Imagine that Desmond, Papachristos, and Kirk (2016) had focused their study on variation in crime reporting rates across Black communities in relation to the proportion of active duty police officers in the population of those communities. Then imagine that they found that communities with more police officers had lower levels of crime reporting and so concluded that “police interactions with citizens reduce their likelihood of reporting crime.” What type of error in reasoning could the authors have been making in this hypothetical example?
a. mechanism failure
b. ecological fallacy
c. reductionist fallacy
d. randomization
e. spuriousness
