There often seems to be a great deal of confusion over these terms, so this is an attempt to help clear it, with an example

Approximately 80% of the US population are Christians.  Approximately 6% are atheists.  This means that there are more Christians than atheists. 

For now, assume that a small group (such as a discussion group, or the group of prisoners in US prisons) consists only of Americans.

If such a group were to be proportionately represented by Christians and atheists alike, you'd expect to see about 80% Christians, and 6% atheists.

So, if there were about 3000 members in the group, then if the group population was proportionately represented as the US population, you'd expect to see about 2400 Christians, and about 180 atheists.

But if there were, upon observation, significantly more than 180 atheists in the group, while the number of Christians was as expected (or lower), we'd say that disproportionately more atheists can be found than Christians in that group, when considering the US population.