Research Methods in Psychology PDF

Research Methods in Psychology Part 1 … that adapts as new data develop. Image Source: Wikimedia Commons  Role of Theory Scientific theories typically explain the relationship between two or more variables. Th e o r y Explanatory & Predictive/Generat...

Research Methods in Psychology Part 1 … that adapts as new data develop. Image Source: Wikimedia Commons  Role of Theory Scientific theories typically explain the relationship between two or more variables. Th e o r y Explanatory & Predictive/Generative  Scientific theories (and the hypotheses they generate) must also be: Testable using currently available research techniques Scientific Falsifiable à it must be Theories possible in principle to make an observation that would show the proposition to be false, even if that observation has not been made Parsimonious à preference for simplicity  Intergroup Contact Theory (Pettigrew, 1998) – Under certain circumstances, positive Examples of intergroup contact can reduce prejudice Psychological toward the outgroup Theories Social Comparison Theory (Festinger, 1954) – People will evaluate their own abilities by comparing themselves to similar others, especially when more objective measures are unavailable Social Learning Theory (Bandura, 1977) – People can learn by observing others, in the absence of explicit behavioural reproduction or reinforcement  Pause & Practice: Hypothesis Generation Each of these theories can be used to generate testable hypotheses – On your own, try to come up with one specific research hypothesis that is based on one of these theories Sample hypothesis: Students who join a study group that includes students from a different racial group will show reduced prejudice toward that racial outgroup compared with students who join a study group that only includes people from their same racial group.  Variable: A characteristic or condition that changes or has different values for different individuals Independent variable: A variable that is manipulated, in order to see its impact on the dependent variable Variables Dependent variable: A variable that is measured, in order to see how it is affected by the independent variable  Conceptual vs. Operational Definitions of Variables Conceptual definition: Akin to a dictionary or textbook definition; the meaning of the term Operational definitions: Definitions of theoretical constructs that are stated in term of concrete, observable procedures Something that can be measured E.g., “Show reduced prejudice” à How do we measure prejudice?  Operational Definitions Sometimes variables are well-defined and easily measured or manipulated  Operational Definitions Some variables are not well-defined and cannot be directly observed Constructs: Internal attributes or characteristics that cannot be directly observed but are useful for describing and explaining behaviour  Choose a construct/variable from the list below (or pick Pause & your own!). Write out a Practice: conceptual definition for the construct. Then try to think of Operational at least three different vs. operational definitions for Conceptual this construct. Definitions Prejudice Anxiety Aggression Intelligence  Operational Definitions Physiological measure Anxiety Behavioural measure Self-reported measure Research Methods Descriptive Methods  Descriptive methods Often concerned with a single Three Big variable of interest Categories of Research Correlational methods Methods Examine associations between two or more variables Experimental methods Examine cause-and-effect relationships between two or more variables  Descriptive Methods Involve the systematic observation and classification behavior Includes: Surveys Focus groups Case studies Observational research  E.g., Survey of Study Strategies N = 177 Survey data from Karpicke, Butler, & Roediger (2009)  Case Studies https://www.utoronto.ca/news/u-t-s-top-undergrad-credits- 40-gpa-his-decision-learn-about-how-i-was-learning https://www.utoronto.ca/news/allie-sinclair-soars-through-university-u-t-s-top-student  Types of Observation Naturalistic observation: Passive observation. Observers do not change or alter ongoing behaviour (at least not intentionally) Participant observation: Active observation. The researcher is actively involved in the situation. Laboratory observation: Systematic observations are made within a laboratory setting (rather than in the ‘real world’). Example: Couples Interaction Scoring System (Gottman, 1979)  Strengths of Descriptive Approaches Case studies and observational research can provide important insights and stimulate further research to test specific hypotheses Surveys allow us to gather large amounts of information quickly and easily Focus groups and interviews can provide rich, detailed information that may be lacking from a survey Resources: https://www.utm.utoronto.ca/iits/information-security-risk-management/resources/cyberbullying (Mishna et al., 2018)  Potential Problems with Descriptive Methods Reactivity (e.g., the Hawthorne effect)  Potential Problems with Descriptive Methods Reactivity (e.g., the Hawthorne effect) – Demand characteristics