Chapter 6 - Methodological Control in Experimental Research Flashcards
(25 cards)
Between-Subjects Design
Contrasts two conditions of an IV (A and B) between two groups of people. Participants are never in both groups.
Random Assignment
Every volunteer has an = chance of being placed into any of the experimental conditions. Most effective with large samples.
Matching
Procedure used to ensure a subject variable is evenly distributed across groups. An example of procedural control.
Within-Subjects Design
Contrasts two conditions of the IV (A and B) within the same group of people. Participants are always in both groups.
Order Effects
When completing the first condition alters one’s performance in other conditions.
Complete Counterbalancing
Way to control for order effects when testing each condition once. All possible sequences of conditions are used.
Partial Counterbalancing
Way to control for order effects when testing each condition once. Some possible sequences of conditions are used.
Ex. Randomly sampling from all possible orders or randomly assign participants to an order using a balanced Latin Square.
Block Randomization
Way to control for order effects when testing each condition more than once. Presents the conditions in random order within a block (one time) and this procedure is repeated a number of times.
Reverse Counterbalancing
Way to control for order effects when testing each condition more than once. Presents the conditions in one order and then again in the opposite order (for however many # of repeats.
Ex.
Group 1 (forward-first): A-B-C - C-B-A
Group 2 (reverse-first): C-B-A - A-B-C
Experimenter Bias
When the researcher does something that leads participants to behave in certain ways.
Protocals
Way to control for experimenter bias. Have detailed descriptions of exactly how the experimenters are to interact w/ participants during each session.
Blinding
Way to control for experimenter bias. Where experimenter and/or subject doesn’t know which condition they’re in.
–> Single-blind studies: participants don’t know which condition they’re in but the experimenter does.
–> Double-blind studies: neither the participants or researcher (usually a research assistant) know which condition the participants are in.
Primary Investigator (PI)
Usually does group assignments.
Participant Bias
Can occur when participants have expectations and beliefs about their role in the study (ex. hawthorne effect, good subject effect, evaluation apprehension/social desirability bias).
Hawthorne Effect
Type of participant bias. When a researcher obtains unrealistic results simply because the research participants are behaving in an abnormal way because they’re being studied.
Evaluation Apprehension (aka Social Desirability Bias)
Type of participant bias. When participants respond different because they wish to be evaluated positively.
Good Subject Bias
Type of participant bias. Occurs when participants persevere/persist through repetitive and boring tasks that they would normally stop.
Reducing Demand Characteristics
Way to control for participant bias. These are the aspects of the study that might reveal study hypothesis and potentially persuade participants to act differently.
Manipulation Check
Way to control for participant bias. Asking participants what they think the hypothesis is before debriefing.
Cross-Sectional Designs
Measures all participants once (between-subjects).
Ex. different ages = different people
Longitudinal Designs
Measures the same participants multiple times (within-subjects).
Ex. different ages = same people
Attrition
When people leave the study. Common in longitudinal studies. The group completing a study may differ from the group that started it.
Cohort
Group of people born about the same time that share a common cultural experience (shares age as well as historical culture/environment).
Cohort Effects
When a cross-sectional study is unable to disentangle the effects of age vs culture/environment.
Ex. people born in the 1940s have a hard time adopting new technology, why?
