Interactive System Design and Human-Computer Interaction Study Guide
Experimental Methodology and Scientific Validity
Internal Validity: The internal validity of an experiment refers to the extent to which a causal conclusion is warranted. This validity can be adversely affected by several factors. Crucially, a noisy environment is a primary factor that hampers the concentration of participants, thereby negatively impacting the internal validity of the study.
Note: Factors such as non-representative samples or non-representative testing environments typically affect external validity, whereas internal factors focus on the integrity of the experimental process itself.
Scientific Methods Comparison:
Observational Method: This method is characterized by high relevance to real-world situations but low precision in controlling variables.
Experimental Method: This is the only scientific method that allows researchers to establish a causal relationship among variables.
Correlational Method: This method identifies relationships or associations between variables but does not prove causation.
Longitudinal Experiments: A longitudinal experiment is defined specifically as a user study that lasts through multiple sessions over an extended period of time to observe changes or trends.
Scientific Investigation and Peer Review:
A controlled experiment is the specific type of scientific investigation designed to test a hypothesis.
Peer Review: This is defined as a specialized process used by the scientific community to select articles for publication, ensuring quality and credibility.
Documentation: It is essential for scientists to record every step of an experiment to ensure it can be repeated or replicated by others.
Replication: When a different scientist repeats an investigation and achieves the same or similar results, they have successfully replicated the investigation, thereby proving its validity.
Scientific Organizations: The American Psychological Association (APA) is the scientific association responsible for publishing a manual containing critical information regarding methodology for experiments involving human participants.
Variables, Data Measurement, and Statistical Analysis
Definitions of Variables:
Independent Variable (IV): This is the specific factor that a researcher changes or manipulates in an experiment (e.g., the type of soil in a plant growth study). It is often referred to by the synonym Factor.
Level: This term identifies a specific possible value assigned to an independent variable.
Dependent Variable (DV): This is the result or outcome being looked for in an investigation (e.g., the actual growth of a plant). It depends on the changes made to the independent variable.
Control Variable: These are the factors that must be kept exactly the same throughout an experiment to prevent them from influencing the results (e.g., plant type, age, and water amount).
Confounding Variable: This is a variable whose value varies systematically with an independent variable, potentially confusing the results and making it difficult to isolate the cause of an effect.
Variable (General): Any factor that can change in an experiment or affect the outcome.
Calculations and Statistics:
H-Index Calculation: The H-index measures both the productivity and citation impact of a researcher. Based on a researcher with 5 articles receiving citations of , and , the H-Index is 3 (calculated by the number of papers that have at least citations).
Independent Variable Effects: If an experiment includes independent variables, there are 7 possible effects they can have on a dependent variable (calculated as for main effects and interactions).
Statistical Software Interpretation: In analyzed outputs, if a null hypothesis can be rejected for , it indicates high statistical significance.
Data Types and Compatible Measures:
Nominal Variable: The only measure that can be calculated is the frequency distribution.
Ordinal Variable: Compatible measures include the frequency distribution and the median.
Human Performance Data: Data regarding human performance (such as text entry speed) generally follows a Normal distribution.
Data Visualization: To show change over time in an experiment, scientists utilize a line graph.
Human Factors, Physiology, and Performance Models
Human Vision and Saccades:
Fovea: The central part of the retina characterized by the highest visual acuity.
Saccades: These are rapid/quick eye movements that generally have a duration of approximately .
Cognitive Performance:
Memory Limits: In a test of cognitive capacity, a human is able to correctly replicate a sequence of 7 numbers in approximately 50\% of cases.
Simple Reaction Time: The delay measured between the occurrence of a single fixed stimulus and the initiation of the assigned response.
Theoretical Models:
Card, Moran, and Newell Model: Presented in "The Psychology of Human-Computer Interaction," this model describes the human processor as consisting of three subsystems: the Perceptive, Cognitive, and Motor processors.
Kantowitz and Sorkins (1983) Model: This model suggests that both the human and the computer are composed of 3 subsystems each.
HCI History, Devices, and Interface Design
Historical Milestones:
Sketchpad: Developed by Ivan Sutherland, this system is considered the first to adopt a "direct manipulation" interface. It dates back to the 1960s.
The Mouse: The invention of the mouse dates back to the 1960s. In the first HCI history user study, the mouse's obvious advantage over other devices was that its error rate was much smaller.
Xerox Star: This was the first commercial computer to feature a Graphical User Interface (GUI).
Device Characteristics and Spatial Relationships:
Direct Manipulation: An interface style allowing users to interact directly with graphical objects.
Natural Mapping: A spatial relationship between a control and a display is considered "natural" when there is a spatial transformation between one of the axes of the control and one of the axes of the display.
Degrees of Freedom: A pendulum has 1 degree of freedom.
Classification: A scrollbar slider is classified as both a control and a display.
Spatial Vocabulary (Rotations and Translations):
Roll: Rotation around the longitudinal (z) axis.
Yaw (Imbardata): Rotation around the vertical (y) axis.
Mouse-to-Pointer Mapping: The spatial relationship between mouse axes and screen pointer axes is faithfully represented as and .
Device Modes: Digital controls support different numbers of modes. For example, the "Navy key" button of the Nokia 3210 supports a smaller number of modes compared to complex laptop keys or toolbars.
Scientific Writing and Reporting
Paper Structure: In a scientific paper, the Procedure subsection is where the specific task carried out by participants is described. Other sections include:
Apparatus: Describes the hardware and software used.
Participants: Describes the demographic and sample size.
Design: Outlines variables and experimental structure.
Experimental Design Advantages:
Between-Subjects Design: A major advantage of this design is that there is no interference (or carry-over effects) between experimental conditions.