Observer/experimenter bias Self-report bias – Social desirability bias – The “better-than-average” effect Image from https://medium.com/usertribe/interview-how-we-avoid-bias-in-customer-research-3d9926e51bf  Summary In psychology, these methods are often used in combination with other methodological approaches May lead to claims about the frequency or prevalence of a behaviour May add rich, qualitative information to a research program that would otherwise be missing this type of detail Research Methods Correlational Methods  Descriptive methods Three Big Often concerned with a single Categories variable of interest of Research Correlational methods Methods Examine associations between two or more variables Experimental methods Examine cause-and-effect relationships between two or more variables  Correlational Methods Involve examining how variables are related (without manipulating any of the variables) Single group of participants, at least two measures (variables of interest) Allow researchers to make claims about associations between variables, but not causal claims Example: Laptop Multitasking School-unrelated laptop use during class time has been associated with lower academic satisfaction, lower end-of-semester GPAs, and lower course performance relative to classmates (Gaudreau et al., 2014) Pause & Practice: - What do these findings tell us? - What do they not tell us?  Correlational Methods Correlational studies tell us about relationships between variables – No relationship (0) – Positive relationship (+) à variables move in the same direction – Negative relationship (-) à variables move in the inverse direction – How strong is the relationship? Example: Laptop Multitasking (Gaudreau et al., 2014)  Correlational Methods CORRELATION ≠ CAUSATION Correlational studies do not tell us whether one variable causes changes in another variable – Why not? Directionality problem A B Third-variable problem C A B Why Bother with Correlational Methods? We can’t always manipulate a variable we are interested in (due to feasibility or ethical concerns) E.g., Cyberbullying research in textbook Why Bother with Correlational Methods? , 2017: Ravizza et al.  Summary Correlational studies are an important component of psychological research as they allow us to examine hypotheses about the relationships between variables However, they do not allow us to make cause-and-effect claims, as tempting as they may be! Research Methods Experimental Methods  Descriptive methods Three Big Often concerned with a single Categories variable of interest of Research Correlational methods Methods Examine associations between two or more variables Experimental methods Examine cause-and-effect relationships between two or more variables  Variable: A characteristic or condition that changes or has different values for different individuals Independent variable: A variable that is manipulated, in order to see its impact on the dependent variable Variables Dependent variable: A variable that is measured, in order to see how it is affected by the independent variable  Pause & Practice: Variables What are the two variables in this study? Which would be the independent variable and which would be the dependent variable? Bonus practice: How might you operationalize these variables?  Pause & Practice: Variables 24 hours Borota et al., 2014  Experiments Involve manipulating an independent variable in order to determine its impact on a dependent variable (which we measure) Are tightly controlled (typically take place in the laboratory) Participants are randomly assigned to study conditions  Experiments: Example Example: Karpicke & Bauernschmidt (2011) Independent variable: The type of studying the participants engaged in 1. Study once 2. Recall once 3. Repeated massed 4. Repeated spaced Dependent variable: Tested one week later Results from Karpicke & Bauernschmidt (2011)  Experiments: The Importance of Control Confound: Anything that may unintentionally vary along with the independent variable – Is there anything else that might be different between experimental conditions? – Confounds limit our ability to make causal claims  Experiments Random assignment: Each participant has an equal chance of being assigned to any condition in the study – Necessary component of an experiment, because this ensures that your different groups are equivalent on average Don’t get these terms confused! Random sample: Each member of the population you are interested in has an equal chance of being chosen to participate  Importance of Random Assignment Random assignment ensures the groups are equivalent, on average, on those variables you might be concerned with (e.g., self-esteem, intelligence), as well as those variables you haven’t even thought of!  