Please enable JavaScript.
Coggle requires JavaScript to display documents.
Reliability and Validity - Coggle Diagram
Reliability and Validity
Lab experiments
Reliability
- Standardised procedure and strict controls allow the study to be easily replicated
- As they are easily replicable, reliability of procedure can be tested using test-retest
Validity
- High internal validity due to level of control an experimenter has
- Low validity, demand characteristics and investigator effects can lead participants ro act in an unnatural way
- Tasks and situations can lack mundane realism, low ecological validity
- Double blind method can increase validity
Field Experiments
Reliability
- Due to natural setting reliability is reduced due to experimenter losing control over environment
- These experiments are often hard to repeat, therefore lack reliability
Validity
- Can have higher ecological validity die to the natural setting
- If participants are aware they are in a study they may experience demand characteristics
- More extraneous variables, which reduces internal validity
Natural Experiments
Reliability
- Standardised procedures allow for some degree of replicability
- Controls are not controlled by experimenter so can be hard to replicate
Validity
- As experimenter cannot completely control independent variables, there may be participant or situational variables that confound results
- Demand characteristics and investigator effects can lead to participants acting unnaturally
- Double blind method can increase validity
Case study
Reliability
- As it is a unique case it is nearly impossible to establish reliability
- If study used methods such as interviews and observations, inter observer reliability can be used
Validity
- Many different research methods are used so triangulation can be used to establish validity of data
- Qualitative method of collecting data can mean it is subjectively interpreted
- Can be hard to generalise
Correlation
Reliability
- Standardised data collection methods can be used, so replicable
- If the methods used to collect the two variables is not reliable, either is the correlation
Validity
- Type of study used are questionnaires or taking measurements, validity is only as good
Interviews
Reliability
- Split half technique can be used to establish internal reliability
- External reliability can be established using test-retest method
- Semi-structure and structured interviews can be very hard to replicate
Validity
- Unstructured and semi-structured interviews can be valid as allows ppts to elaborate
- Interview question may not accurately measure what they mean to
Questionnaires
Reliability
- Split half technique can be used to establish internal reliability
- Easy to replicate closed questionnaires
- Open questions can be hard to replicate
Validity
- Open questionnaires can allow more detail therefore higher validity
- Concurrent validity- comparing results of a new questionnaires with an already established one
- Demand characteristics, all answer how they think the investigator wants them to answer
Observations
Validity
- If coding system is flawed, then the results
Reliability
- Natural observations can be near impossible to replicate
- Inter-observer reliability can be used