
You are doing research and you have gathered data in order to prove something. For example higher advertisment results into higher sales or certain type of diet helps to build weight.
But wait, how do you do that? You can’t just put down your answer based on your understanding. You have to prove it by doing statistical tests in order to validate and prove them to others.
All data can be categorised into 2 types.
- Numerical Data
- Categorical Data
Numerical Data :- Data which is in form of numbers are numerical data. this can be further divided into
Categorical Data:- Data which is not numeric and qualitative in nature is called as categorical data. It can be further classified into
Deciding which type of test to be done depends upon the type of variables you want to compare.
1. Numerical to Numerical
Suppose you want to find relationship between height and weight, both are numerical in nature. You can simply use a correlation matrix to find the data. Correlation values ranges from – 1 to 1. Where -1 is perfectly negatively correlated and +1 is perfectly positively correlated, a correlation of 0 would signify no correlation.
2. Categorical to Numerical
Suppose you want to prove there is difference between Male’s and Female’s height. You can use students t-test or if there are many variables to compare you can use ANOVA.
3. Categorical to Categorical
Suppose you want to compare relationship between Preference for Ice cream flavour and Gender, you can use Chi Square rest.
This are some of the test which you can do to find relationship but there are many more you can try.







