Lecture 2- Research Designs, Validity, Reliability And Sample Selection Flashcards
Qualitative Research
- Describes phenomena using verbal means
- Looks for meaningful patterns
- Useful for describing rate phenomenon or an unstudied population
Quantitative Research
- Turns psychological experiences into numbers
- Performs inferential stats tests
- Qualitative can be turned into numbers using a coding system
Naturalistic Observation
- Collects data about naturally occurring behaviour
-Unique features:
•researcher/observer is unobtrusive
•researcher must remain covert to record behaviour
•researcher tries not to let the PPs know they’re being watched
-Awareness of being watched may alter behaviour
- Good: get accurate descriptions
- Bad: purely correlational, cannot isolate causes
Field Studies
- Conducted in real world
- Researcher designs questions + tasks that change behaviour
- Taps into more specific behaviours
Lab Studies
- Conducted in lab
- Better control over variables and settings
- Support stronger claims of cause-effect relationships
- Highly artifical = low generalisability
- Simulations can help
Correlational Method
-Non-experimental method
-Observes and describes relationships between variables
•how much they covary and in which direction
- Used when manipulation is not possible, ethical or preferable
- Can predict behaviour
- Cannot infer causes
- Good in early exploratory stages
- Good for relating naturally occurring variables
-Problems:
•Third variable problem
•Direction of causality
-Solutions:
•Control for possible covariates
•Longitudinal assessment
Experimental Approach
-Observation under controlled conditions, systematically varying 1 or more variables
-Used to determine cause using canons of deductive logic:
•temporal precedence: effect -> cause
•agreement: effect when cause is present
•difference: absent when cause is present
•concomitant variation: rule out alternative explanations
~Basis of experimental method
Experimental Method
- Manipulation of IVs
•At least one variable has to have 2 levels which are systematically changed by experimenter
-Observer effect on DV
-Control extraneous variables
-Random assignment to groups
•Evenly distributes abnormalities or irregularities across groups
-Replication necessary
Experimental method: Advantages
- Better support for cause-effect relationships
- Greater control over all aspects of research
- Ability to precisely manipulate variables of interest
- Findings more enduring over time- can be a basis for more applied research
- May need fewer PPs than non-experimental methods
- May be easier to write up findings
Experimental Method: Disadvantages
-Experiments are artificial
•May not translate to real life situations
•May not be generalisable to other populations
- May cause more reactionary response from PPs
- Difficult to manipulate certain constructs
Basic Research Goals
- Advance knowledge
- Identify relationships among constructs
- Identify relationships among causes
Applied research goals
- Solve problems
- Yield larger effects
- Predict future events
Internal validity
- Extent of cause-effect claims
- The higher the internal validity, the lower the external validity
Internal validity threats
- History: external events other than treatment between conditions
- Maturation: changes over time due to fatigue/age
- Selection: sampling bias for different conditions
- Attrition: condition-sensitive drop out
- Testing: pretesting suggests hypothesis
- Instrumentation: change in observer measurement criteria, lack of standardisation
- Statistical regression: extreme PPs will tend to be closer to the mean on retesting
External validity
- Extent findings can be generalised
- Higher the external validity, lower the internal validity
