L8: Observational Research Flashcards
0
Q
Driver distraction
A
- anything that distracts from the primary task of driving safely to destination
- delays the recognition of info needed to drive safely
- effects; lane keeping ability, speed control, greater ‘eyes off road’ time, increased RT to unexpected events
1
Q
Observational research design
A
Descriptive research in which the researcher observes and systematically records the behaviour of individuals in order to describe the behaviour
2
Q
Driver behaviour studies from a number of methodological approaches
A
- epidemiological data collection: use of mob phones up to 10 mins before a crash was associated w a 4x increase in likelihood of crashing
- focus groups and interviews about driving habits
- survey
- experimental via simulator
3
Q
Driving simulator vs. real world
A
- Simulator fidelity
• high vs low - external validity
• absolute vs relative - many simulator validation studies
• high-fid sim better
• low-fid sim results still defensible - relative validity only essential element necessary for meaningful simulator research
- on road testing provides many benefits but has limitations
- driving simulation provides a convenient and safe method for assessing driving behaviors
4
Q
Observation w/o intervention
A
- naturalistic observation, physical traces, archival data
- how ‘ecologically valid’?
- how easy will it be to explain the observations?
- what influence does the observer have on the situation?
5
Q
Problems with naturalistic observation
A
- interesting behaviors may occur infrequently
- some events do not occur in public
- hard to examine processes during observation
- no control over circumstances
6
Q
Observation w intervention
A
- Experimental manipulations
- participant observation (when researcher engages in same activities)
• disguised observation: infiltrates groups and observes, ethical problems, experimenter bias and expectancy
• undisguised observation: camera in car study - intervention allows control: allows comparison with diff groups, can vary the settings explicitly
7
Q
Types of sampling
A
- continuous
- time
- event
- situation
8
Q
Continuous sampling
A
- all behavior that occurs within a specified time period is coded, contributes to validity and utility of the study by providing a complete account of all behaviour
9
Q
Time sampling
A
- choosing random/fixed intervals for making observations either systematically or randomly
- just the behaviour that is occurring at the exact moment that your beeper tells you to record the behavior
- can get an estimate of the duration of the behaviour
• whereas continuos sampling only notes when behaviour changes - may capture a range of situations
- may miss infrequent events
10
Q
Event sampling
A
- observe only occurrences of a particular event (e.g. Secondary activity whilst driving)
- focuses on events that are rare but may not generalize
11
Q
Situation sampling
A
- studying behaviour in diff locations and under diff circumstances and cond
- enhances external validity
- is likely to have high ecological validity
- may be expensive to carry out
12
Q
Recording behavior
A
- qualitative records of behaviour: observers must be carefully trained behaviour according to established criteria
- quantitative: frequency or duration
13
Q
Analysis of observational data: data reduction
A
- observational data are summarized through the process of data reduction
- researchers quantify the data in narrative records by coding behaviors according to specified criteria, for example, by categorizing behaviors
- data are summarized using descriptive measures such as frequency counts, means, and standard deviations
14
Q
Observer reliability
A
- extent to which independent observers agree in their observations
- increased by providing clear definitions and events to be recorded, by training observers, and by providing feedback about discrepancies
- assessed by calculating % agreement or correlations, depending on how the behaviors are measured and record
• number of agreements divided by number of opportunities for agreement
