Human Factors in Class: Lectures 1-6

WWII

- Begin to see that work requirements exceeded capabilities of people - Despite high motivation and skill level - Examples: radar & sonar operators, pilots, anti aircraft gunners - Research showed that performance suffered when equipment was not designed to fit the human - Key new idea: Required psychologists and engineers to work together to develop components (e.g., Paul Fitts, Alphonse Chapanis)

Definition of HF/E from HFES

- Ergonomics (or human factors) is the scientific discipline concerned with the understanding of interactions among humans and other elements of a system, and the profession that applies theory, principles, data and methods to design in order to optimize human well-being and overall system performance. - Ergonomists contribute to the design and evaluation of tasks, jobs, products, environments and systems in order to make them compatible with the needs, abilities and limitations of people.

2000-2010

- Greater pubic awareness - Mobile computing - Consumer technologies - Design thinking - IT companies driving many innovations--APPLE

Objectives of HF/E

- Help users to successfully perform intended action with a system - Optimize human-system interactions

19th century: 1840s--Ignaz Semmelweis:

- In Vienna, 1 out of 10 mothers died when giving birth from what is known as "childbed fever" - Semmelweis found that the doctors were not washing or disinfecting hands - Hand washing reduced mortality to below 1%

Human Machine systems

- Integration of human and machine in an environment to achieve set of functional requirements. - Procedure for human machine system development: 1. Define objective for system. 2. Determine requirements system must fulfill. 3. Design components to address requirements. 4. Integrate human control through interface design.

WWI

- Introduction of equipment that not all soldiers could operate (e.g., warplanes) - Development of selection tests to identify people suited for jobs - A new idea: Recognize individual capabilities and limitations

1980-1990

- New challenges, methods, & theories - PC revolution propels HF and human computer interaction into limelight - "ergonomic" & "user friendly" products - e.g., office furniture, PC input devices - Motivated by disaster - e.g., Three Mile Island (1979), Bhopal (1984), Chernobyl (1986), Challenger (1989) - Liability and litigation

2010s-Present

- New societal issues around technology (i.e. Privacy, Ubiquitous computing(can now do work while walking bec of computers), Automation and work) - New sensor technologies - Design for underserved populations--UNIVERSAL DESIGN) (i.e. Aging population and diverse users)

4th century BC: Hippocrates:

- Noted the toxic lead in the mining industry • Earliest recognition of "environment worker health" - How a surgeon's workplace should be designed and how the tools he uses should be arranged

Cockpit flap and gear controls

- Pilots of P 47's, B 17's, and B 25's frequently retracted wheels after landing - In all 3 aircraft, controls for the wheels and flaps looked identical and were co located - In the C 47, on which such mistakes were rare, the controls were separated and activated differently - After attaching a rubber wheel to the gear control and a wedge to the flap control, the accidents ceased

some physical work analysis methods

- RULA (rapid upper limb assessment) technique - strain index - lumbar motion matching

1990-2000

- Regulation and institutionalization - in some industries (aviation, surface transportation, military, manufacturing) - resistance in others (medicine, small business-->bec experts in field) - Safety culture - Internet and collaborative computing

Post WWII

- Research fueled by reconstruction - European Productivity Agency --"Fitting the task to the worker" project - Concerns for productivity and social conditions - Expansion in military laboratories - Role of the cold war - New Idea: Shift to emphasis on systems over components -F becomes a profession and discipline • Professional Societies formed

1960-1980

- Research fuelled by space program - Expansion into commercial sector: - Pharmaceuticals, computers, cars, workplace design, consumer products - Introduction of control engineering models and methods: - Transfer functions used to prediction human machine system performance - Still, relatively unknown to the average person

objectives of HF/E research

- describe understand predict relationships between: independent (controlled) system features/variables and dependent (observed) system outcomes/variables -quantify human capabilities and limitations (i.e. how soft can you hear?) -validate theories of human behavior and performance -evaluate human performance with systems under different conditions (i.e. different interface designs, training, etc.)

objectives of HF/E work

- enhance HSI: improve performance, increase safety, increase user satisfaction - using the HF/E design cycle: understand, create, evaluate - pragmatic goals - human centered design

Roadblocks to effective design

- reduce the gulf of evaluation -reduce the gulf of execution

Task sequence:

