INTRODUCTION

For testing causal models, it is well agreed that experimental research is the gold standard. For example, Farrington (1979) calls experimentation the most conclusive methodology for testing hypotheses. Despite its key role in the prevention of and protection against crimes and renewed interest in the concept, experimental tests of guardianship remain scarce. In their article on guardianship for crime prevention, Hollis-Peel et al. (2011) found that most of the research on guardianship makes use of macrostructural and survey data, relying on cross-sectional and nonobservational approaches. The authors point out that there is a distinct lack of, and subsequently a specific need for, quasi-experimental and experimental designs in guardianship research.

Following the call for more experimental research on guardianship by Hollis-Peel et al. (2011), the aim of this chapter is to explore in more detail if experiments would be useful in guardianship research. We look into experimental research in the realm of guardianship and review the literature. In the course of this chapter we propose to refine the concept of guardianship, distinguishing guardianship before a crime takes place (“preventive guardianship”) from guardianship during or after a crime (“repressive guardianship”). The distinction between phases of guardianship made in this way has, of course, consequences for the way experimental research into guardianship can and should be designed. We observe that experimental work into repressive guardianship, of which a number of examples in the 1970s and 1980s can be found, raises concerns about its validity, whereas experimental work on preventive guardianship is altogether absent.