ABSTRACT
The objectives of research were to find out 1) which one provided better mathematics learning achievement, DL, PBL or Direct learning model, 2) which ones had better mathematics learning achievement, those with high, those with medium or those with low independency, 3) in each learning model (DL, PBL and Direct), which ones had better mathematics learning achievement, those with high, those with medium or those with low independency, and 4) in each independency level of students, which one provided better mathematics learning achievement, DL, PBL or Direct learning model.
This study was a quasi experimental research with a 3 x 3 factorial design. The population of research was all of the 8th graders of Junior High Schools in Banyumas city in the school year of 2014/2015 using KTSP curriculum. The sample was taken using stratified cluster random sampling. The sample of research consisted of 300 students: 102 for the 1st experiment class, 98 for the 2nd experiment class, and 100 for control class. The instruments used for collecting data were mathematics learning achievement test, student learning independency questionnaire. Before used for collecting the data, the instruments of achievement test and learning independency questionnaire was tried out first. The assessment on content validity for test and questionnaire instrument was carried out by validator. The reliability test on test instrument was carried out using KR-20 formula, while the reliability test on the questionnaire instrument was conducted using Alpha Cronbach formula. The test variance and questionnaire internal consistency was obtained using Karl Pearson’s product moment correlation. Balance test was conducted using one-way Anava test, with α = 0.05, it could be concluded that the experiment and control groups were in balance. Prerequisite test included normality test using Lilliefors test method and homogeneity test using Bartlett method. With α = 0.05, it could be concluded that the sample derived from normally and homogeneously distributed population.The hypothesis testing in this study was carried out using a two-way anava test with different cells. Furthermore, this study undertook post-anava advanced test using Scheffe method.
Considering the result of research and discussion, the following conclusions could be drawn (1) the students treated with DL and PBL learning Models provided better learning achievement than those treated with Direct learning model. The students treated with DL had mathematics learning achievement equal to those treated with PBL. (2) The students treated with high learning independency and medium learning independency provided better learning achievement than those treated with low learning independency. The students with high learning independency had learning achievement equal to those with medium independency. (3) In each learning independency group of students, DL and PBL learning models provided better achievement than the Direct one, and DL learning model provided learning achievement equal to the PBL learning model did. (4) In each learning model, the students treated with high learning independency and medium learning independency provided better learning achievement than those treated with low learning independency. The students with high learning independency had learning achievement equal to those with medium independency.
Keywords: DL Learning Model, PBL Learning Model, Direct Learning, Learning Independency