whether or not the man had moved to the place of current residence
since 2003; and (3) if he had not moved since 2003, the
number of units in his current building of residence.
We limitedthe sample tomen who:hadsmokedatleast 10 years
to make it more likely that the person’s lung cancer was caused by
smoking; had not moved since 2003 to ensure that the individual
had resided in the location sufficiently long to be known to neighbors;
and lived in a building with at least 15 units for there to be
a sufficiently large pool of neighbors. There are in total 255 persons/units
diagnosed with lung cancer that satisfied our inclusion
criteria.
Finally, our screener survey revealed that five persons who lived
in four communities had been diagnosed with lung cancer outside
the 2011 time window.6 To make the information intervention sij
correspond more closely to persons diagnosed with lung cancer
during the 2011 time window, wedroppedthese four communities,
yielding a net sample of 251 communities.