Questions for ‘What happens when you look at crime by the numbers’


To understand who is likely to be a victim of crime — and where — many law enforcement officials are turning to the numbers. They are analyzing past reports of what has happened.


To accompany feature ‘What happens when you look at crime by the numbers’


Before Reading:

  1. Communities are doing a lot to fight crime. What measures have you heard about in your community? Does it seem to be working?
  2. What groups collect data on crime? Do you trust the numbers they’ve been coming up with?

During Reading:

  1. According to the story, what are some examples of possibly faulty or flawed data?
  2. What kind of crime did Kristian Lum focus on in Oakland, Calif., and what did her computer model conclude?
  3. What did Lum conclude about the biases in the data fed into her model?
  4. Give some examples of why some illegal activities may not get picked up in a region’s crime statistics (based on the story).
  5. What is PredPol and what did it conclude about the race of the criminals being analyzed in Oakland?
  6. What did Reid Amin conclude from his study on violent crime hotspots about the types of people responsible?
  7. Lum says to understand crime, analysts shouldn’t rely only on data collected by police departments. What other groups does she recommend turning to for useful information?
  8. What community of people did Greg Ridgeway study in New York City?
  9. Which people were more likely — or less likely — to shoot, based on his findings?
  10. Why did Ridgeway take issue with a “double-digit” quote by President Trump? Why did he find the statement misleading?

After Reading:

  1. Come up with three other types of data that are collected to better understand people or their behavior. Brainstorm what kinds of limitations may exist in those data sets, based on the issues raised in this story.
  2. Some people argue that it’s easy to “lie with statistics” — essentially by misleading people as to what the numbers mean or neglecting to include information that would influence how you would interpret trends from the numbers. Create an example where you do that. Then brainstorm what type of questions others should ask you to figure out if  the statistics you had offered might be misleading.


  1. According to a Sept. 2016 report by the Federal Bureau of Investigation, there were 1,197,704 violent crimes across the United States in 2015. The FBI said this was an increase of 3.9 percent. Then how many crimes must there have been the year before?
  2. The FBI reported that the estimated number of U.S. murders in 2015 was 15,696. What share (on a percentage basis) was that of all violent crimes that year?
  3. The FBI reports that “firearms were used in 71.5 percent of the nation’s murders.” How many murders did that come to in 2015?