1/144
Human Factors Engineering
applies information about human behavior, abilities, limitations, and other characteristics to the design of tools, machines, systems, tasks, jobs, and environment for productive safe, comfortable, and effective human use
Goal of Human Factors
enhance performance, increase safety, and increase user satisfaction
What do Human Factors Engineers do?
design the optimal person-machine, expert on topic of what people can and cannot do, and specifies training and education
Look at Human Factors Cycle
OK
Sense
our sensory abilities extract information from the environment and convey it to the brain
Perceive
we decide what is present in the environment based largely on our previous experiences; people interpret things differently
Teach and Learn
feedback is important; hard to control what a person absorbs
Store and Retrieve from Memory
accessing information needing to use
Respond
after making a decision, it must be executed; people tend to overestimate their abilities
Look over the Stroop Tests
OK OK
Look Over Human Factors Cycle Scenario in Lecture 2
OK OK OK
When does Human Factors become Science?
the result is generalized to a wide range of systems and results can predict future outcomes
Research
used to know whether we’ve accomplished (or could accomplish) the goals of HFE
Basic Research
generalized to a wide range of tasks, settings, and people
Applied Research
specific to particular tasks, settings or people
Continuum
research between basic and applied
Research Design
seek to explore the relationship between 2 or more variables
Independent Variables
casual/control; what you want to influence
Dependent Variables
outcome; results and what we measure
Confounding Variables
things that could influence results but not interested in
Types of Dependent Variables
attitude/motivation, preferences, knowledge, skills, emotion
Experimental Research
seeks to prove that the change in the dependent variable was caused by the independent; lab/field study
Descriptive Research
analyzing relationships that exist between variables, even without control; historical data
Tradeoff
experimental control vs. real-world relevance
Steps of a Research Method
define problems and hypotheses, experiment, conduct the study, analyze data, and draw conclusions
Define Problem and Hypotheses
what questions do you want to answer? what cause/effect relationship do you expect to see?
Specify Experimental Plans/Methods
define independent and dependent variables, describe design, and detail procedure, materials, participants, etc.
Conduct the Study
follow the plan; best practice is to do a pilot test first if uncertain of methods
Analyze the Data
best practice suggests using both quantitative vs qualitative
Draw Conclusions
were hypotheses supported by the results of the experiment?
Statistically Significant
many use standard p-values of 0.05 or 0.01; may relax to 0.10 with exploratory or subjective data
Practical Significance
real-world meaningful impact of a research finding
Experiment Types
laboratory, field, simulation
Labratory Experiment
high experimental control, low relevance
Field Experiment
moderate experimental control, moderate relevance
Simulation Study
moderate experimental control, moderate relevance
Field Study
low experimental control, high relevance
Best Practice Suggests ____ is Good for HFE Research
field study
Common HFE Research Methods
survey, observation, focus group, interview, task analysis
Additional Considerations with HFE Research
hypotheses should be stated prior to experiment; use multiple methods; reliability and validity
Reliability
consistent in measurement
Construct Validity
measure what you intend to measure
Internal Validity
whether conclusions about casual relationships can be made
External Validity
how valid it is to generalize the results of the study
ALWAYS focus on the ethical treatment of human subjects
true
NEVER remember the Code of Ethics for Engineers
false
Human Subjects Research
a systematic investigation designed to develop or contribute to general knowledge, which involves the collection of data from or about living human beings
Why must research be reviewed?
ensure the rights and welfare of human subjects are adequately protected in research; both federal and state laws require this protection
Nazi War Crimes
Nazis conducted large scale sterilization and typhus fever vaccine experiments on unknowing concentration camp populations; Trials of Nuremberg
Tuskegee Syphilis Study
around 600 black men were recruited without informed consent to test syphilis treatment; they were not directed to take penicillin an effective treatment in the 1940s
1948 Nuremberg Code
informed consent, valid research, good of society
1979 Belmont Report
respect for persons, benevolence, justice
What Type of Activities do Human Factors Designers do?
talk to users and other designers
6 Stages of Human Factors in Lifecycle
front end analysis; iterative design and testing; system production; implementation and evaluation; system operation and maintenance; system disposal
User-Centered Design
the creation process is done around the user; support user needs rather than making the user adapt to the system
Information Sources
databases, design standards, principles and guidelines
Databases
anthropometry, simple task times
Design Standards
VDT design, workstation design
Principles and Guidelines
textbook material
Front End Analysis
used to understand the users, their needs, and the demands of the task; can take a bit of time but needed
User Analysis
identify the user population
Function Analysis
what goals are the users trying to accomplish
Task Analysis
what steps must the user take to accomplish these goals
Task Analysis - Step 1
identify tasks (consider information, task sequence, location)
Task Analysis - Step 2
collect task data (observe, think-aloud verbal protocol)
Think-Aloud Verbal Protocol
when you watch/record user doing work and have them think out loud while they do the task
Task Analysis - Step 3
summarize the data (lists, outlines, flow charts)
Task Analysis - Step 4
analyze task data
Iterative Design and Testing
identifies design features and/or constraints; QFD
Quality Function Deployment (QFD)
allows to see how important human factors are (see lecture 3 for pic)
Between Subjects Design
each combination of independent variables is administered to a different group of subjects (participants only see one version of software)
Between Subjects Pros and Cons
pro: more info/focus into each version
con: hard to really tell which is the best version
Within Subjects Design
all combinations are administered to all subjects (each participant sees all software versions)
Within Subjects Pros and Cons
pro: easier to see which version performs the best
con: could cause some bias from seeing all versions
Mixed Designs
use both between and within subjects (EX. subject uses one font size but both colors)
Multiple Dependent Variables
efficient to measure how the causal variables affect several dependent variables at once (EX. measure time and errors)
Two-Group Design
independent variable has two levels (control and experiment group)
Multiple Group Design
independent variable has multiple levels
Factorial Design
more than one independent variable; all possible levels are combined and evaluated, including interactions
Mean
statistical average
Standard Deviation
amount of “spread” or variation in the data
Inferential Statistics
analyze a relationship due to data
Inferential test if DV is normally distributed
ANOVA
ANOVA
compare means between 2+ groups
Inferential test when you have Categorical Data
Chi-Square
Chi-Square
compare proportions across 2+ groups
Inferential test when you want to predict an outcome
Regression
Linear Regression
if DV is interval and normal
Logistic Regression
if DV is ordinal or categorical
Significance
specify confidence level; typically 0.05
If p-value is < 0.05…
results are not due to chance; statistically significant
Look at Lab 4
OK
What Impacts How we See an Image?
visual angle, placement on retina
Visual Angle
a function of height and distance
Placement on Retina
determines type of receptor cells; rods or cones
Vision Acuity
clarity and sharpness of vision
3 Primary Acuity Measurement Techniques
Snellen, Landolt Rings, Vernier Acuity
Snellen Acuity
the eye test at the doctors where you identify letters
Landolt Rings
identifying where the opening in the ring is
Vernier Acuity
determining if two lines are connected or not