Learning mathematics is a challenge in the current educational context. This is demonstrated by a high percentage of students who both fail to develop these numerical-type skills and have a low academic performance. This research performs a quasi-experiment to develop mathematical skills and knowledge. The sample consisted of 54 students (20 female and 34 male) between 10 and 14 years old (M[subscript age] = 11.11, SD = 0.79) organized into two groups. First, an experimental group (N = 26) that received a technological intervention of learning virtual objects and second, a control group (N = 28). The results shown a statistical significant interaction between the experimental group and the post-test, which suggests that the intervention improves the skills to solve root exercises F[subscript (1, 52)] = 10.41, p = 0.002, [eta][superscript 2] = 0.17, with a significant main effect F[subscript(1, 52)] = 1184.02, p = <0.001, [eta][superscript 2] = 0.96. The knowledge of root exercises problem-solving variable has no meaningful interaction with the intervention F[subscript(1, 52)] = 2.14, p = 0.15, [eta][superscript 2] = 0.04. Nevertheless, there was a direct effect between the pre- and post-test measurement differences and the intragroup factor F[subscript(1, 52)] = 17.95, p = <0.001, [eta][superscript 2] = 0.26. The results are discussed about the benefit generated by the use of virtual learning objects in favor of developing mathematical skills and based on previously published studies.