مقارنة بين بعض الاختبارات المعلمية و اللامعلمية في القطاعات العشوائية الكاملة (دراسة تطبيقية)

Compression of some parametric and nonparametric tests in the Randomized complete Blocks (Applied study)

The hypotheses assumed by the researcher and the  cabability to improve their correction are considered as the important thing that discussed by the scientific research. Thus, their importance apper in the experments which have an important rule in the scientific and technology progress. The statistics and epically The experimental design has a main rule in this research’s and study’s concerning with hypotheses testing in the design and analysis experiments . The concept of hypotheses testing concerned with the concept of statistical inference which started with estimation of parameters which belong to the population of study depending on the sample from this population , After that test the hypotheses  about the estimated parameters and how much are fit to the population parameters . This method of testing is called a parametric testing which are depend on some of assumption same as the assumptions of analysis of variance ( Normality, Constant variance , and Independent variables ).When we loss one of these assumption we will use another type of testing called nonparametric or Free- distribution  testing . Therefore, the main objective of  these study to introduce a comparative study between the parametric and nonparametric testing in the randomized complete design by using the two way ANOVA and using F-test to represent the parametric test and Friedman and Quade to represent the nonparametric test. On the other hand, the application consist areal data in the health field, including sample with (540) cases of (Abortion). The aim is to study the effect of four main factors (baby weight, time of pregnancy, age of mother and Mother’s  job  ) on the (Abortion), Where we classify each factor into many levels to study the effect of each two factors on (Abortion). Consequently we compose Six experiments which are the combination of these four factors. After that , We test the assumption of ANOVA, Then we find that we loss the assumption of normality for the factors. When we test the assumption of  constant variances by using the levene test, then the results show that the significance of homogenous of variances is differ from experiment to other .The results  of F-test show the significance differences of these factors from experiment to other which invite us to conduct the multiple comparisons by using Duncan method . Also, the results of the nonparametric test show the significance differences for treatments and blocks for all experiments.