Method
Mental strain is a measurement, not an impression
A design critique calls an interface heavy. Ergonomics declines to treat that word as technical at all, and defines two quantities underneath it that can be measured and compared.
ISO 10075-1:2017 states its position on the loose term plainly: mental workload is an umbrella covering the domain, with no standardised meaning of its own. The terms it does define are mental stress, the external load imposed by the task and conditions, and mental strain, the effect of that load on the person. ISO 10075-3:2004 sets the requirements an instrument must meet to measure them. Between the adjective and the quantity sit two definitions and a measurement standard.
The word that circulates in product review is cognitive load, and it is used as a judgement: this screen has too much of it, that flow reduces it. The usage is not empty — the person saying it has observed something real — but it has no referent that two people can check against each other. One reviewer's heavy screen is another's dense but orderly one, and the disagreement has no resolution procedure.
Ergonomics has spent several decades on the same territory and arrived somewhere more disciplined, partly by declining to standardise the popular word. ISO 10075-1:2017 is explicit in its scope that mental workload functions as a generic term for the field and is not treated as a technical term within the document. What the standard defines, in its terms and definitions at clause 3, is a pair: mental stress, the totality of assessable influences impinging on a person from external sources and affecting them mentally, and mental strain, the immediate effect of that stress within the individual, depending on their own preconditions including habitual and actual coping.
The separation is the useful part. Stress is a property of the task, the interface, the environment and the time available; it can be altered by design and is the same for everyone exposed to it. Strain is what that produces in a particular person, and varies with their experience, state and capability. A design change acts on stress. A measurement taken from a participant reports strain. Conflating the two produces the familiar unresolvable argument, in which one party describes the interface and the other describes an experience of it.
The instruments
ISO 10075-3:2004 exists because measurement in this field needed requirements rather than enthusiasm. It sets out principles and requirements for methods of measuring and assessing mental workload, including the precision a method must offer and the conditions under which a given class of instrument is appropriate. An instrument meeting those requirements produces a number that can be compared across designs, which is the property the adjective lacks.
The most widely used subjective instrument is NASA-TLX, developed by Hart and Staveland and published in 1988 after extensive development work at NASA Ames. It decomposes the construct into six dimensions — mental demand, physical demand, temporal demand, performance, effort and frustration — rated on scales, with a weighting procedure derived from pairwise comparisons that the person performs for the task at hand. Hart's 2006 retrospective reported that the instrument had by then been used in several hundred published studies across domains, and noted that many users adopted the raw unweighted ratings rather than the full weighting procedure, which has become a common variant in its own right.
The decomposition is what distinguishes it from an impression. A design that scores high on temporal demand and low on mental demand requires a different change from one scoring the reverse. The composite number is the headline; the profile across the six dimensions is the finding.
Subjective instruments are one of three families, and the others answer different questions. Performance-based measurement, including the secondary-task technique, infers spare capacity from what a person can do concurrently with the primary task. Physiological measures index strain through variables the person does not control. Each family has known weaknesses, and the reason to name the family being used is that each is informative about something slightly different.
What the numbers support
The practical gain is not precision for its own sake. It is that a quantity can carry a comparison and an adjective cannot. Two candidate designs measured under the same tasks and conditions produce two profiles, and the difference between them is a finding that survives the meeting. A design measured before and after a change establishes whether the change did what it was intended to do. A design measured against the range reported for comparable tasks establishes whether the demand it imposes is ordinary or exceptional.
None of this requires laboratory apparatus. The subjective instruments are short, are administered immediately after the task, and have published procedures. What they require is that the task be specified, the conditions held constant, and the measurement taken from participants rather than from the team.
The profile in practice
The change is to how a demand problem is specified and closed. Where the complaint is that a screen feels heavy, the remedy is whatever the loudest reviewer proposes, and nothing establishes afterwards that it worked. Where the finding is a measured profile, the remedy addresses the dimension the profile identifies, and the same measurement taken again establishes whether it moved.
In the interface the difference is specific. A high temporal demand is addressed by removing the deadline the design imposes: allowing the task to be left and resumed with state intact, rather than by simplifying a layout that was never the constraint. A high mental demand is addressed by moving what the person is holding in memory into the display, so that the comparison or the carried value is in front of them rather than retained. A high frustration score with adequate performance points at a path that works and costs, which is a different defect from one that fails. The profile tells the team which product they are looking at, and the adjective does not.
Sources
ISO 10075-1:2017, Ergonomic principles related to mental workload — Part 1: General issues and concepts, terms and definitions. Scope statement on mental workload as a generic term, and the definitions of mental stress and mental strain at clause 3. Definitions are paraphrased, not quoted. - ISO 10075-3:2004, Ergonomic principles related to mental workload — Part 3: Principles and requirements concerning methods for measuring and assessing mental workload. - Hart, S. G., & Staveland, L. E. (1988). Development of NASA-TLX (Task Load Index): Results of empirical and theoretical research. In P. A. Hancock & N. Meshkati (Eds.), Human Mental Workload (pp. 139–183). North-Holland. - Hart, S. G. (2006). NASA-Task Load Index (NASA-TLX): 20 years later. Proceedings of the Human Factors and Ergonomics Society 50th Annual Meeting, 904–908. doi:10.1177/154193120605000909. Source of the usage count and the observation about raw unweighted ratings. - Wickens, C. D. (2008). Multiple resources and mental workload. Human Factors, 50(3), 449–455. Background on the resource account underlying secondary-task measurement.
All sources are paraphrased; nothing above is quoted. The three-family taxonomy of measurement approaches is standard in the workload literature rather than attributable to a single cited work.