This study aims to analyze wether (1) there is a difference in students' mathematical problem solving abilities and self regulation learning (SRL) taught contextual learning models with students taught cooperative learning models geogebra assisted; (2) there is an interaction between learning models with early mathematical abilities of students mathematical problem solving abilities and SRL. This research is a quasi experimental study. The population is the all of students of Santo Yoseph Private Middle School Medan grade VIII. The instruments used in this study are test mathematic problem solving abilities and questionnaire for SRL. The results of the study are there is difference in students mathematic problem solving abilities and SRL taught through geogebra assisted contextual learning models with students taught through geogebra assisted cooperative learning models and there is an interaction between learning and students' initial mathematical abilities towards students mathematical problem solving abilities and SRL.

The objectives of this study are to describe: 1) validity of learning tools developed based on the Realistic Mathematics Education approach; 2) practicality of learning tools developed based on the Realistic Mathematics Education approach; 3) the effectiveness of learning tools developed based on the Realistic Mathematics Education approach; 4) improvement of students' mathematical problem solving skills using learning tools based on the Realistic Mathematics Education approach that has been developed. This research is a development research. The development model used to develop learning tools in this study is the Dick and Carey development model. Before the field trial was conducted, the learning tools were validated by three experts and two practitioners. From the results of the validation, it was found that: (1) the learning tools developed were valid with the total average validity of student books (4.33), teacher books (4.28), learning implementation plans (4.42), and student worksheets (4.44). Subjects in this study were seventh grade students of senior high school (SMP) 1 Medan which consisted of two classes, namely VII-Herodotus as a trial class I and VII-Pythagoras as a trial class II. (2) learning tools that had been developed are effective, seen from classical learning completeness are fulfilled, in which ≥ 85% of students were passed, student activities within the specified tolerance limits, and students' responses to learning are in good categories; (3) the average increase in students' mathematical problem solving ability from trial I to trial II was 0.18 points, namely from 3.02 in the first trial to 3.20 in the second trial, with an increase in classical mastery learning 14%.