methods Case Study Correlational Experimental Quasi-Experimental Natural Experiments Analogue Experiments Longitudinal Studies Research Approaches Research Approaches Nomothetic Idiographic Nom = name, category Idio = individual, one Examines what makes Examines what makes people alike people different Focus on patterns of Focus on pathology of psychopathology one client at a time Uses experiments, Uses case studies, epistemologies, longitudinal studies correlations Example: Depression Nomothetic Idiographic What makes rates of What factors in an depression higher among individual’s childhood put women than men? them at risk for depression? Are drug therapies most How can I help this client effective in combination change their negative self- with talk therapy? talk? Does depression look What ways can I make my different across different Latina clients feel more ethnic groups? comfortable with me? Nomothetic or Idiographic? A study on whether high-status individuals are more prone to personality disorders A study that follows an individual in rehab following a stroke A study on why one woman develops PTSD, but her identical twin does not A study that measures daily mood of children with a parent in prison Research Methods Validity Internal validity How well the study is tightly controlled and offers only one possible explanation External validity How well the study can be generalized, or how it informs us about different people Case Study Deep dive into one single individual’s life and their psychopathology Usually focuses on people who are unique for some reason Idiographic Case Study Advantages Allow for close examination of many different factors that account for psychopathology Ex: Genie’s case suggests importance of teaching language early Can develop new therapies Ex: Genie learned best through small frequent rewards Case Study Drawbacks Lack of internal validity Several possible explanations for findings Did Genie lack language due to genetic disabilities, or lack of nurturing? Lack of external validity Hard to draw conclusions about other people from studying just one Is Genie similar to other kids with language delays (e.g. Deaf children) or not? Correlation Examines co-relation, or amount of overlap, between two factors Strong correlation: huge overlap, factors closely tied Weak correlation: some overlap, factors loosely tied Correlation Examines co-relation, or amount of overlap, between two factors Positive correlation: as one factor increases, the other factor also increases Negative correlation: as one factor increases, the other factor decreases Correlation Advantages Better external validity If income-education link holds across 1000s of participants, then likely to continue to hold in other groups Ability to predict risk Depression severity correlates with suicide risk — suggests interventions should focus on most depressed individuals Correlation Disadvantages Lack of internal validity Correlation does not imply causation Examples? Experiment Manipulates one variable (the independent variable) to measure its impact on another (the dependent variable) Experiment Advantages Reasonable external validity If done right, should generalize across groups Bigger sample sizes improve external validity Excellent internal validity If done right, independent variable is only possible cause of dependent variable Experiment Advantages Reasonable external validity If done right, should generalize across groups Bigger sample sizes improve external validity Excellent internal validity If done right, independent variable is only possible cause of dependent variable Experiment Must be carefully controlled to make sure that indep. variable is only possible cause of dep. variable Otherwise, confounds make study impossible to interpret Confound: a variable that influences the dependent variable but isn’t the independent variable Experiment Disadvantages Confounds 50 participants with depression given SSRI (experimental group), and 50 a placebo (control group) Participants who aren’t getting better tend to drop out 30 drop out of the control group 20 drop out of the experimental group After 1 year, the remaining experimental group less depressed on average than the remaining control group… but why? Correlational Study Experimental Study Measures two Manipulates one variables, without variable, then manipulating either measures its effect on Cannot determine the other cause Can determine cause Excellent external So-so external validity validity Excellent internal Poor internal validity validity Experiment Disadvantages Ethics Cannot give people OCD in a study — would be morally wrong Instead, must use workarounds Animal research Analogue experiments Natural experiments Longitudinal studies Quasi-Experimental Studies Quasi-Experiment A study that can give us some causal information, but isn’t a true experiment Analogue experiments Natural experiments Longitudinal studies Analogue Experiment Seeks to imitate a real experience in a controlled lab setting, to observe its effects Examples Try to induce depression in rats Have participants use a driving simulator while drunk Provoke participants to see how anger influences heart rate Natural Experiment Takes advantage of an event that occurred naturally (in the real world) to see what is different before and after, or between matched pairs Examples Identical twin smoker and nonsmoker Measure of contagion fear in 2019 vs. 2021 Rates of depression in cities with and without lead pipes Longitudinal Study Looks at the same individual many times over a long period of time Examples Mortality among nuns study Seven-Up series Recidivism among parolees Quasi-Experiments and Validity Analogue experiment Excellent internal validity — tightly controlled lab setting Poor external validity — not directly studying behavior of interest Natural experiment Poor internal validity — many possible differences between conditions besides one variable Excellent external validity — studying real behavior as it occurs Quasi-Experiments and Validity Longitudinal study So-so internal validity — causes must always precede effects, but causes aren’t controlled So-so external validity — benefit from real behavior, drawback of small sample sizes