| Home > RippleStat > Statistical Decision Theory | |||
|
|||
|
Statistical Decision Theory |
This page under construction
Decision TheoryDecision theory is all about "making decisions under conditions of uncertainty." It is one set of concepts applied to the Hypothesis Testing Model of Inferential Statistics, Signal Detection Theory (aka "Sensory Decision Theory"), and "Error Management Theory." The screens (aka "cards") in this application (aka "stack") provide interactive practice with various aspects of these models. |
Instructions and Explanation: Built into the programEach of the ways decision theory is applied in this program has instructions and explanation built into the program. When the user clicks on the "Info" button, a seocnd window opens with this information. The "Info" window in this image is displaying the information for the Hypothesis Testing Model of Inferential Statistics. |
Cell LabelThis screen (aka "card") provides the user with an opportunity to simulate a research situation where a prospective new "treatment" is being compared with an established treatment. Then a practitioner can try the new technique and decides whether or not the new treatment makes a difference. |
Trials for Signal Detection Theory TaskThis simulation is included to show the formal similarity between Signal Detection Theory and the Hypothesis Testing Model of Inferential Statistics. |
More InsructionsIncluded to show the Instructions for the Signal Detection Theory simulation. |
Which Error is Critical"Error Management" theory recognizes a basic truth of making decisions under conditions of uncertainty: It is impossible to make a correct decision every time. Error Management basically continues the discuss of "Which decision error is worse: Type-I or Type II?" When should Type-I errors be minimized? When should Type-II? Some time ago Chris Spatz made a presentation at a teaching conference about where the 5% Type-I error criterion arose. Here's some of that information as I remember it. |