Week 2 Research Design: Basic Principles Flashcards
What is the purpose of research design?
To provide answers to research questions in the most accurate, valid, objective and economical way possible. To make inferences to the population
What ways can you control variance?
Maximize systematic variance, Control extraneous systematic variance and minimize error variance
How to maximize systematic variance?
Research be designed and conducted so the experimental conditions (IV & DV) are as different as possible.
Control extraneous variables.
Define systematic variance and unsystematic variance?
Systematic variance: treatment effects
Unsystematic variance: error variance
What are the four ways to control extraneous variables?
- Eliminate the variable (cannot generalize)
- Randomization (only method to control all extraneous variables, experimental groups statistically equal in all ways)
- Build into design
- Match participants (good for repeated measures design)
How to minimize error variance?
Control conditions and use reliable measures
What is error variance?
The variability of measures due to random fluctuations (individual differences & errors of measurement)
Name the non-interpretable designs.
- Single-group, post test only design
- One-group, pretest-post test design
- Post test-only design with non-equivalent groups
These can help to generate hypotheses
What is a single-group post test only design?
One group with a treatment then observation
G1: X O
No pretest obs
no control group to compare
What is the one-group pretest-post test design?
Pretest obs, treatment, post test obs
G1 O1 X O2
does not control for confounding variables
What are confounding variables?
Extraneous variable that interferes with IV and DV:
- history
- maturation
- measurement
- statistical regression
- selection
- instrumentation
- mortality or attribution
What is statistical regression?
Lowering of extremely high or raising of extremely low scores due to mean (retesting situation)
What is the post test only design with non-equivalent groups?
Two groups one w/ treatment one w/o then obs
G1 X O
G2 O
no randomization, therefore no equivalence before treatment
What are the criteria of a research design?
Answer research Q?
Control extraneous variables?
What is external validity?
Extent to which the results can be generalized to other subjects, conditions, times, places (population)
Is of more interest in applied research than basic research
What is internal validity?
Extent to which researcher can accurately state that the IV produced an observed effect.
When all effect (as measured by DV) is due to the IV, then internal validity has been achieved.
Must control confounding variables to achieve.
What are the characteristics of experiments?
- manipulation of one or more factors (IV/s) to measure effects (DV) the manipulation has on behaviour
- internal validity
- reliability (replicatable experiment and results)
- sensitivity (power)
- external validity
Research design steps?
- Problem statement
- theoretical constructs
- IV and DV
- Hypotheses and stat hyp
- population
- # of subjects required
- define procedure
- anticipate threats to validity
- create controls
- determination of stat analysis to be performed
What are subject artifacts?
- Demand characteristics
Altruism, obedience and evaluation apprehension
What are experimenter artifacts?
- Noninteractional (observer, interpreter and intentional effects)
- Interactional (biosocial and psychosocial effects)
- Situational (environmental) effects
- effects of researcher’s expectancy (E may behave differently to get + results)
What are the controls to reduce threats to validity?
- general control procedures
- control over S/E effects
- control of selection and assignment (randomization)
- Control through specific experimental design (B, W, M)
Procedures to control demand characteristics and experimenter effects?
- Deception (needs debriefing)
- Placebo controls
- Double blind procedure
- limit E communication with S’s
- Automate
- Experimental setting and procedures low key and nonthreatening.