click below
click below
Normal Size Small Size show me how
D293 Section 4
Assessment an Learning Analytics Section 4 "Bias & Accessibility"
| Prompt | Answer |
|---|---|
| Why Should We Mitigate Bias? | It Ensures Accurate Measurement, Promotes Equity, Provides Valid Data |
| Construct Validity Bias | Inaccuracy in how an assessment measures what it is intended to. Does it accurately measure knowledge or skills? How we construct "build" our assessments. |
| Content Validity Bias | Failure of an assessment to fully represent its subject matter. {Content that was not taught} Example: World History Questions on a US History Test |
| Item Selection Bias | Biased Test Item choices that favor or disadvantage certain groups. Example: Math Word Problems with Football Jargon. |
| Predictive Validity Bias | Inaccuracy in forecasting future outcomes across different groups. Does it accurately predict future abilities? Example: Using a written test instead of a Driving Exam to predict driving skill. |
| Irrelevant Knowledge | Any uncommon information that is unrelated to the assessments targeted learning objectives and goals. |
| Cognitive Load | Overwhelming the working memory's limited capacity handle only a few bits of information at a time. |
| Define: UDL | Universal Design for Learning |
| P.O.U.R. Principles | Perceivable, Operable, Understandable, Robust |
| P-Perceivable | Information and UI components must be presented to users in way they can perceive using their 5 senses. Example: Make it accessible to learners who are hard of hearing or color blind |
| O-Operable | UI components and navigation must be usable. Example if there is a drag and drop, have an alternative method for those who cant use a mouse. Another Example: Text-To-Speech |
| U-Understandable | Information and UI navigation must be understandable, clear, easy to navigate, and use language at the appropriate grade level. |
| R-Robust | Content must be able to be accessed across multiple platforms and devices. |