-Describes the order of tasks and the relationships between tasks over time -Useful in determining how long a set of tasks will take to complete or in estimating the number of people required to complete them -Specific task sequence information include: • Goal or intent of task • Sequential relationship (what must proceed or follow) • Triggers or event that start a task sequence • Results or outcome of performing the tasks • Duration of task • Number and type of people required • Tasks that will be performed concurrently

annett's dichotomy of ergonomics methods

-analytic: help HF analyst gain understand of mechanisms underlying interaction between human and machines (system requirement specification) - evaluative: used to estimate parameters of selected interactions between human and machines (i.e. experimentation involving performance measurement, psychometrics

general HF/E research approach

-begin with question/hypothesis (expectation) on human performance -use empirical methods to obtain answer -two stages of scientific inquiry 1. gather information--observation and data collection 2. analyze data and make inferences in academic settings, HF/E research involves design of experiments and statistical analysis

some psychophysical research methods

-electrodermal measurement -heart rate and heart rate variability for estimating mental effort -blood pressure measurement to evaluate physical and cognitive workload

methods and tools for different domains

-high risk domains: probably vee -workplace: plan do check act -consumer products: scrum, allows you to stay flexible because what people like changes

some macroergonomics research methods

-interview methods -questionnaires/surveys -focus groups for organization analysis -field study

some behavioral/cognitive research methods

-observation of task performance -interviews for usability assessment -focus groups for usability analysis -critical decision method (i.e. interviews with emergency operators regarding decision making and work methods during critical events)

categories of contemporary HF/E research methods

-physical work analysis methods -psychological methods: allow for inference of cognitive state based on physical measures -behavior and cognitive methods: ... (fill in) -teamwork analysis methods: assessing team communications and collaboration relative to outcomes -macro ergonomics methods: analysis of how work environment design factors influence system performance

plan-do-check-act-cycle

-see picture -improves designs of processes, very iterative -4 phases: 1. act: implement the best solution 2. plan: identify your problems 3. do: test potential results 4. check: study results

some teamwork research methods

-simulation training for teams -team task analysis -team behavior rating scales -team communications analysis

work system modeling

-used to ID interrelationships among people, processes, and outcomes in complex socio-technical systems -feedback loop identification and causality analysis

Define the Purpose of the Task Analysis

Define what design considerations the analysis should address, for example: • defining training requirements • identifying software/hardware design requirements • redesigning processes • assessing system reliability • evaluating staffing requirements • estimating workload The defined purpose will influence the information that is gathered

Information flow:

Describes the communication between people and the roles that people and automated systems play in the system Examples: • individuals and their roles and responsibilities • skills and knowledge of the individuals • flow of information • information needed and information generated by each individual

Middle ages

Design of armor and swords probably had to involve some sort of anthropometry (human dimensions study) -diff types of people who use it

Applied research

Development of theory, principles, and findings that are relatively specific with respect to particular populations, tasks, products, systems, and/or environments

Basic research

Development of theory, principles, and findings that generalize over a wide range of people, tasks, and settings

level of automation

Function allocation: • Distribution of control functions to human and machine based on capabilities and capacities. • Referred to as "levels of automation". • Range from manual control to full automation.

Analyzing and Describing Systems

Generic functions of human machine system components (Sanders & McCormack): - Sensing Perceiving information. - Information storage Memory. - Decision making Selecting among alternatives - Action Carrying out an action (e.g., the physical world) Endsley and Kaber's model: - Monitoring Scanning displays to perceive system status. - Generating Formulating options or strategies for achieving goals. - Selecting Deciding on a particular option or strategy. - Implementing Carrying out the chosen option.

19th century: Wojciech Jastrzębowski (1799 1882)

One of the main creators of ergonomics - "An Outline of Ergonomics or The Science of Work based upon the truths drawn from the Science of Nature" (1857) - ergon (work) , and nomos (principle or - Noted 4 chief considerations - With what creatures do we share work? 1. In what periods of life are we best suited for work? 2. In what manner may we proceed in this work? 3. What are the benefits to be drawn from work?

Results of ineffective design

Procedures that require people to behave in unnatural ways... Operator (deliberate) violations of poor procedures:

NSF (National Science Foundation) funding historically focused on basic research but is now open to applied research

Some applications pose unique demands that lead to unique behaviors and, consequently, new variations on models of human performance

Method selection

Speed accuracy tradeoffs • Fast and dirty • Slow and methodological Cost benefit tradeoffs • Financial and resource constraints • Efficacy • Number of usability studies required Tool availability • Expertise required

19 and early 20th century: Frank and Lillian Gilbreth

Time and Motion Study: - Taylor's ideas fell short when it came to managing the human element - Skilled performance and fatigue - Motion studies used to reduce the number of motions in performing a task - A new idea: adapt equipment and procedures to people

Descriptive research:

What is the situation or what is happening? (e.g., gender relates to income • Collect wide range of social and economic indicators to identify structural relations of phenomenon • Often times independent variable(s) cannot be manipulated • Leads to questions of why relationships exist

