Research Designs in Applied Social Psychology Flashcards
Chapter 4
described the most common
trade-offs researchers must consider as a ‘three-horned dilemma’.
Joseph McGrath
What consists of three-horned dilemma?
✧ PRECISION
✧ SITUATION GENERALIZABILITY
✧ PEOPLE GENERALIZABILITY
Good Researcher Must:
⌘ Know the unique advantages and disadvantages for each research design.
⌘ Conduct studies that make the best useof each design’s advantages while minimizing disadvantages.
⌘ Carefully construct research programmed using multiple designs that compensate for each method’s weaknesses.
FEATURES OF TRUE EXPERIMENTS: MAXIMIZING PRECISION
Manipulator and Random Assignment
occurs when the experimenter systematically varies the level of one (or more) variables while holding constant other variables that may have an effect.
Manipulator
- The manipulated variable (potential cause)
Independent Variable
- outcome variable (effect)
Dependent Variable
occurs when the experimenter assigns participants to groups on an arbitrary basis. The use of random assignment therefore ensures that every participant has an equal chance of being assigned to any condition in the experiment.
Random Assignment
is an additional variable that systematically varies with the independent variable and influences the dependent variable.
Confound
- occurs when the presence of an individual difference influences the outcome variable.
Person Confound
- occurs when the experimenter unwittingly varies two or more variables at once
Procedural Confound
refers to variables that influence the dependent variable
Noise
indicate that the effect that one or more independent variables has on the dependent variable depends on an additional independent variable.
Interactions
has become the most common laboratory method for studying social behavior.
2X2 Factorial Design
Statistical interactions are also useful because they identify the of a theory: conditions that demarcate when a theory does or does not apply. Boundary conditions are important to study because no theory that is of applied use will be correct under all circumstances (see McGuire, 1973)
boundary conditions
is that research laboratories are often situated on university campuses, and it is very easy and cost effective to utilize this resource.
Incidental Reason
Using homogenous samples reduces noise.
Purposeful Reason
What are the disadvantages of true experiments, maximizing precision?
Incidental Reason, Purposeful Reason
What are the advantages of true experiments, maximizing precision?
Confound, Noise, Interactions, 2X2 Factorial Design, Boundary Conditions
Difficulty with these approaches, however, is that researchers can rarely, if ever, find the one ‘perfect’ situation that captures the range of situations of interest or the ideal complement of samples that will define the degree to which an effect generalizes to different people. Difficulty with these approaches, however, is that researchers can rarely, if ever, find the one ‘perfect’ situation that captures the range of situations of interest or the ideal complement of samples that will define the degree to which an effect generalizes to different people
TRUE EXPERIMENTS: MAXIMIZING PRECISION MINIMIZING DISADVANTAGE
: there are multiple ways of interpreting the realism of a laboratory experiment.
Aronson and Carlsmith (1968)
● The physical setting of the study is like the real-world setting in which the phenomenon of interest typically occurs.
● style that focuses on the ordinary, everyday aspects of life. It avoids grand narratives of heroic deeds, instead delving into the minutiae of daily existence.
Mundane Realism
● by asking if it ‘feels’ realistic on psychological dimensions.
● style that delves into the inner thoughts, feelings, and motivations of characters. It aims to provide a deep understanding of their psychology, often exploring their subconscious minds and the complexities of their personalities
Psychological Reaction
investigates the relationship between different measured variables, typically with an interest in determining how these variables interrelate in naturally occurring situations.
Correlational Research
What are the features of Correlational Research?
FEATURES
⌘ Researchers must generate a fixed set of observations about a group of people.
⌘ Correlational research takes a more passive approach to studying variability.
Behaviors of interest can be measured within natural settings, thereby increasing the chances that the research findings are representative of what occurs naturally in the ‘real world’.
Advantages
ranges from–1 to +1. The absolute numerical value indicates the magnitude of the relation, with higher numbers showing a stronger connection between the variables. The sign indicates the direction of the relationship.
Correlation Coefficient
indicates that as the value of one variable increases, so does the other variable.
Positive Correlation
indicates that as the value of one variable increases, the other variable decreases.
Negative Correlation
In order to increase generalizability to situations
Correlational research forfeits measurement
precision.
✧ Inability to infer causality
✧ Reverse causality
✧ Third variable problem
Disadvantages of Correlational Research
an independent variable that can influence the outcome of a given statistical trial, but which is not of direct interest.
Covariates
used in correlational studies whenever reverse causality offers a plausible alternative explanation.
Longitudinal Designs
research method that resembles a true
experiment but lacks one or more of the
essential
elements,
such as random
assignment of participants to experimental
and control groups.
Quasi-experimental Design
In this design, one of the independent variables is manipulated (treatment variable) and one is measured (person variable).
Person-by-treatment quasi-experiment
(by using laboratory methods) are more likely to suffer from artificiality.
Closely resemble true experiments
(by going into the field) are more likely to suffer from the influence of confounds and noise.
Resemble correlational studies
- selects participants
from the larger population in a random fashion
Random Sampling
- population list of locations rather than people
Cluster Sampling
- discrepancy between the results obtained in a representative sample and the results one would have obtained if everyone in the population of interest had been studied.
Sampling Error
Formula of SE
Z*(σ/√n)
is the z-score corresponding to the desired level of confidence
Z
is the population standard deviation.
σ
is the sample size.
n
The best way to minimize the disadvantages of survey research is to:
pair a sampling procedure with either a true experimental or a correlational design.