**Description** The most influential theory about tax decisions is the standard economic model of tax evasion (Allingham & Sandmo, 1972). However, empirical evidence for the model’s predictions is ambiguous and its behavioral implications conflict with observed compliance rates in most countries. A recent first process tracing experiment using MouselabWEB evaluated whether participants’ compliance decisions are in line with the behavioral predictions as well as the implicit process assumptions of the Allingham and Sandmo model (Kogler, Olsen, Mueller, & Kirchler, under review). On the behavioral level, there was a general decrease in compliance with increasing financial attractiveness of tax evasion, however, overall compliance levels were much higher than the standard model would predict. Against the predictions, providing additional explicit information about the expected value (EV) had no influence on tax compliance. However, it is possible that explaining and providing EV information actually would lead to more decisions in line with the A&S model, but this first study did not capture this effect due to specific characteristics of the research design (i.e., clarity of instructions, parameter calibration, or the concrete presentation of explicit EV information). Thus, the aim of this current study is to test the core research question with an improved design, investigating whether tax decisions in the lab are influenced by EV considerations. Accordingly, all parameters, including the EV, are explained in more detail prior to the experiment and we include more check items to control for understanding of instructions, parameters, and the concept of EV. Additionally, we either (i) only explain the EV, (ii) explain and present it as a numerical value, or (iii) explain and present it as a visual cue and compare these three groups against (iv) a control condition where no explicit EV information is provided. We expect these changes to result in a more critical test of the hypothesized effect. Furthermore, it allows us to test whether the different presentation modes of EV information have distinct effects. If, however, the more detailed explanation in combination with explicit expected value information does not lead to more choices in line with the model, this study would be further evidence that EV considerations do not guide tax behavior - not even in a more artificial context.