DI Question, relating to a table that adds to > 100%

[arielle]

Approximately how many people, in thousands, in the U.S. population had private health insurance in 2004?
181,262
185,983
191,230
203,816
245,254

There are three tables to look at for this set of questions, however only two are relevant for this question. First, it's pretty straightforward how to get the US population from this question from table #1, it's approximately 291,000 (in thousands).

The confusion comes from the next step, interpreting what % of the US population has private health insurance. The third table is titled: U. S. Population Health Insurance Coverage Type Percentages: 1999 to 2008
It has three main titles of Private Health insurance (2 subtitles listed beneath that), Gov't Health Insurance (3 subtitles listed beneath that), and Not Covered (no subtitles underneath).
Theoretically, the question is very straightforward (it is a 200-500 level question) as it just asks you to add the two %'s of the subtitled columns beneath private health insurance (which adds to 70%) and then multiply that % by the total US population in 2004 (291,000).
However, if you add the %'s across all of the columns it comes out to approximately 115%, which does not make any sense UNLESS we are to ASSUME that various people are covered under multiple options? If I see a similar type of problem on the real GRE should I just go ahead and assume this assumption is valid? Thank you for your help.

[jen]

I think that is a reasonable assumption. Pie charts should always add to 100% unless otherwise specified, but bar charts can describe situations in which people fall under more than one category. (I'm not surprised, though, that the question didn't depend on the "overlap.")

PP133-142 of the Official Guide to the GRE Revised General Test contains math conventions that answer some questions like this. I checked and didn't see this issue addressed, but you might want to check out that section for other questions about "ground rules" for the GRE.

Jen