# Histogram with # on edge of bin?

1. Oct 2, 2013

### doubled

let's say I have 10 data points. And one of my data points is 270.

Then let's say two of my bins are 260-270 and 270-280. Which bin would you put the 270 in?
Or would such a choice of bin range be inappropriate and new ranges have to be chosen?

2. Oct 2, 2013

### Office_Shredder

Staff Emeritus
When you construct your bins they should be literally non-overlapping - how you decide to do this is up to you. If your values are all integers, then most people would probably think it's more natural for the bins to be 260-269 and 270-279, if your values are arbitrary real numbers then the probability you pulled an integer is zero and you should be rethinking what the heck is going on (or more likely just arbitrarily picking whether everyone rounds down or up).

I have seen histograms often described as having intervals for example as 259.5 to 269.5 and 269.5 to 279.5 in order to encompass the same integer range, but look like there aren't any gaps.