Related Experiment Videos
Subjective rating scales: science or art?
1Department of Psychology, University of Warwick, Coventry CV4 7AL, UK. j.annett@warwick.ac.uk
Ergonomics
|February 6, 2003
Summary
Subjective rating scales in ergonomics are scientific if intersubjectivity is achieved. While human judgment is ordinal, these scales can test theories and understand constructs like fatigue, but general standards remain challenging.
Area of Science:
- Ergonomics
- Psychometrics
- Human Factors
Background:
- Subjective rating scales are prevalent in ergonomics for assessing workload, fatigue, usability, and comfort.
- The scientific validity of these subjective measures is frequently debated.
- This paper examines the scientific rigor of subjective scales in ergonomics research.
Purpose of the Study:
- To debate the scientific nature of subjective rating scales in ergonomics.
- To explore the role of intersubjectivity in validating subjective measurements.
- To discuss the implications of conscious experience and psychometric properties for scientific measurement.
Main Methods:
- Argumentative and critical review of existing literature and concepts.
- Discussion of intersubjectivity as a criterion for scientific probity.
- Analysis of human judgment, error sources, and bias in subjective reporting.
- Examination of psychometric properties and measurement scales (ordinal vs. interval).
Main Results:
- All empirical observations, including 'objective' ones, are inherently subjective; intersubjectivity is key to scientific validity.
- Subjective measures, though often ordinal, can be used to develop and test theories, especially for latent variables like fatigue.
- Dissociation between different measures and the preference for physical over subjective measures are clarified through a construct validity perspective.
Conclusions:
- Subjective rating scales can be scientific when intersubjective agreement is maximized.
- While precise quantitative prediction is difficult, subjective scales are valuable for hypothesis testing and design evaluation in ergonomics.
- Establishing generalizable design standards based on subjective measures for broad constructs like workload is impractical due to sampling complexities.