Introduction to Research Methods in Psychology

Goals of a science

1. DESCRIBE 2. PREDICT 3. EXPLAIN

Systematic empiricism

Use of controlled, systematic observations

Empirical questions

Ask questions to which the answers are observable (ex: falsifiable questions)

Generates public knowledge

Accumulate knowledge and allows science to be self-correcting

Basic Research

Achieving a more detailed and accurate understanding of a phenomenon

Applied Research

Direct application to real-world problems

Operational definitions

A definition of a variable in terms of precisely how it is to be measured

Self-report measures

Those in which participants report their own thoughts, feelings, and actions

Behavioral measures

Those in which some other aspect of a participant's behavior is observed and recorded

Physiological measurements

Those that involve recording any of a wide variety of physiological processes, including heart rate and blood pressure

Converging operations

Good operational measures often include some combination of measures

Levels of measurement

Nominal = categorical, Ordinal = numbers represent a rank order, Interval = numbers spaced so that intervals have the same interpretation throughout, Ratio = numbers have a true zero point that represents the complete absence of the quantity

Reliability

The consistency of a measure

Test-retest reliability

A personality or memory capacity; if tested again, the scores should positively correlate with the first time

Internal consistency

Quality of measurement; consistency of people's responses across the items on a multiple-item measure

Inter-rater reliability

Extent to which different observers are consistent in their judgements

Validity

Extent to which scores from a measure represent the variable they intend to

Face validity

Extent a measure appears 'on its face' to measure the construct of interest

Content validity

Extent a measure 'covers' the construct of interest

Criterion validity

Extent to which people's scores on a measure are correlated with other variables (known as criteria) that one would expect them to be correlated with

Discriminant validity

Extent scores on a measure are not correlated with measures of variables that are conceptually distinct.

Hawthorne effect

Observer bias where individuals modify their behavior in response to being observed.

Expectancy bias

A phenomenon where a researcher's expectations influence the outcome of a study.

Clever Hans

A horse that appeared to do math but was actually following the researcher's movements without the researcher realizing.

Unobtrusive measurements

Strategies to minimize bias problems by employing methods that do not interfere with the subjects.

Blind procedures

Experimental methods where participants are unaware of the treatment they receive to reduce bias.

Double blind procedures

Experimental methods where both the participants and the experimenters are unaware of the treatment assignments.

Deception in research

Using misleading information in research if justified, to prevent bias.

Subject anonymity

Ensuring that participants' identities are kept confidential to reduce bias.

Non-threatening setting

Creating a low-key environment for participants to minimize anxiety and bias.

Experiment

A study that includes at least two variables: a dependent variable (DV) that is measured and an independent variable (IV) that is manipulated.

Dependent variable (DV)

The variable that is measured in an experiment.

Independent variable (IV)

The variable that is manipulated by the experimenter.

Levels of IV

The IV must have 2 or more levels for a valid experiment.

Critical experiments

Experiments that pit two theories against one another to determine which is valid.

Popperian logic of falsification

The principle that any scientific theory should be potentially falsifiable.

What-if experiment

An experiment conducted without a compelling theory, primarily to observe outcomes.

Replication

Repeating an already published experiment to confirm original results.

IV manipulation strength

The IV manipulation must be strong enough to produce the desired effects.

Floor effect

A situation where all participants score at the low end of a measurement scale.

Ceiling effect

A situation where all participants score at the high end of a measurement scale.

Control variables

All variables that the researcher keeps constant in an experiment.

Multiple Independent Variables

A common practice in psychological sciences that allows for better control and generalization of effects.

2x2 factorial design

An experimental design that evaluates the effects of two independent variables, each with two levels.

Atypical antipsychotic medications

Medications used to treat schizophrenia that are different from traditional antipsychotics.

Interaction effects

When the effects produced by one independent variable (IV) are different at the different levels of another IV.

Main effects

Changes in a dependent variable across the levels of a single independent variable, averaged across all other independent variables.

Multiple Dependent Variables

Having more than one way to measure outcomes, which improves generalizability.

Experimental designs

Research designs that test the effects of independent variables on dependent variables.

Between subjects designs

Experimental design where different groups receive different treatments, ensuring no treatment outcomes affect each other.

Within-subjects designs

Experimental design where all participants are exposed to all levels of the independent variable.

Order effects

Changes in participants' responses due to the order in which treatments are administered.

