A study by Brower et al (1981) used interviews to establish views about different types of space. They found that people believed they would feel safer in homes that were well fenced and had markers such as house signs, plants and cars visible, than in homes with no clear territorial markers. This again shows the importance of defensible space in combating crime.
Ley (1974) carried out observations of delinquent boys in institutions to see how territory affected aggression. In this study it was found that when territory was clearly established and controlled (dominant boys controlled the best areas in the institution) there were few aggressive incidents. However when the dominant boys were transferred and replaced, this disrupted the territorial control and aggressive incidents rose, after a few weeks territory was re-established and aggression fell again. This suggests the importance of established areas of territory for reducing aggression and giving a sense of order.
The first evaluation issue is ecological validity.
The study by Smith et al (1981) showing that gender affects the amount of territory claimed was an observation. Observational research has a high degree of ecological validity, however observers can be affected by their own biases and some behaviours may be missed during the observational period. Ethical guidelines state that observational research should only be carried out in public places in order to protect privacy. The study was carried out on beaches and it may be that the behaviour of the males cannot be generalised to other situations – further research would be needed. Mercer’s research however does support the finding of Smith that males require larger territories than females. This study used a questionnaire design which may provide socially desirable or dishonest responses. In this case the participants were students and it could be possible that the male students claimed they regarded more space as theirs in an act of bravado. In addition since these participants were students the results may again not be generalisable to other populations. These methods of study are non experimental and it is therefore not possible to show cause and effect.
Similarly Newman used a correlation to interpret the findings from his study and found a relationship between undefended space and levels of crime. Correlations cannot show cause and effect, therefore other causes of these findings cannot be ruled out. One other possible explanation may have been that the different estates compared by Newman were simply in high or low crime areas, or that policing tactics of the areas was different.
However in contrast Brower et al (1981) used interviews to try and establish how people felt about areas of defensible space. Interviews are a good way of obtaining rich and detailed data, however unlike Newman’s data which was factual (recorded crime figures) interview data is qualitative and needs to be interpreted by the researcher. Inevitably the way this is done may be influenced by the researchers views and therefore biased. In addition to this, as in the research by Mercer, there are issues relating to the honesty of people interviewed and providing socially acceptable responses to be considered.
The research by Ley was an observation and therefore, as in Smith’s observation, this research could be argued to have greater ecological validity than research using laboratory methods, however there are problems again of observer bias affecting what is recorded. Also this study used institutionalised delinquents as participants and therefore it is difficult to generalise the results to any other situations, although the findings do have implications for institutions. Ethical issues may be particularly relevant in this study as carrying out observations of institutionalised participants may be a breach of their right to withdraw from the study. It could also be argued that if the observations were carried out by a researcher their presence may have affected the behaviour, on the other hand if video cameras were used covertly there may be a problem in ensuring that all behaviour was observed, in addition to the ethical issues mentioned.
Designers and architects could use the research mentioned to ensure that working areas meant for males and females have varied sizes in order to take account of different gender needs for space. When designing housing estates architects need to ensure that the semi public areas are defensible in that they are overlooked and have markers to suggest ownership. They should also include barriers and fenced areas around homes and use plants and foliage as markers.
In institutional design it would seem important to use design to clearly mark out areas in order to try and create established boundaries and reduce aggression caused by disorderly space use. It may be possible to include specific time slots for prisoners to use semi public areas to help reduce any aggression caused by dominance of most desirable areas.