The menu Adaptive Questionnaire allows you to specify the options of the adaptive questionnaire process.
You can choose the target node, and if required a target state, to run an adaptive questionnaire based on a particular target value. It is also possible to specify the desired number of monitors that will serve as an ordered list of questions. This wizard also allows you to Edit Costs associated with the observation of the variables.
Once the adaptive questionnaire is launched, a new toolbar is displayed:
The button resets the adaptive questionnaire.
The buttonstops the adaptive questionnaire, deletes all monitors and removes all observations.
Adaptive Questionnaire Based on a Target Variable
The Adaptive Questionnaire allows the automatic and dynamic reordering of the monitors while taking into account the information that is being contributed to the knowledge of the target variable and taking into account the corresponding cost of the questions. The target monitor is highlighted with a pink background at the top of the list, followed by the monitors in the order of their relevance.
The monitors of the observed variables appear at the end of the list (with a gray background) in the order of their observation. The most recently observed variable thus appears right after the least relevant not-yet-observed variable. You can also remove an observation or change an observed value by double-clicking on the respective monitors. Furthermore, you may observe any variable that is not in the list of proposed questions. Simply select the variable manually and set evidence via the monitor.
Please note that none of the translucent nodes in the graph will be proposed as questions.
Adaptive Questionnaire Based on a Particular State of a Target Variable
This questionnaire is similar to the previous one. Here, however, each question's importance is measured in terms of its knowledge contribution towards a particular value/state of the target variable, as opposed to its contribution towards the overall knowledge of the target variable. In case the target variable has only two states, both questionnaires are equivalent.