2.4.1. Assessing the User's Awareness Phase

In the first step of the reasoning engine, the user's current activity level and awareness phase are assessed to determine what type of support they need. The assessment is based on two evaluations of the user's physical activity level, namely an objective evaluation (i.e., whether the user meets the norm) and a subjective evaluation (i.e., whether the user thinks he/she is sufficiently physically active). Using that information, the user is assigned to one of four categories, each representing an awareness state regarding their physical activity. The resulting categories are summarized in Table 1. Category 1, in which users believe that they are sufficiently active but objectively do not meet the norm, is more or less comparable with the *precontemplation* phase in Prochaska's Transtheoretical Model of behavior change [12]. In this phase, people are uninformed or under informed about the consequences of their behavior and education about the consequences is needed. Category 4 is similar to the *maintenance* phase, while Categories 2 and 3 are comparable to the *action* phase in Prochaska's model. Because our system can objectively measure whether people meet the norm, we can determine an awareness phase in a simpler and more accurate way than with the questions that are often used for determining the stage according to the transtheoretical model, and without the strong assumption that people always go through all phases.


**Table 1.** User categories based on objective/subjective evaluation.

Based on the categorization of the awareness of users, the system determines which type of support (i.e., *education*, *coaching* or *feedback*) the user needs. This assessment is reflected in the type of motivational messages that the user receives from the app (see also Section 2.5). In addition, for users who receive coaching, the system guides them to choose a coaching domain, prompts them to set a specific goal, and predicts the most promising coaching determinants, as further explained below. This evaluation is repeated every three weeks, in order to continuously tailor the system to the user's current state. Thus, instead of treating all users the same, their specific needs and wishes are taken into account. This should lead to improved user acceptance and adherence, and consequently to increased effectiveness of the intervention [13].
