Non-Parametric Tests in Statistics

Spread the love
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  

In statistics, Non parametric tests test which does not make any assumption as to the form of distribution in the population from which the sample

is drawn i.e to say that the functional form of the distributions is not known. The test is called non parametric tests or distribution free test.

Assumptions of Non parametric test

Non-parametric tests follow some assumptions. They are,

  • Sample observations are independent
  • The variable under study is continuous.
  • The pdf is continuous.  
  • Lower order moment exists.

Advantages 

There are some advatages of non parametric test,

  • Non parametric methods are readily comprehensive very simple and easy to apply and do not require complicated sample theory.
  • No assumption is made when the sample is drawn from population.
  • Non- parametric tests are available in deal with the data which are given in rank.

Disadvantages 

  • It can be used only if the measurements are nominal and ordinal even in that case if a parametric test exists it is more powerful than non-parametric test.
  • Non Parametric tests are designed to test statistical hypothesis only and not for estimated the parameter.
  • A large number of different type of table is required.

The steps for testing procedure of Non-parametric test

Step 1 : Hypothesis test

State the null hypothesis  (Ho) and it’s alternative hypothesis (H1)

Step 2 : Statistical test 

Among several test which might be given research design, choose that test the which most closely approximates the conditional research in terms of the assumptions the test based.

Step-3 : Significance level 

Specify a level of significance (α)  and size (n )

Step-4: Sampling distribution 

finding the sampling distribution statistic test under the assumption that Ho.

Step 5 : Critical Region 

In the basis of step 2,3 and 4 above the region of rejection for the statistic form

Step 6 : Decision 

If the value of the test statistic is on of rejection, the decision is to reject wise accept.

Difference between Paramettric and Non-parametric test

Parametric Test

Non- Parametric Test

     Normality of the distribution

  Non- normality of the distribution

     Homogeneity of variance  

 Homogeneity of variance

     Scale of measurement is interval.

 Scale of measurement is ordinal and nominal. 

     It depends upon on parameter of population.

 It does not depends upon or parameter of population.

     It is not always easy to apply.

 Easy to apply and does not need complicated sample theory. 

     It makes assumptions as to the form of distribution to population.

 It does not make assumptions as to the form of the distribution to population.

     Here assumptions are too strictly.

Here assumptions are less strict. Such : continuity, discreteness, symmetry etc. 

     Here, Outlier exists.

Here Outlier does not exists.


 

Leave a Reply

Your email address will not be published. Required fields are marked *

Releated

Wilcoxon Signed Rank Test: Step by Step Procedure

Spread the love
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  

Spread the love          The wilcoxon signed rank test determine whether the median of the sample is equal to some specified value. The Wilcoxon signed rank test utilizes both the sign of the difference as well as the magnitude of the differences, hence it provides more information than the sign test, it is often more powerful non-parametric […]

Run Test in statistics: Step by Step procedure

Spread the love
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  

Spread the love          In many situation, we are interested to whether the data collected for statistical analysis is random or not. In most all statistical inference we assume that the data is in a random sample. If the randomness of a sample is suspected, we test it’s randomness before, going to statistical analysis. In detail, There […]

Sign Test: Step by Step calculations for small and large sample

Spread the love
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  

Spread the love          The sign test is the oldest of all non-parametric tests. and it is so called as usually covert the data for analysis to a series of plus and minus sign. The test statistic consists of either the number of plus sign or the number of minus sign. According to wikipedia, “The sign test is […]