Most deep learning models trained on time series are too complex for humans to comprehend. Explainable AI aims to interpret such complicated models by shedding light on the black box models. Therefore, users can understand how a model makes a decision and hence trust the model. Does a user engage with the visual explanation, and if so, does it increase the trust in the model outcome? This question is central. To gain insight into user behavior and answer the question, we conducted a user study on examining the sieve lifetime in continuous production of textile fibers operated in a filtration machine. Participants in the user study were divided into two groups: one provided with explanations, and the other not. We then observed users via an eye tracker during decision-making to assess their visual attention. The results indicate that users, provided with explanations, trusted the model more often and made more accurate decisions.