Traditional Statistics (Descriptive & Inferential)

 


There are two types of statistics: descriptive statistics and inferential statistics. Each of these components is important because it offers varied techniques to reaching various goals. Descriptive statistics describe what happens in a population or when data is collected. In contrast, inferential statistics allow scientists to extrapolate findings from a small sample to a larger population. There are some key differences between the two types of statistics.

 

DESCRIPTIVE STATISTICS

When most individuals hear the word "statistics," they immediately think of descriptive statistics. The goal of this statistical field is to characterise. To define the features of a set of data, numerical measures are utilised. This statistics area comprises a range of items, including:

  •          The average, or measure of the centre of a data collection, consists of the mean, median, mode, or middle.
  •          The spread of a data collection, which may be measured using the range or standard deviation.
  •          Descriptions of overall statistics, such as the five-number summary
  •          Two examples of measures are skewness and kurtosis.
  •          The investigation of data links and correlations.
  •          The graphical representation of statistical results

These measures are important and helpful because they allow scientists to see patterns in data and hence make sense of it. Descriptive statistics may only be used to describe the population or data set under consideration: The results cannot be generalised to any other group or population.

 

TYPES OF DESCRIPTIVE STATISTICS

Descriptive statistics are used by social scientists in two ways:

The mean, median, and mode, which are calculated and presented as measures of central tendency, show overall patterns within the data. A mean is the mathematical average of all of a data collection, such as the average age at first marriage; a median is the midpoint of the data distribution, such as the age in the centre of the range of ages at which individuals first married; and a mode is the most frequent age at which people first marry.

Spread measurements, which include the following, determine how data are scattered and connected to one another:

  •          A data set's range is the whole collection of values.
  •          The frequency distribution describes how frequently a certain value appears in a data set.
  •         Quartiles are data set subgroups formed by partitioning all values into four equal pieces over the range.
  •          The mean absolute deviation is the average of how much each result deviates from the mean.
  •          Variance, which indicates how much variety exists in the data.
  •          The standard deviation measures data dispersion compared to the mean.

Spread measurements are typically shown in tables, pie and bar charts, and histograms to aid with data pattern interpretation.

 

INFERENTIAL STATISTICS

Complex mathematical computations are employed to develop inferential statistics, which allow scientists to infer patterns about a larger population based on a study of a sample drawn from it. Inferential statistics are used by scientists to investigate the relationships between variables within a sample and then generalise or forecast how those variables will relate to a larger population.

It is usually hard to study each member in a group. So scientists choose a representative subset of the population, known as a statistical sample, and analyse it to learn more about the population from which the sample was drawn. Inferential statistics is classified into two types:

  •          By surveying a statistical sample, a confidence interval offers a range of values for an unknown population parameter. This is expressed as a range and the degree of certainty that the parameter falls inside it.
  •          By evaluating a statistical sample, scientists might make a claim about the population using significance tests or hypothesis testing. By design, this technique has a high degree of unpredictability. This might be expressed as a level of importance.

By surveying a statistical sample, a confidence interval offers a range of values for an unknown population parameter. This is expressed as a range and the degree of certainty that the parameter falls inside it.

By evaluating a statistical sample, scientists might make a claim about the population using significance tests or hypothesis testing. By design, this technique has a high degree of unpredictability. This might be expressed as a level of importance.

 

DESCRIPTIVE VS INFERENTIAL STATISTICS

Although descriptive statistics can help you uncover things like data distribution and centre, they cannot be used to draw broad statements. In descriptive statistics, measurements such as the mean and standard deviation are given as numerical values.

Even while certain formulae in inferential statistics are same, such as the mean and standard deviation, the focus is different. Inferential statistics start with a sample and work their way up to a population. This information about a population is not numerically stated. Instead, scientists portray these elements as a range of potential numbers accompanied by a level of confidence.


Descriptive vs. Inferential Statistics (thoughtco.com)

Comments

Post a Comment

Popular posts from this blog

Historical Development of Big Data

Image
Prior to 1980, corporations stored and analysed data on mainframe computers. Two important technologies had a critical role in the early creation of data centres as we know them today. The first was the advent of personal computers (PCs), which grew in popularity as Microsoft's Windows operating system became the global standard. Sun Microsystems' invention of the network system protocol, which allowed PC users to access network files, was the second. Following that, microcomputers began to occupy mainframe rooms as servers, and the spaces were nicknamed data centres. Several key events have affected the industry's evolution since then, most notably the advent of virtualization software and the shift to cloud computing. The timeline below demonstrates these and other major turning points in the history of big data.   In 1943, the United Kingdom developed the first data-processing computer capable of breaking Nazi codes. In 1945, ENIAC, the first electrical general-p...
Read more