曝光台 注意防骗
网曝天猫店富美金盛家居专营店坑蒙拐骗欺诈消费者
June 2007
Research Criterion
Validity
Face Validity
Content Validity
The variable(s) being sampled must be part of a
domain, such as a body of knowledge or a set of
behaviors
– Performance of a specific task (optometrists testing for
glaucoma, etc.)
– Achievement tests to measure the knowledge
Must include various facets of the performer’s
(subject’s) performance, rather than just one facet.
Construct Validity
A project supported by AIRBUS and the CAAC
June 2007
Research Criterion
Validity
Face Validity
Content Validity
Construct Validity
Does the measure really measure the “construct”
being studied?
Is the experiment really looking at the underlying
issues/problems of interest?
Is the right question being asked?
Construct validity is based on a judgmental
assessment of an accumulation of empirical
evidence regarding the measurement of the
variable in question.
A project supported by AIRBUS and the CAAC
A project supported by AIRBUS and the CAAC
June 2007
Research Criterion
Sensitivity
Criterion should be measured in units that are
commensurate with the anticipated differences
that one expects to find among subjects.
A project supported by AIRBUS and the CAAC
June 2007
Research Criterion
Sensitivity
How gross or fine should the scale be? How to
decide?
Sample size
Differences in applicability
– Would a finer scale make the data significantly more/less
comprehensible and/or useful?
A finer scale is not always physically or
technologically possible
A project supported by AIRBUS and the CAAC
A project supported by AIRBUS and the CAAC
June 2007
Effect of Human Reliability
Human reliability is the inverse of the fact that
being human involves errors.
As a methodology, this is a method for collecting
and then analyzing quantitative data to predict
the likelihood of human error.
Theoretically, this involves discussion and
explanations of how the errors are introduced and
produced.
A project supported by AIRBUS and the CAAC
June 2007
Effect of Human Reliability
As a measure, human reliability is simply the
probability of successful performance of a task
by a human.
Same as systems reliability – where humans are
the element of the system being studied.
The measure is expressed as a probability of a
successful outcome.
A project supported by AIRBUS and the CAAC
A project supported by AIRBUS and the CAAC
June 2007
3. Examples of research
Research for design of product and system
Computer desk
Cockpit design
The Amsterdam Tram
Improving safety and efficiency in operating system
(human errors management)
HFACS
June 2007
A project supported by AIRBUS and the CAAC
Porter, Gyi and Robertson (1992)
Tilting computer desk
A project supported by AIRBUS and the CAAC
A project supported by AIRBUS and the CAAC
June 2007
Problem with existing desks
A need to improve working posture
Desks not deep enough
Computer screen too low
Viewing distance too short
Lack of adjustability
A project supported by AIRBUS and the CAAC
June 2007
A project supported by AIRBUS and the CAAC
A project supported by AIRBUS and the CAAC
June 2007
Methodology
Literature
2 hour experiment
16 subjects
Anthropometric data
A project supported by AIRBUS and the CAAC
June 2007
Methodology
Posture data
Discomfort data
Design feature checklist
Performance
Mood checklist
A project supported by AIRBUS and the CAAC
A project supported by AIRBUS and the CAAC
June 2007
Body part discomfort map
A project supported by AIRBUS and the CAAC
June 2007
Results
The prototype design encouraged a better
posture (confirmed by the literature)
Less reported discomfort in the neck, buttocks,
wrists and hands
A project supported by AIRBUS and the CAAC
A project supported by AIRBUS and the CAAC
中国航空网 www.aero.cn
航空翻译 www.aviation.cn
本文链接地址：
AIRWORTHINESS AND FLIGHT STANDARDS(31)
