Week 4 Experimental Control Considerations PDF

**[Considerations for experimental designs: ]** Experimental control allows IV to influence DV and be the only possible causation. How? IV the only explanation: How? 1. Expectancy effect -- be avoided 2. Demand characteristics -- be avoided 1. Experimenter expectancy effect: Unintentio...

**[Considerations for experimental designs: ]** Experimental control allows IV to influence DV and be the only possible causation. How? IV the only explanation: How? 1. Expectancy effect -- be avoided 2. Demand characteristics -- be avoided 1. Experimenter expectancy effect: Unintentionally or intentionally manipulate experiment due to cognitive bias by experimenter. How? By knowing which condition participants are in. Result? Cause threat to internal validity. Solutions: How to avoid? - Double blind study - Automated instructions - Conditions run at the same time 2. Demand characteristics: Any parts of experiment that reveal the purpose of the study. Problem: may affect the results of DV due to lack of control in IV Result: cause threat to internal validity. Problems with participants: how they affect? - Being "good participants" - Being socially desirable - Mess up results - [Straightforward manipulation] = create a state of IV to manipulate DV (ex: create a quiz to measure influence of IV on DV) - [Staged manipulation] = indirectly create a state of IV to manipulate DV (ex: play a podcast while reading to influence comprehension and performance on quiz, make someone mad to measure anger) \*\*must debrief - [Filler task] = task that move the attention of participant away from or conceal real purpose of study - Easy - Cheap - May result in demand characteristics - Expensive - Hard to pull off and prepare - Hard to replicate - May avoid demand characteristics **[Operationalization considerations: Section 2]** Forms of measuring DV: 1. Self report 2. Behavioural response 3. Physiological response Ceiling and floor effect- influence DV results negatively [Ceiling effect:] High performance of ps on measurement which does not allow to see effect of IV on DV. [Floor effect:] Low performance of ps on measurement which does not allow to see effect of IV on DV. *The level of strength of IV and sensitivity of DV influence results.* *How?* [Strength of IV:] Strong IV = extreme causations (usually influence all population in the same way) Weak IV = detailed causations (able to see variety in results of the population) Strong IV used for new research and variety in DV. Weak IV used for gap in knowledge and less variation in DV. Sensitivity of DV: Sensitive DV = detailed effect (sensitive to small levels of causation) Tough DV = generalized effect (not sensitive to small levels of causation only catch big ones) Sensitive DV used for gap in knowledge and variety in DV. Tough DV used for new research and target a particular manipulation. \*\* two most important: - Gap in knowledge: weak IV and sensitive DV - New research: strong IV and tough DV *Asking questions in experiment is essential.* *Why?* - *Analyze trends* - *Correlational research* - *DV in experiment* - *Manipulation check in experiment* *It is essential what questions to ask and how to ask them.* *Why?* - *To know how ps respond* *Types of answers:* - *Beliefs and attitudes* - *Facts & demographics* - *Behavioural responses* [Self-report:] What is the form of asking questions (most common) 1. Basic information: facts, demographics, anything related to study 2. DV operationalization: what you are manipulating Wording questions: 1. Keep simple (avoid complicated words and definitions) 2. Avoid: - [Negative wording] (ex: do you not consider yourself not worthy? Vs. Do you consider yourself unworthy?) - [Double barrel] (ex: do you believe in nurture and nature?) - [Loaded questions] (ex: do you agree to put the miserable restaurant on close?) \*\* miserable adds bias and unnecessary wording [Question validity:] Ensure the questions are assessing what mean to be. [Other factors to consider for asking questions:] a. [Pre-test/ pre-existing questions: ] use of these materials allow use to measure question validity. b. [Response sets:] Ps who have a tendency to respond to questions in a particular way or bias. Solution? Put a question to ensure ps are reading the questions. c. [Social desirability:] Ps have tendency to be socially desirable and accepted therefore may not answer truthfully. Solution: keep answers anonymous. d. Open-ended vs. Close-ended questions: [Open-ended:] - More detail and rich answers - Different interpretation - Time consuming - Ambiguous - Common themes identify - Direct - Easy to interpret - Limited detail - Fast