Carcass wrote:
OE
Rather than calculating, think in terms of the slope of the line connecting successive data points for assault. A greater rate of increase means a more sharply sloping downward line: The line from 2006 to 2007 is steeper than the one from 2005 to 2006, so choice (B) is correct, and the line from 2008 to 2009 is steeper than the one from 2007 to 2008, so choice (D) is also correct. For all the other years, the decrease in assaults is less than the previous year, so only choices (B) and (D) are correct.
What about 2006? 2005 to 2006 decrease is also tangible as against 2004 to 2005 - which hasn't been specified.