top of page

The Cambridge Randomiser 2.0

The Cambridge Randomiser 2.0, developed and maintained by Ben Linton from the Metropolitan Police Service, allows users to create -- free of any charge -- random assignment sequences, with up to four treatments and a control, and up to 15,000 randomisations.  

Once set up, administrators can grant free access to an unlimited number of users who can login to the randomiser and receive a randomisation.

Resources: Text
Resources: News & Resources


Experimental criminology is the use of advanced experimental methods to answer key questions about the causes and responses to crime.

  • How much crime does prison prevent or cause for different kinds of offenders?

  • Does visible police patrol prevent crime everywhere or just in certain locations? What is the best way for societies to prevent crime from an early age?

  • How can murder be prevented among high-risk groups of young men?

For a 6-minute video presentation by Cambridge University Lawrence Sherman describing one such experiment, please see this Cambridge Ideas Video.

These and other urgent questions can be answered most clearly by the use of a research design called the "randomized controlled trial." This method takes large samples of people--or places, or schools, prisons, police beats or other units of analysis--who might become, or have already been, involved in crimes, either as victims or offenders. It then uses a statistical formula to select a portion of them for one treatment, and (with equal likelihood) another portion to receive a different treatment. Any difference, on average, in the two groups in their subsequent rates of crime or other dimensions of life can then be interpreted as having been caused by the randomly assigned difference in the treatment. All other differences, on average, between the two groups can usually be ruled out as potential causes of the difference in outcome. That is because with large enough samples, random assignment usually assures that there will be no other differences between the two groups except the treatment being tested.

Do you have useful resources to share with our community?

Resources: Text
bottom of page