Exam 2 - Chapter 9 (Conducting Experiments) Flashcards
Manipulating the Independent Variable
To manipulate an IV you have to:
- create an operational definition.
- Turn a conceptual variable into a set of operations that can be presented to the participants.
**The manipulation happens when the researcher changes the conditions exposed to the participant. **
Strength of the IV Manipulation
Manipulation of the IV should be strong enough to create an effect (if one exists), but not too strong that it hurts the study.
- The amount of potential impact of the IV on the DV.
Strong Manipulations = Important in early stages of research in order to demonstrate that a relationship does exist
Example:
- Weak Manipulation:
- Hot – Cold
- Hot – C_ld
Stronger Manipulation: - Hot – Cold
- Hot – ____
Manipulations can be TOO Strong
Ecological and External Validity → when a manipulation is unrealistic to the real-world, it is too strong. (ie: inducing anxiety via monster chase = unrealistic)
Ethics → Using extreme scenarios to manipulate IV can be unethical (ie: Stanford Prison)
Curvilinear relationships → we might not see the relationship when using too strong of a manipulation (ie: Performance – Arousal)
Types of Manipulations:
Straightforward Studies vs. Staged Studies
Straightforward Studies: Easy situations are used to manipulated the IV by presenting written, verbal, or visual materials to the participants.
Staged Studies: Complex situations are used to manipulate the IV, often using simulations of real-life.
- Used for One of Two Reasons:
- To create a psychological state (frustration, anger, affect self-esteem, etc)
- To simulate a real-world situation
Manipulation Checks
A measure used to determine if the manipulation of the IV had its intended effect on the participant.
- Includes another measure (that IS NOT the Dependent Variable) to see if the manipulation worked. (ie: self-report)
- Provide evidence for the Construct Validity of the manipulation
Risk: manipulation checks might create Demand Characteristics
- solutions: disguise the check OR do the check after measuring the DV.
- BUT: Post DV measurement risks that IV effect may have weakened. – Instead do it in a Pilot Study
Pilot Study
A small-scale study done before the main experiment to test the procedures (aka a test-run).
- A Pilot Study can be done instead of a Manipulation Check to Demand Characteristics in the main study
Demand Characteristics
Anything about the experiment that unintentionally indicates to participants how they should act, making people change their behavior to conform to what they think they should do.
- Could be smth that reveals the true purpose of the study (ie: Title of study, Instructions)
- Could be smth that leads people to believe an inaccurate purpose of the study
Expectancy effects (Experimenter Bias)
The intentional or unintentional impact that an experimenter’s expectations about the outcome of the study has on participants.
Can happen through:
- Instructions (wording),
- behavior (friendly/cold behavior),
- Informing participants of the purpose of the study
Controlling for Experimenter Bias:
- Double-blind study (participant + experimenter are unaware of the group the participant is in)
- Automated procedures/instructions
- Multiple experimenters
Types of Dependent Variable Measures
- Behavioral Measures
- recall, recognition, cued recall, etc.
- Self-Report Measures
- give real-time self-reported measures instead of asking for predictions
- Physiological Measures
- hard to fake, but take more work/background research
Test Example Question:
What features of an experiment handle this issue: Expectancy effects / experimenter bias
Answer: Double-Blind Study, Automated procedures/instructions, Multiple experimenters, and more.
Test Example Question
What features of an experiment handle this issue: Demand characteristics
Answer: Use of Deception (make participants think you are studying smth you are not), ambiguous (unobtrusive) measures, disguising the measure (usually used for manipulation checks)
Test Example Question:
What features of an experiment handle this issue: Confounding variables
Answer: Random assignment, matching or control groups, covariates
Test Example Question:
What features of an experiment handle this issue: Sensitivity of the DV
Answer: Pilot Testing the DV (helps ensure sensitivity is just right), using Multiple Measures,
Test Example Question:
What features of an experiment handle this issue: Strength of the manipulation
Answer: Manipulation checks, Pilot Studies that check the strength of DV measure,