The descriptive measure of some characteristic of population is referred to as Parameter. These characteristics include population mean(μ), population standard deviation(σ), population variance(σ²).
The descriptive measure of some characteristic of a sample is referred to as Statistic. These characteristics include sample mean( x¯ ), sample variance(s²), sample standard deviation(s).
Let’s begin with the knowledge of the most frequently yet confusing terms in statistics. Once you start dealing with data sets, you’ll come across these terms endlessly, hence you must know their true meanings:
Nominal Scale: When data are labels or names used to identify the attribute of an element, then nominal scale is used. In this(titanic) data set, name and sex represents nominal data. This type of data is not fit to be used in its raw form, unless numbers are assigned to them. Nominal scale is the lowest level of data measurement. Think of ‘nominal’ as ‘name’. It helps to remember.
Ordinal Scale: When the data can be sorted in an inherent order, then ordinal scale is used. This scale is used to rank or order objects. For example, if a data set can be ordered in Ranks (increasing or decreasing), where ranks are in numerical order, then it’s ordinal. Think of ‘ordinal’ as 'order. It helps to remember.
Interval Scale: When the data is numeric in nature, then interval scale is used. But, the difference
between two numbers should be meaningful. For example: 3 students scored 88, 78, 59 marks in their statistics test. The meaningful difference in these marks will help us know which student has the highest and lowest marks.
Ratio Scale: This scale possess all the properties of interval scale with meaningful ratios of two values. A ratio scale must contain a zero value that indicates absolute void at zero point. For example, if you say, I ran 0(zero) km today. This means you didn’t run at all. But, if you say, today’s temperature is 0(zero) degree, this means there exists a temperature which is too cold. Hence, temperature doesn’t have ratio scale.
Therefore, summing up,
Population: Population is commonly referred to collection of data used for statistical investigation to draw meaningful conclusion, which can later be used by companies for informed decision making. For example: 2200 people boarded the titanic ship. This number represents population.
Sample: A small representative of population is known as sample. For example, if you have download the data set, you would realize, you have only 891 values in train data. This small representation is the sample of people boarded titanic ship. Statisticians prefer to work on sample rather than population as it saves time, money and resources.