19th and 20th century: Frederick Taylor (1880-90s)

Work Study of Frederick Taylor - Taylorism or Scientific Management (one of the first management consultants) - By analyzing work "one best way" of doing work can be found - Design for more efficient work

Operational Sequence Diagram (OSD):

a graphical flow-chart that captures the sequence of activity and categorize the operations into various behavioral elements

understand

accident investigation, observation, task analysis

team

authority, communication patterns, responsbilities

Field research

conducted in real worldscenarios: • i.e. measure time it takes to package produce • i.e. understand occupational ergonomics of real staff Pros: • Promotes realism and generalizability of results • Motivated participants Cons: • Costly • Difficult to schedule • Limited control of variables

organizational

corporate culture, rewards structures, staffing levels

Goal:

end condition or reason for performing the tasks

systems design approaches

formalizaed methods for approaching the design of human machine systems -vee process -plan-do-check-act-cycle -scrum approach

Laboratory research

happens in controlled environment Pros: • Low cost • High level of control of variables • Study can be replicated with accuracy Cons: • Low external validity • Small sample size

evaluate

heuristic eval, usability testing, system evaluation

human tech approach

identify a human or societal need, a problem with studying, and then tailor the technology to the specific relevant HF

psychological

information content, structure, cause/effect, reaition

strength assessment

lower limb muscle strength estimated using dynamometer

political

policy agenda, budget, allocations, laws, regulations

differences between analytic and evaluative methods (annett's dichotomy)

see picture

vee process

see picture -anything large in scale that requires a lot of structure (i.e. govt projects, planes, etc) -tests are based on the specifications (requirements guide testing) -also takes into account long term maintenance -not necessarily iterative, it has a system lifetime that it covers

Human Tech ladder

see picture -examine what is likely to constrain actions at different "rungs" of the human tech ladder -levels (bottom up): physical, psychological, team, organizational, political

scrum approach

see picture -product backlog->sprint backlog->spring-> working increment of the software -good for IT, small projects, able to stay flexible and fast -in these situations there is high uncertainty, don't know methods involved... sooo set small goals that are achieved very quickly

physical

size, shape, location, weight, color, material

Tasks:

specific activities needed to carry out a function

create

task equipment environment training selection organization

Functions:

the general transformations needed to achieve the goal

cognitive engineering (woods and roth)

• "Applied cognitive science that draws on knowledge and techniques of cognitive psychology and related disciplines to provide the foundation for principle-driven design of person-machine systems." • The workplaces of today have increased the cognitive workload for workers; "more and more we create or design cognitive environments." • How do we build tools to help individuals solve problems and accomplish tasks?

supervisory control

• A human operator of a complex system intermittently programming and continually receiving information from a computer that closes an autonomous control loop of a system (Sheridan, 1992). • Human monitoring of automated system for error conditions. • Human intervention in system control loop. • Error recovery and return to automation.

What is Task Analysis

• A way of systematically describing human interaction with a system to understand how to match the demands of the system to human capabilities • Study of the cognitive or physical activities required of a person or team to achieve a goal • Results can differ based on which users are studied even when the technologies are similar

types of research methods

• Basic vs. Applied • Quantitative vs. Qualitative • Descriptive vs. Explanatory • Field vs. Laboratory

Focus of HF/E: Why should you know human factors?

• Because bad design is easy • Safety incident costs can bankrupt companies • Retrofitting and training are not cost effective • We can design systems that human can't manage without automation

Human Machine Interaction

• Characterizing systems and system types: (What is the degree of manual versus automated control?) • Manual Complete human control. • Mechanical / electromechano Humans act as controllers but computers filter information between human and system. • Automated Supervisory control of fully autonomous systems.

Two General Types of Systems (Based on error feedback):

• Closed loop systems: • Open loop systems:

How to Perform a Task Analysis

• Define the purpose and identify the data required • Collect task data • Interpret task data • Innovate from task data

Descriptive methods

• Describe population of users, e.g., anthropometry studies • Use means, SD, and frequency, no formal statistical analysis

Location and environmental conditions:

• Describes the physical world in which the tasks occur Examples: • Path of motion • Location where particular tasks should be performed • Physical structures (walls, partitions, desks) • Tools and their location • Conditions under which the tasks are performed

Different Analytic Methods

• Direct observation • Focus groups and interviews • Survey Design • Accident analysis • Time motion studies • Contextual Inquiry • Task analysis

• Optimize human-system interactions:

• Enhance performance • Increase safety • Increase user satisfaction

Reduce the gulf of evaluation

• Gap between system and user • 4 -> 1 • Figure out if the system did it right • Improve visibility -> support user's interpretation of system state

