ABSTRAK
Tujuan penelitian ini adalah untuk menentukan: 1) manakah yang memberikan prestasi belajar matematika lebih baik antara siswa yang dikenai model pembelajaran kooperatif tipe TSTS dengan Concept Mapping, model pembelajaran kooperatif tipe TSTS, atau model pembelajaran ekspositori pada materi persamaan garis lurus; 2) manakah yang mempunyai prestasi belajar matematika yang lebih baik antara siswa dengan kemampuan penalaran matematis tinggi, sedang, atau rendah pada materi persamaan garis lurus; 3) pada masing-masing model pembelajaran, manakah yang memberikan prestasi belajar matematika lebih baik antara siswa dengan kemampuan penalaran matematis tinggi, sedang, atau rendah; 4) pada masing-masing kemampuan penalaran matematis, manakah yang mempunyai prestasi belajar matematika lebih baik antara siswa yang dikenai model pembelajaran TSTS dengan Concept Mapping, model pembelajaran TSTS, atau model pembelajaran ekspositori.
Penelitian ini merupakan penelitian eksperimental semu dengan desain faktorial 3x3. Populasi dari penelitian ini adalah seluruh siswa kelas VIII SMP Se-Kabupaten Kebumen tahun pelajaran 2015/2016. Pengambilan sampel dilakukan dengan teknik stratified cluster random sampling. Instrumen yang digunakan untuk mengumpulkan data adalah tes prestasi belajar matematika dan tes penalaran matematis. Sebelum digunakan untuk pengumpulan data, instrumen tes dan angket telah diuji cobakan terlebih dahulu. Penilaian validitas isi dilakukan oleh validator. Uji reliabilitas instrumen tes prestasi dan tes penalaran matematis menggunakan rumus KR-20 dan daya pembeda menggunakan rumus korelasi momen produk dari Karl Pearson. Uji keseimbangan menggunakan uji ANAVA satu jalan. Uji prasyarat meliputi uji normalitas menggunakan metode uji Lilliefors dan uji homogenitas menggunakan metode Barttlet. Teknik analisis data yang digunakan adalah uji ANAVA dua jalan dengan sel tak sama.   
Berdasarkan uji hipotesis diperoleh kesimpulan sebagai berikut: 1) Prestasi belajar matematika siswa yang dikenai model pembelajaran TSTS dengan Concept Mapping lebih baik daripada prestasi belajar matematika siswa yang dikenai model pembelajaran TSTS dan ekspositori pada materi persamaan garis lurus, prestasi belajar matematika siswa yang dikenai model pembelajaran TSTS lebih baik daripada prestasi belajar matematika siswa yang dikenai model pembelajaran ekspositori pada materi persamaan garis lurus. 2) Prestasi belajar matematika siswa kemampuan penalaran matematis tinggi lebih baik daripada prestasi belajar matematika siswa kemampuan penalaran matematis sedang dan rendah pada materi persamaan garis lurus dan prestasi belajar siswa kemampuan penalaran matematis sedang lebih baik daripada prestasi belajar matematika siswa kemampuan penalaran matematis rendah pada materi persamaan garis lurus. 3) Pada model pembelajaran TSTS dengan Concept Mapping, prestasi belajar matematika siswa kemampuan penalaran matematis tinggi lebih baik daripada siswa kemampuan penalaran matematis sedang maupun rendah, dan prestasi belajar matematika siswa kemampuan penalaran matematis sedang sama baik dengan siswa kemampuan penalaran matematis rendah; pada model pembelajaran TSTS, prestasi belajar matematika siswa kemampuan penalaran matematis tinggi sama baik dengan siswa kemampuan penalaran matematis sedang, prestasi belajar matematika siswa kemampuan penalaran matematis sedang lebih baik daripada siswa kemampuan penalaran matematis rendah, dan prestasi belajar matematika siswa kemampuan penalaran matematis tinggi lebih baik daripada siswa kemampuan penalaran matematis rendah; pada model pembelajaran ekspositori, prestasi belajar matematika siswa kemampuan penalaran matematis tinggi lebih baik daripada siswa kemampuan penalaran matematis sedang maupun rendah, dan prestasi belajar matematika siswa kemampuan penalaran matematis sedang sama baik dengan siswa kemampuan penalaran matematis rendah, 4) Pada siswa kemampuan penalaran matematis tinggi, model pembelajaran TSTS dengan Concept Mapping, model pembelajaran TSTS, dan model ekspositori menghasilkan prestasi belajar matematika yang sama; pada siswa kemampuan penalaran matematis sedang, model pembelajaran TSTS dengan Concept Mapping menghasilkan prestasi belajar matematika yang sama dengan model pembelajaran TSTS, model pembelajaran TSTS menghasilkan prestasi belajar matematika yang lebih baik daripada model pembelajaran ekspositori, dan model pembelajaran TSTS dengan Concept Mapping menghasilkan prestasi belajar matematika yang lebih baik daripada model pembelajaran ekspositori; pada siswa kemampuan penalaran rendah, model pembelajaran TSTS dengan Concept Mapping menghasilkan prestasi belajar matematika yang lebih baik daripada model pembelajaran TSTS, model pembelajaran TSTS menghasilkan prestasi belajar matematika yang sama baik dengan model pembelajaran ekspositori, dan model pembelajaran TSTS dengan Concept Mapping menghasilkan prestasi belajar yag lebih baik daripada model pembelajaran ekspositori.