Random Assignment: Video Explanation Video Link: https://www.youtube.com/watch?v=V_GIjFw6RZE&t=6s  Why are “Double-Blind” Experiments Ideal? Recall: Observer/experimenter bias & demand characteristics In a double-blind experiment, both the participants and the experimenters who interact with them are unaware of which condition the participant is in  Participants: Samples & Populations Population: The group that you want to be able to generalize your findings to Sample: The group of individuals from this population who are part of your study Random samples vs. convenience samples Psychology’s WEIRD Problem Western, Educated, Industrialised, Rich and Democratic To learn more (listen or read transcript): https://www.abc.net.au/radiona tional/programs/allinthemind/w eird-psychology/12766212  Quasi-Experiments Experimental design where random assignment is not possible - E.g., researcher takes advantage of pre-existing groups Risk of potential confounds limits the claims that a researcher can make – But they can be very useful for studying variables where random assignment isn’t feasible or ethical!  Field Experiments Experiments that occur in real-world settings (the “field”) rather than the laboratory Random assignment is possible; however, researchers have less control over the study Participants are often unaware of the study Example: See “Psychology Takes on Real World Problems” in Chapter 2  Summary Experiments involve the manipulation of an independent variable in order to measure its effect on a dependent variable Random assignment and other strategies are used to avoid potential confounds which threaten the internal validity of our experiment  Thinking Critically About Research Claims What was actually measured (or manipulated) (i.e., construct validity) à The degree to which the variables in a study truly represent the abstract, hypothetical variables (i.e., constructs) in which the researcher is interested Who were the participants/respondents? For causal claims à are they really justified? Research Methods Thinking Critically About Research  Types of Descriptive methods Research May lead to claims regarding the frequency of some Methods behaviour & Correlational methods Subsequent May lead to claims regarding Claims the association between two variables Experimental methods May lead to claims regarding the causal relationship between two variables Validity vs. Reliability  Interrater reliability Types of Reliability Test-retest reliability  Types of Validity Construct validity Operationalizations External validity Generalizability Internal validity Causality Each focuses on a different way of searching for plausible alternatives  Construct Validity How valid are the measures used in the study? – i.e., How accurately (or appropriately) have the variables been operationalized? Are there alternative constructs the measure may be assessing? What are the limitations of the measure (or operationalization) that was used?  Physiological measure Anxiety Behavioural measure Self-reported measure E.g., Is the measure capturing anxiety or something else, such as sadness or anger? Is it measuring acute or chronic or clinical anxiety? Etc.  External Validity How well would we expect the results of the study to generalize to people and contexts besides those in the study itself – i.e., what factors other than the ones measured/manipulated here might also be necessary for the claim to apply? Who were the participants in this study? Where and when was it conducted? Would we expect these results to generalize to other individuals, situations, and time periods? Pause & Practice: In a study on anxiety and job performance, what factors might affect the generalizability of the results? - How long someone has been in the job - Type of job - Job security - Sociodemographic variables - Etc.  Internal Validity How well has the study established a cause-and-effect relationship between variables – i.e., Are there confounds in the experiment? Are there plausible alternative explanations for the observed differences between the groups?  Pause & Practice: Thinking Critically About Claims For each of the (real) headlines below, think about: a. What type of claim is being made (descriptive, associative, or causal)? b. What questions would we ask (or what additional information would we want to know) in order to assess the validity of the claim? “Canadians level of trust in “Walking reduces risk of breast cancer” charities dropping: Study” “Mediterranean diet tied to better health late in life” Research Methods Intro to Statistics & Open Science  Types of Statistics Descriptive statistics Organize data into meaningful patterns and summaries – They describe the data – Percentages, counts, averages, correlations, etc.  Types of Statistics Inferential statistics Allow us to extend conclusions from a sample to a population – They allow us to make inferences based on data E.g., “Based on these data, we predict that between 35% and 55% of college student use practice problems to study” – These inferences are always probabilistic Hypotheses are never “proven”! Replication & Reproducibility Reproducibility: A study can be duplicated in method and/or analysis Replicability: A study about a phenomenon produces similar results from a previous study of the same phenomenon – Close/Exact Replications – Conceptual Replications Video Link: https://www.youtube.com/watch?v=j7K3s_vi_1Y Is There a “Reproducibility Crisis” in Science? A Cultural Shift in Psychological Research What does it mean when a study “fails to replicate?”