The model may have been created at a different hierarchy level than the level at which the values for the responses are set.

For example, a decision is made for a specific Issue/Group/Name/Channel combination with the following values: Sales/Phones/Samsung Galaxy S/Call Centre. When the response is captured, if the value for Channel is then set to CallCentre or Call Center, or any other different spelling, there is a mismatch between the context values of decision and response. Consequently, the response cannot be assigned to a model and is not captured.

1. In the header of Dev Studio, click ConfigureDecisioningModel Management.
2. Click Show last responses section, and then check the values that are used when the response is captured for potential mismatches.
3. Ensure that the context keys of an adaptive model at the response time are the same when the response is captured. Pay attention to spelling and case.

   You might need to ask a strategy designer to review the strategy that you use.

The outcome values set in the adaptive model rule are different than the outcome values received in the response.

For example, the positive outcome of the adaptive model rule is set to Accept, but the value of the pyOutcome property that is passed at response capture is Accepted. The model cannot learn from such a response.

1. In the header of Dev Studio, click ConfigureDecisioningModel Management.
2. Check whether the outcome values in the responses received match the outcome values configured in the adaptive model rule.
3. Add the outcome values that you get from the responses to the Outcome tab of the adaptive model rule instance.
   For more information, see Defining outcome values in an adaptive model.

The system does not save responses to the pxAdaptiveAnalytics data set.

1. Check the data flow run and ensure that responses are saved into the pxAdaptiveAnalytics data set.