Linear algebra is a crucial part of mathematical training for engineering students and there are many problems in the application of this subject in engineering, physical, and even mathematical disciplines. In this scope, a study on the sustainability of learning this subject and how it can be improved is considered necessary. This article aims to delve into how linear algebra learning is effective and at examining three critical parameters: a) Comparison of summative assessment test results on specific topics with the ones produced in the previous year to assess student learning sustainability. b) Testing the topics based on the approach taken, i.e., based on other mathematical topics and real problems. c) Comparing the effectiveness of learning of problems solved in an analytical fashion and the learning of problems solved with mathematical software's help. The methodology used is based on constructionism that depends on the type of activity given in both the design phase and management phase. From the obtained results of the study, the state of learning linear algebra was measured after one year. From the mathematical retest, it turns out that learning algebra was not effective and the exam grades showed temporary learning. There is a skewed relationship, in the next test those with better grades perform worse than those with lower grades. While in the applied aspect the learning was more sustainable. The scrutiny accentuates the noteworthiness of coordinating the importance of coordinating the semiotic systems for them to produce meaningful and durable learning.