The problem with non-peer reviewed articles is that you don't know how they selected the data, etc. Scientific research requires you set your hypothesis before

The problem with non-peer reviewed articles is that you don't know how they selected the data, etc. Scientific research requires you set your hypothesis before you get the results. Clearly this didn't happen in this case. Why did Telegraph chose to use this particular metric ("crimes solved using CCTV") rather than for instance "crime rate change as a result of CCTV" (which would take into account deterrence). Why did they choose to use total number of CCTV cameras rather than the number of active CCTV cameras ? One of the points of peer review is spot biases like this. I'm not saying that CCTV works (or that it doesn't) but newspapers or non-peer reviewed research often falls far below an acceptable standard for evidence.