This Adaptive Questionnaire page is accessible via the following URL: https://simulator.bayesialab.com/#!questionnaire or by selecting Go to Adaptive Questionnaire in the upper right corner
This page allows you to run all the public adaptive questionnaires. The private ones are only accessible if you know their Identifiers and Passwords.
They are three panels:
- The Questions panel where the monitors are sorted with respect to the importance of the corresponding questions
- The Answered Questions panel that contains all the questions that have been observed, or skipped
- The Output Variables panel that contains information on the Target variables.
Selecting the first line (the one without any name) allows to close your current adaptive questionnaire and load the blank page.
Clickingbrings up information about the simulator.
The monitors are dynamically sorted according to the information they bring to the Target variables, given the current set of observations.
When costs are associated with the variables, they are also taken into account for defining the hierarchy of the variables. The cost corresponds to the price to pay to get evidence on the variable (the states do not matter).
Changing the state, the value, or the slider position automatically sets the variable as Observed. The corresponding evidence is set to the node in the underlying Bayesian network in order to compute the posterior probability distributions of all the other nodes.
Upon entering an evidence, the corresponding monitor is moved to the Answered Questions panel and the order of the questions is automatically updated to take into account this new piece of evidence.
When the question is the Answered Questions panel, the evidence can be modified or removed by unchecking the Observed box. If it's removed, the monitor goes back to the Questions panel.
When you cannot answer the most efficient question (the one in the top left corner), you can click the icon(upper right corner of the page) to skip this question. This will move the corresponding monitor to the Answered Questions panel.
Clicking theicon removes all the pieces of evidence that have been set on the input variables. All the input variables are then moved back to the Questions panel.
If you are interested in measuring what is the impact of your observations on the output variables, you just have to select the optionavailable by clicking the icon (upper right corner of the page).
The deltas are displayed between brackets in green and red (for positive and negative impacts respectively).
With gauges, the variations are also qualitatively shown with green and red arcs (for positive and negative impacts respectively).
By default, the variations are computed with respect to the previous state. This allows to measure the impact of each piece of evidence. If you are interested in measuring the impact of a set of evidence, you can define a reference state by clicking by clicking the icon.also available
The iconallows you to store the current set of observations you have entered in a scenario file. A name can be associated with the stored set of observations.
When scenarios are available, the iconis added in the toolbar and allows you to load a set of observations that has been stored in a scenario file.
When observations are stored in a scenario file, clicking the icon allows to manage your observations:
- You can clear the stored observations to reset the scenario file,
- You can save the scenario file on your local disc,
- You can load a scenario file that is on your local disc.
When observations are stored in a scenario file, clicking theicon allows to enter into Simulation Mode
The two iconsallow to go back to the previous set of observations, and advance to the next set of observations respectively.
You can change the set of observations when you are in Simulation Mode. However you cannot store these new observations in a scenario file. You first need to quit the Simulation Mode.
Different layouts are available for each panel by clicking the icon.