It means they had to sift through 9 minorities to find 1 minority guilty party, and 9 whites to find 1 white guilty party.
The only difference - the volume of sifts was much more for minorities.
Here's a way to look at it:
100 whites are stopped, and of those, 11 are arrested (11%)
900 minorities are stopped, and of those, 99 are arrested (11%)
Same 11% arrest to stop ratio, 9x as much volume for minorities
Another way to look at it:
Whites were stopped at a rate equal to 28.5% of what you would expect given their population, and arrested at 31.4% of that same rate.
Blacks were stopped at a rate equal to 239.1% of what you would expect given their population, and arrested at 208.7% of that same rate.
Hispanics were stopped at a rate equal to 114.3% of what you would expect given their population, and arrested at 128.6% of that same rate.
Asians were stopped at a rate equal to 25.0% of what you would expect given their population, and arrested at 33.3% of that same rate.
The statement that minorities are 9x as likely to be stopped comes from the fact that 10% of arrests are of white people, and 90% of arrests are of minority people. I think they are counting Asians as minorities for this study. Whites and Asians make up 47% of NYC's population, but comprise only 13% of stops and 15% of arrests.
http://www.nytimes.com/imagepages/2010/05/13/nyregion/13frisk-gr.html?ref=nyregion