Reduce the gulf of execution

• Gulf (or gap) between goal and action • 2 -> 3 • Figure out how to operate a system • Improve usability -> reduce gulf of execution

Categories of Required Data

• Hierarchical relationships • Information flows • Task sequence • Location and environmental condition

Motivations for studying HF/E

• Human error ≠ cause of system failure • Complex system designs may not be compatible with humans (Wickens, 1993) • Example of a complex system: Boeing 737 Max 8 (2018-2019) --Changes in aircraft design and underlying systems led to two major aviation accidents: Ethiopian Airlines Flight 302 (157 fatalities), Lion Air Flight 610 (189 fatalities) --Flight dynamics changed, Automation (MCAS) used to compensate --Training issues --Situation awareness and information overload • Need for safety in complex sociotechnical systems • Usability in simple technologies

Help users to successfully perform intended

• Human operator provides input control to a system • System interprets commands and processes • System provides feedback to human on system status • Human operator understands information

Task Analysis Purpose

• Identify actions, system requirements, and personnel requirements • Designing procedures, instruction manual, training programs • Allocate functions • Designing interfaces (e.g. displays, workstations, etc.) • Designing the evaluation of the system

Notes Re: Data Collection

• Limitation: Over reliance on existing systems • Analysis should focus on basic user goals and needs or system level functions and constraints rather than on the exact way they are carried out using the existing system • You want to analyze the task data to identify new design concepts to help users achieve their goals rather than design to fit the current tasks

Interpret Task Data

• Lists, Outlines, and Matrices • Flow charts, timelines, and maps • Hierarchies and networks

Implications of function allocation

• Low or high workload. • Low or high system/situation awareness. • Skill decay.

Task Data Collection Methods

• Observe task performance - Direct observation - Automatic data recording - Think aloud verbal protocols and probe questions • Ask users or experts - Talk through - Interviews - Surveys and questionnaires • Use existing information - Documentation - Training materials

Procedures that require people to behave in unnatural ways...

• Stay alert for extend periods of time (fatiguing processes) • Multitasking (divided and shared attention) • Accurate and precise monitoring (vigilance tasks) • Interruptions during tasks (goal changes and re-orientation costs)

Task Data Collection

• Task analysis is conducted by interacting extensively with multiple users (experts) • Capture how tasks and subtasks are performed to achieve goals • Record instances where the current system makes it difficult for users to achieve their objectives (opportunities for redesign and improvements) • Note: task analysis is NOT a usability study

Human factors vs ergonomics

• Terms are synonymous. • "Human factors" typically used in U.S. • "Ergonomics" typically used in elsewhere. • Some have attempted to differentiate terminology: • Human factors - Focused on human cognition, decision making, and mental demands. • Ergonomics - Focused on physical demands and physical workload responses. • Course will use "human factors" to refer to area of study and will focus on human information processing issues in systems design.

Why HF/E is unlike other disciplines

• The objects of analysis (humans) exhibit immense variability • Design decisions are often based on scientific principles as opposed to scientific laws • HF is an inter-disciplinary field

Operator (deliberate) violations of poor procedures:

• To be more efficient • To meet time constraints • To reduce high workload • To promote comfort • To analyze errors • To prevent accidents • Develop heuristics (rules of thumb) for greater effectiveness

Why would we care about the history of a discipline?

• Understand how the evolution of technology leads the development of HF • Understand how the relationship between people and technology has changed (in the West) • Develop awareness of the motivations of different industry sectors • See examples of the practical problems that led to human factors concepts

Task Analysis Outcome Task analysis results in an information base that includes:

• User goals • Functions • Major tasks to achieve goals • Information required • Inputs and outputs

Front end Analysis

• Who are the users? • Why do users need the product and what are their preferences? • What are the environmental conditions under which the system or product will be used? • What is the physical and organizational context of the users' activity? • What major functions must be fulfilled by a person, team, or machine? • When must tasks occur, in what order, and how long do they take?

Explanatory research:

• Why is this happening? • Typically use measurement or experimentation to establish causal explanations - e.g., gender is associated with specific types of training or life responsibilities that mediate income level

Explanatory methods

• e.g., Want to know effects of touch screen button size on user performance (e.g., why are users pushing wrong button) • Hypotheses (button size, physical ability, etc.) and experimentation • Make inferences based on statistically reliable data

Hierarchical relationships describe:

• function, task, subtask relationships • how tasks are composed of subtasks • how groups of tasks combine into functions

Quantitative and qualitative research methods differ in:

• types of questions • analytical objectives • types of data collection instruments • forms of data they produce • degree of flexibility built into study designs