Studying students' problem-solving abilities in physics education research has consistently shown that novices focus on a problem's surface features rather than its physical principles. Previous research has observed that some electricity and magnetism students confuse electricity and magnetism concepts, often presented in parallel problems (or problems with similar surface features). This confusion has been referred to as interference. It is essential to compare students' performance in these problems to evaluate their understanding of these topics. The present work focuses on the students' understanding of interactions between charged particles (i.e., electric force) and electric currents (i.e., magnetic force). We present and compare the findings on students' conceptions when analyzing electric and magnetic interactions for different systems of field sources. We conducted this study with engineering students finishing a calculus-based course on electricity and magnetism. We administered a written, open-ended questionnaire with two sets of three items: one version contained only electricity problems, and the other contained only magnetism problems. Each item in the electricity version of the test had a parallel counterpart in the magnetism version. We used a phenomenographic approach to analyze our data to identify categories that emerged from students' answers. We identified four main ideas in the results: (a) the rule of signs (ROS), which does not evidence a complete conceptual understanding of electric interactions; (b) the force-field confusion due to the similarity of electricity and magnetism contexts; (c) the importance of semiotic representation when answering an electricity and magnetism problem, where the student's choice of representation indicates their understanding, and (d) the interference phenomenon, in which we find evidence of other factors besides those produced by the timing of instruction and administration of the tests.