Topic 1 Control Flashcards
Lab setting: Advantages and disadvantages of extraneous variables
a. Control of extraneous variables
i. Non-experimental studies have more external validity than lab experiments
ii. Field research is warranted when ethical or practical problems preclude the degree of control need for experimentation
b. Disadvantage in terms of eternal/ecological validity
Research setting as a preparation
a. Content for data collection
i. Preparation- environment of entity constructed for the purpose of data collection, determines whether a relationship between variables of interest can be measured
1. ***RT methodology is simple and reliable
b. Which situation will provide best relationship between IV and DV
Purpose of instrumentation of the response
a. Validity and reliability of measurements and instruments
i. Increasing the sensitivity of the research is to improve the measurement
ii. Characteristics: takes response out of the realm of casual observation and makes it reliable
1. Standards of reliable measurements: MMPI, Beck Depression Inventory
What are the effects of holding a constant?
- *remains across all conditions**
a. Extraneous variable can’t be eliminated, might be held constant throughout the experiment
i. Present for all participants, or ex: time of day if it’s a factor
Importance of balancing
- *equally distributing the uncontrollable variation across all conditions**
a. Most important that can’t constant in psychological research is participant variables
i. Ex: gender, education, SES, age
b. **Counterbalancing- participant serves two or more different experimental conditions
Subject as own control: advantages and disadvantages
What are the advantages and disadvantages of having a subject as control?
i. advantages
1. 1 of the most powerful
2. Reduces variation caused by individual differences
3. Provides more statistical power
ii. Disadvantages
1. Not always possible (when there are possible carryovers or sequence effects)
Importance of random assignment (between subject)
a. Random selection- every person in the population must have equal chance of being selected
b. Random assignment- determine who gets what condition
i. Eliminates statistical bias
ii. Also referred as random ordering of variables, random ordering of treatments
iii. Assures that sample represents population (generalization- internal validity)
iv. Stat analysis provides means of testing if results are from chance alone.
1. If subjects aren’t randomly chosen statistical test are invalid
Matching (group trait)
a. Experimental and control groups are equated on one or more variables
b. Justification for matching
i. Important variables on which subjects differ controlled by matching
ii. Easy method to Pretest subjects before assigning conditions
Building in nuisance Variables
a. Variables which can’t be easily controlled are made IV’s as means of control
b. Different from confounding variable- 2nd IV that independent from 1st one
i. Additional IV makes a factorial design where interaction occurs
Statistical control
a. mathematical way of seeing how subjects differ from nuisance variable correlated with the DV
i. Analysis of Covariance
b. Helpful when random assignment and matching are not possible
Direct Replication
a. Direct replication- same exact experiment. Only done when systematic replication has failed
Systematic Replication
b. Systematic replication- different experiment where certain results should be found if original is valid
i. Test external validity- different subject, species, situations
ii. test construct validity- different ways of measuring the theoretical concepts
iii. test stat validity- in all replication, both direct and systematic
Importance of counterbalancing?
b. **Counterbalancing- participant serves two or more different experimental conditions
i. Ex: whether stop sign should be red or yellow
ii. General principle
1. Repeated treatment experiments, each condition must be presented to each participant the same amount of times and same number of times
2. Each condition must proceed and follow all other conditions on equal number of times
