Scope of Measurement Flashcards
List 5 possible ways of measuring driving behaviour
- Observational methods
- In car/instrumental vehicle measures
- Driving simulator methods;
- Tests of attributes not specific to driving that may predict driving performance
- Questionnaires or interviews
How would observational methods be conducted, and what are the advantages?;
Disadvantages?
Counting/classifying/rating behavioural events; taking direct measurements; Measures actual driver behaviour, and drivers are unaware they’re being monitored;
Level of detailed information is limited, and experimental control is difficult
How would in car/instrumental vehicle measures be conducted, and what are the advantages?;
Disadvantages?
Physiological measures/eye tracking; vehicle movement;
behavioural events; Detailed information can be collected, and they measure actual behaviour;
Drivers are aware they’re being monitored, and experimental control is difficult.
What is measured in driving simulator methods, and what are the advantages of these?;
Disadvantages?
Response times, and judgements of standardised events; Experimental control can be excellent, and detailed information can be collected;
Not real, must be validated; and can be expensive.
How would you go about testing attributes not specific to driving that may predict driving performance, and what would be the advantages of this approach?;
Disadvantages?
Measure accuracy and response times while doing standardised tests; scoring of response quality using criteria; Excellent experimental control; more likely to be well established, and can collect detailed information; Not real (must be validated); and may lack specific context of driving
What are the advantages of using questionnaires or interviews?;
What are the disadvantages?
Excellent experimental control; can collect detailed information; easy and cheap to collect large samples;
Not real (must be validated); may be hard to collect
dynamic, complex stimuli; and relies on participant honesty and insight (prone to reporting bias)
According to Lees et al. (2010) what are the 2 key dimensions based on driving performance?
Fidelity and experimental control
How could we use models such as the speed questionnaire to predict behavioural outcomes, such
as crash involvement?
Combine measurements of lots of different crash predictors/IVs (e.g. level of distraction, speeding
propensity, traffic violations, etc), and use multiple regressions to predict crash involvement outcomes (DV)
What are the four fundamental principles of human measurement?
Scales of measurement; Standardisation; Reliability; Validity
What are the four scales of measurement? Provide an example of each
Nominal - categories (e.g. male vs. female);
Ordinal - order of ranks; distance between not meaningful (e.g. political parties ranked from left to right wing);
Interval - interval between is meaningful but zero point arbitrary (e.g. time of day/temperature);
Ratio - zero means absence of quantity (e.g. distance)
Which scale of measurement do we normally treat psychological scales as?
Interval, so we can do parametric tests, even though scales are usually ordinal (rank order of individuals)
Define standardisation and provide an example
Ensuring different measurements are in equivalent units before comparing them; converting raw scores into z or t scores (e.g. converting inches, cm or feet into metres)
When is it fine to use raw scores in psychology?
When comparing like with like (e.g. reaction time measured in milliseconds)
Define reliability, and provide some examples in the context of psychology
The extent to which a measurement tool gives consistent measurements; e.g. do people get more or less the same
score if they repeat a questionnaire 3 times?; do they give similar responses across 3 questions measuring the same thing?
Define validity, and provide some examples
The extent to which the measure actually measures what it’s supposed to measure; e.g does a proposed diagnostic test for a disease actually work?; does a questionnaire score reflect actual behaviour? does an IQ test score reflect actual intelligence?
