Numerical literacy is a skill that must be possessed by students. It is the ability to use numbers and mathematical symbols to solve practical problems in daily contexts. In statistics courses, it is identified that 73.3% of students could solve paired t-test problems using the SPSS 21 program, but 86.36% of these students could not understand the symbols and numbers in the output of the program. This indicates that students' numeracy literacy skills need further analysis. This study aims to analyze students' numeracy literacy by using paired t-test statistical problem-solving. The type of this research was qualitative descriptive research with a case study method. This study employed a purposive sampling technique. The subjects were 30 students of mathematics education. Data were analyzed using a triangulation technique between two assignments. This study has obtained sixth major results. First, 53.3% of students cannot write statistical hypotheses correctly. Second, 73.3% of students cannot determine the solution at the output of the SPSS 21 program. Third, 60% of students cannot conclude their answers correctly. Fourth, subjects with low numerical literacy skills could understand the meaning of the questions. Fifth, subjects with medium numerical literacy skills should get training on numeracy literacy-based questions to improve their numerical analysis skills. Sixth, subjects with medium numerical literacy skills could conclude the solution correctly and tend to reflect on the process before concluding a final solution.