There are several factors that have an effect on physics learning for engineering students: from strong alternate conceptions, to attitudes toward their learning or expectations of the use of physics in their other courses during their undergraduate education, or, even further, to their professional practice once they graduate. One of the factors proven to have a strong effect when using active learning strategies has been students' reasoning. Studies show that physics learning does not correlate to students' reasoning level for traditional teaching. On the other hand, physics learning has a positive correlation to students' reasoning level for those taking active learning physics classes. There are several instruments to assess reasoning, such as the Lawson test. This test is well known among the physics education research community and is often used to measure fundamental reasoning elements through simple context situations. The test consists of 12 pairs of multiple-choice items in which, for each pair, the second item consists of the different reasoning students might have to answer the first item. Although Lawson test's results have been documented in several different studies, often the use is limited to analysis using the general score or, at most, using the different Lawson test levels of reasoning. The objective of this study was to conduct an in-depth examination of the tests results by analyzing students' paths for each pair of items. To that end, 500 undergraduate engineering students took the Lawson Test. For each pair of items, the analysis looked at the most common students' paths when answering the question correctly or incorrectly. By combining the results for each dimension of the test, an indepth analysis on students' reasoning was performed for two dimensions. Moreover, by characterizing students' paths, we believe that physics education researchers will have another tool to design activities to develop students' reasoning skills and therefore, increase engineering students' physics learning.