Are more methods and types of data better than one?
- How to enter the group
- How to leave the group
- Whilst you're spending time with the people, how will you find an unbiased meaning?
- 'Going native' - when the researcher takes part in criminal acts of the group
- Specific look/personalities needed
- Covert - Can't take a notebook
- Experimenter bias
control variables e.g. weather
not true environment
people don't want to be tested on/studied
- Types of question (open/closed) - may lead answer to what researcher wants
- Training interviewers (especially unstructured)
- Gender of interviewers e.g. men interviewing female prisoners
- Time consuming (especially unstructured)
will people want to participate?
may affect sample size
- Interviewer bias
- Ethnicity of the interviewers e.g. will black offenders talk openly to white interviewers?
Lose track of them (change address/pass away)
How accessible is the secondary data?
Permission need to be obtained?
Will data be interpreted in the way they wanted?
Some sources are written to be published therefore they can give a biased view
Mostly middle-aged women answer
Quick to send
Low response rate e.g. Jackson and Sunshine 13% response rate
People may lie e.g. under exaggerate their crimes for fear of reprimand
Karl Popper and the…