Kata Kunci: Two Stay Two Stray (TSTS), Concept Mapping, Model Pembelajaran Ekspositori, Kemampuan Penalaran Matematis.
ABSTRACT
The purpose of this research was to determine: 1) which one of TSTS with Concept Mapping, TSTS, and direct learning models that resulted in better students’ mathematics learning achievement on equation of a straight line; 2) which one of the students with the high mathematical reasoning ability, the moderate mathematical reasoning ability, or the low mathematical reasoning ability who had better mathematics learning achievement on equation of a straight line; 3) for each learning model, which one of the student with the high mathematical reasoning ability, the student with the moderate mathematical reasoning ability, or the student with the low mathematical reasoning ability that resulted in better mathematics achievement; 4) for each mathematical reasoning ability, which one of TSTS with Concept Mapping learning model, TSTS learning model, or direct learning model that resulted in better mathematics achievement.
This research was a quasy experimental research with 3x3 factorial design. The population of this research was all of the eighth grader students of Junior High Schools in Kebumen Regency in the academic year of 2015/ 2016. The sampling was conducted by stratified cluster random sampling technique. The instruments used to collect the data were the students’ mathematic test result and mathematical reasoning ability. Before being used to collect the data, the test were tried-out. Validity test was conducted by the validator. Instrument reliability test was conducted by using KR-20 formula and distinguishing ability by using product moment correlation by Karl Pearson. Balance test was conducted by one-way ANAVA test. Prerequisite test consisted of normality test by using Lilliefors testing method and homogeneity test by using Barttlet method. Data analyzing technique used were two-ways ANAVA test with different cells.  
Based on the hypothesis test, it was concluded as follows. 1) The mathematics learning achievement of the students who were treated by TSTS with Concept Mapping learning model was better than the mathematics learning achievement of the students who were treated by TSTS and direct learning method on equation of a straight line, the mathematics learning achievement of the students who were treated by TSTS learning model was better than the mathematics learning achievement of the students who were treated by direct learning method on function; 2) The mathematics learning achievement of the students with the high mathematical reasoning ability was better than the mathematics learning achievement of the students with the moderate and the low mathematical reasoning ability, and the mathematics learning achievement of the students with the moderate mathematical reasoning ability was better than the mathematics learning achievement of the students with the low mathematical reasoning ability on equation of a straight line; 3) In TSTS with Concept Mapping learning model, the mathematics learning achievement of the students who had the high mathematical reasoning ability was better than the students who had the moderate and the low mathematical reasoning ability, while the mathematics learning achievement of the students who had the moderate mathematical reasoning ability was same as the students who had the low mathematical reasoning ability; in TSTS learning model, the mathematics learning achievement of the students who had the high mathematical reasoning ability was same as the students who had the moderate mathematical reasoning ability and was better than the students who had the low mathematical reasoning ability, while the mathematics learning achievement of the student who had the moderate mathematical reasoning ability was better than the students who had the low mathematical reasoning ability; in direct learning model, the mathematics learning achievement of the students who had the high mathematical reasoning ability was better than the students who had the moderate and the low mathematical reasoning ability, while the mathematics learning achievement of the students who had the moderate mathematical reasoning ability was same as the students who had the low mathematical reasoning ability; 4) The high mathematical reasoning ability category students who were treated by TSTS with Concept Mapping, TSTS, and direct learning models had same mathematics learning achievement; the moderate mathematical reasoning ability category students who were treated by TSTS with Concept Mapping learning model had same mathematics learning achievement as the students were treated by TSTS and had better mathematics learning achievement than the students were treated by direct learning models, while the students were treated by TSTS learning model had better mathematics learning achievement than the students were treated by direct learning model; The low mathematical reasoning ability category students who were treated by TSTS with Concept Mapping learning model had better mathematics learning achievement than the students were treated by TSTS and direct learning models, while the students were treated by TSTS learning model had same mathematics learning achievement as the students were treated by direct learning model
Keywords: Two Stay Two Stray, Concept Mapping, Direct Learning Model, Mathematical Reasoning Ability.