Exploring the Best Method for Finding the Mode in Statistical Data

[kay]

till now i used the normal method for finding out the mode of a given data that is just simply look for the most frequently occurring observation and label it as the mode. But recently I have encountered another method for finding out the mode in which it was also stated that my old method for finding out the mood was incorrect. What this new method involved was that we grouped all the items in 2s and 3s and using this method we basically found out around which observation did all the other observations concentrate around.
So my question is that why is this grouping method used for finding out the mode can't we just look at the highest frequency observation and tell that this is the mode?

 

[PWiz]

For continuous data, the modal class is the one with the highest frequency density, not the highest frequency, because the original values are lost, and to account for that we need to estimate the frequency of the original discrete entries by dividing the frequency with the class width.

 

[statdad]

If you have raw data - a list of individual values - a mode is a value which occurs most often. But - if there are no duplicate values, you won't have a mode. In that case it is sometimes suggested to count how many items of data are in convenient groups (0 to 9, 10 to 19, 20 to 29, as an example where all the values are between 1 and 30) and see which group contains the most data values. If it is the 10-19 group, take the mode to be the midpoint: 15