Analysis of the population is a difficult thing to do because besides being very complicated and complex, it is also not time and cost effective. This is the background of inference statistics necessary and important to do. The concept of inference statistics allows the researcher to perform an analysis using data from a sample to estimate an unknown population parameter. Inference Statistics is divided into 2, namely Parametric and Non Parametric.
 
 

1. Parametric

 
It is a statistical inference method whose model determines the existence of certain conditions (assumptions) of the distribution of population data. The application of parametric methods in general is to predict or test unknown parameters from certain distributions that are considered in accordance with data conditions, for example the assumption of normal distribution that must be met in the regression model. In addition, when viewed from the amount of data, it is usually large, at least greater than or equal to 30 data.
 
2. Non Parametric
 
The term non-parametric was first used by Wolfowitz, in 1942. Non Parametric statistical methods do not require assumptions about the distribution of population data (data distribution is not known and do not need to be normally distributed). Other terms that are often used for non parametric statistics are distribution free statistics and assumption-free test. Usually, the data used in this method are not that large, less than 30 data.
 
 
 
There are several statistical analysis techniques that can be used both for parametric and non parametric:
 
a) Correlative
 
Correlative analysis techniques are used to determine the relationship or correlation between one variable and another. For example, in a simple linear regression analysis that looks for the relationship between the dependent variable Y and the independent variable X.
Examples for each method:
• Parametric: Linear regression, Pearson
• Non parametric: Spearman Kendal, Kai
 
 
 b) Comparative
 
Comparative analysis techniques are used to determine differences in values (parameters), for example the average of a group with other groups or to find out differences in the retirement age of employees of company A and company B.
Examples for each method:
• Parametric: ANOVA
• Non parametric: Mann-whitney, Kruskal-wallis
 
 
To understand, you can see the picture below which is a summary of what has been explained to show the types of statistics and analysis techniques.