Practice effects

Improvements in participants' performance due to repeated exposure to the same task.

Fatigue effects

Declines in performance due to participants becoming tired over the course of an experiment.

Context effects

Differential carryover effects where the context of one treatment influences the response to another treatment.

Mixed designs

Research designs that incorporate both between-subject and within-subject variables.

Counterbalancing

A technique used to control for order effects by varying the order of treatments across participants.

Memory enhancement

The improvement of memory performance, which can be tested through different treatments.

Group Equivalence

Ensuring that two groups are comparable in terms of certain baseline measures.

Matching

Pair participants in the two groups on some baseline measure (ex: IQ, age, memory score).

Randomization

A technique where everybody has an equal chance of being assigned to group A or B.

Demand Characteristics

When the participant's expectations about the purpose of the experiment or what is expected of them influences their behavior.

Hawthorne Effect

Named after an electric plant that conducted experiments on worker productivity; refers to changes in behavior due to awareness of being observed.

Blind Participants

Making participants unaware of the conditions of an experiment to reduce bias.

Experimenter Effects

When the experimenter's expectations inadvertently influence the outcome of the experiment.

Double-Blind Experimental Design

A design where condition/group assignment is unknown to both the participant and the experimenter.

Quasi-Experimental Designs

Research designs where the independent variable cannot be manipulated or randomly assigned.

Nonexperimental Research

Research that lacks the manipulation of an independent variable.

True Experiment

Used when a researcher has a specific research question about a causal relationship between two variables and can manipulate the independent variable.

Quasi-Experimental Research

Research where the independent variable cannot be manipulated or participants cannot be randomly assigned to conditions.

Single Variable Research

Research questions or hypotheses that focus on a single variable rather than a statistical relationship between two variables. (milgram obedience study)

Milgram Obedience Study

An example of nonexperimental research as it observes one variable (extent to which participants obeyed) under the same conditions.

Correlational Research

Research questions about a non-causal statistical relationship between two variables.

Verbal Intelligence and Mathematical Intelligence

An example of a research question exploring the correlation between two types of intelligence.

Marijuana Use and Schizophrenia

An example of a research question investigating the relationship between substance use and mental health.

Adverse Childhood Experiences

Research question examining the relationship between the number of adverse childhood experiences and adolescent conduct problems.

Correlational Research Pitfalls

Linearity, directionality, third-variable problem, spurious correlations.

Prediction

When the researcher wants to explain how variable X predicts (non-experimental 'causes') variable Y.

Experimental Research

Research where the dependent variable (DV) is continuous (e.g., age, depression level, reaction time) and the independent variable (IV) has 2-6 levels (e.g., diagnosis, gender, experimental condition).

Non-Experimental Research

Research that investigates the relationship(s) between two or more continuous variables (DVs) using correlation and linear regression.

Relationships between Ordinal Variables

Investigates relationships such as race/ethnicity, parenting styles, and employment status.

Impact of Parenting Styles

The effect of parenting styles on the presence/absence of child behavior problems.

Chi-Square Test

A test for goodness of fit between observed and expected values.

Milgram Experiment

An experiment where participants were told to deliver shocks to another person.

Correlation vs Regression - Similarities

Both qualify the direction and strength of the relationship between two numeric variables.

Correlation vs Regression - Differences

Regression establishes how X causes Y to change, while correlation treats X and Y as interchangeable.

Correlation Advantage

Provides a more concise (single value) summary of the relationship between two variables.

Regression Advantage

Offers a more detailed analysis of relationships used for prediction, can account for confounding variables, and analyze interactions.

Internal Validity

The extent the design of a study supports the conclusion that changes in the IV caused any observed differences in the DV.

External Validity

The validity of applying the conclusions of a scientific study outside the context of the study.

Generalizability

The extent the results can be generalized to and across other situations, people, stimuli, and times (REAL WORLD APPLICATION).

Non-Experimental Research Validity

Non-experimental research tends to have higher validity.

Median

middle number (50 percentile)

Mode

most frequently occurring number

Range

Xmax - Xmin

Standard deviation

average distance between individual observations in a dataset and the mean of the dataset

Histogram

frequency (count) as a function of observed values

Percentile rank

# of data points less than current value, converted to a percentage

Bar chart

a graphical representation of data using bars

Line chart

a graphical representation of data points connected by lines