To be able to have neuro-adaptive tools, it is useful to be able to forecast performance when doing interventions. In this paper we use results from an experiment collecting biometric data from different sensors, capturing EEG, eye-tracking, physiological state and facial expression (through cameras). Earlier work has shown the possibility of using such data in causality-analysis. In this paper we investigate to what extent we can use these results for forecasting the effect of different interventions. The paper shows that the basic forecasting results and the models utilizing causality information are significantly better for forecasting performance than models not being based on the causality information. This increases the possibility of using forecasting in Neuro-adaptive modeling tools.