Abstract
The aim of this research is to find out how many students at SDN Batu Ampar 09 Pagi can use the differentiation learning model as part of the independent curriculum to improve their grades. Quantitative methods based on quasi-experimental designs were used. The sampling method used was saturated sampling. If we look at the N-gain score, the researchers found that the experimental class that used the differentiated learning model outperformed the control class. The results of the classification of experimental class mathematics learning outcomes showed that five students got very good scores, eight students got medium scores and eight students got low scores. Meanwhile, one student in the control class got a very good score, eleven students got a medium score and nine students got a low score. Random assignment, that the probability of each research object is the same. There are several ways that researchers can control additional variables, including by conducting pre-tests and post-tests.