Application of Non-Parametric in education, research,
Eureka Education . In quantitative research, data analysis techniques can be used, is descriptive and inferential. If only descriptive data analysis techniques related to the collection, processing, analyzing and presenting data without some or all close with a level of significance. While inferential data analysis work on estimation (estimated) population parameters and testing hypotheses about the units of quantitative (statistics) is based on the sample. As is understood, data analysis techniques are inferential in two parametric and non-parametric. The purpose of this is parametric. Quantitative data collection through research samples to draw with the aim to draw conclusions about the population from which the sample was taken. Basically, the conclusions about a population in terms of the price estimate population parameters showed. Price statistics which is what we call the estimator or estimators. The accuracy of the estimate of the parameter estimates depends between the statistics and price parameters of the size of the gap can fulfill its function order parameter data analysis to perform required processed some requirements such as, data is the data interval, complying with the rules of normality and homogeneity. However, if one of those conditions is not satisfied, then the data analysis, the use of non-parametric.
in the early stages of learning Statistics in the study, nonparametric be can rarely studied. So that when the processing of research data, if they do not qualify parametrically, some novice researchers or students are not with the use of non-parametric analysis familiar. Here are some examples of non-parametric analyzes were used to test different in the case of educational research.
Some examples of the use of non-parametric is to look for different test against some groups or samples independently, as example: There are two class is used in a research experiment. A class is a class experiment using PBL combined learning methods with a concept map, while the other class is a class control with PBL course. The dependent variable (bound) in this research is understanding the concept. If researchers want to measure differences, understanding the concept of students before and (were used with their pre-test and post-test) after treatment, the data are the data pairs ( couple independent ). In parametric data analysis used the t-test (t test), whereas in the non-parametric used Wilcoxon rank test-mark .
Then, when the study to determine differences, aims to understand the concept of these two classes, parametric analysis used the t-test (t-test), but if the requirements are not met parametric, non-parametric data analysis was used Man Whitney test . In the case of the research, there are three groups or three classes will be tested (for example: grade 3 or groups of 3 using model Jigsaw learning), then, knowing the different understanding of the concept of the three classes or groups, the analysis of used non-parametric Kruskal Wallis .
0 Komentar