Let’s take an example. We are given a dataset of Sex Ratio of all the states of the India. In this case, Central Tendency is a single
number which denotes the “Centre” of the data. We can say that the mean of the sex ratio of all states is 940. We can also say that the Median of sex ratio of all Indian states is 946.6. Here 940 or 946.6 is a single number which represents the Centre or Location of our dataset and is known as Central Tendency. While the central tendencies are good for describing the location of data, it fails to tell anything about the shape of data. That is, how disperse or how close the data points are in a distribution. Central Tendencies are
measured mostly using
- Mean – Calculated Average
- Median – middle value
- Mode – most occurring value
A Mean is a sample mathematical average of data points.
In our previous example, we may ask the following question –
- What is the average sex ratio of all the states of India?
- (N+1/2)th term, if N is ODD
- ((N/2 th) term + (N/2 +1)th term)/2
Mean Vs Median
Mean doesn’t represent correct picture when the data has outliers, or the distribution is skewed.
For e.g., let’s pocket money of 5 students in thousand rupees are 5,7,8,12. Then the average pocket money is 8k. But when an outlier say a very rich student say Mark whose pocket money is 100,000k joins them, then average becomes almost 20,000k. This is gross misrepresentation as 5 out of 6 students has very little pocket money compared to Mark. But, Median pocket money in this case is 8k which makes sense.
The median gives a better idea of central tendency in case of skewed distribution.
The following diagram summarizes Mean Vs Median with another example.
The mode is the most frequent score in our data set. Let’s say our data-set has following data-points.
In this case 3 occurs the most number of times and hence 3 is the mode of our data-set. Mode is used when we want to know the most likely outcome in our data-set. Let’s say we are dealing with categorical data. For e.g. religion of people in India and we know that 80% people living in India are Hindus. Hence Hindu become the mode of our data.
The following video deals with measures of central tendency in detail.