- Thread starter emekadavid
- Start date
- Tags histogram

I have not heard of these terms. Are they for skewed data and putting a ceiling or floor on them?

1) Create a bunch of bins

2) Count how many observations fall into each bin

3) Draw a histogram from that

The question the OP is asking is does it matter if when an observation falls on a bin edge (so maybe the first bin is 1.5-2.5 and the second is 2.5-3.5 and you have a data value of exactly 2.5) does it matter if you put it in the lower bin or the upper bin.

What will you be using the histogram for in particular?

Histograms are generally just used to get an idea of what kind of common distribution the data follows (e.g. normal, skewed). If the histogram is only a little off from a common distribution, then it shouldn't really matter if the histogram is left open or right open. It will still only be a little off.