 |
Data source Strength Weakness
|
|
|
|
Data source Strength Weakness
· Targeted focus on case study phenomena
• Allows for the voice of the researched to be heard in their own words
• Insightful – provides perceived causal inferences
Questionnaires · User anonymity
• Practical – they are a fairly simple way to gather large data sets
• Scale – able to cover a large sample
• Systematic and quantifiable
• Generalizable – able to generalize findings to a wider population
· Incomplete recollection and/ or interviewee unwilling to explore depth of issues
• Can be less representative due to sample sizes
• Reflexivity – the interviewee expresses what the interviewer wants to hear
· Bias due to questions posed
• Cost of administration and data analysis
• Response rates often difficult to secure
• Often difficult to provide context and meaning (they often count incidence and prevalence rather than research the dynamic processes)
Documents and artefacts
· Stable repeated review
• Unobtrusive, exist prior to case study
• Exact names etc.
• Broad coverage – extended time span
• Insightful into cultural features
• Insightful into technical operations
· Retrievability difficult
• Biased selectivity
• Reporting bias, author bias
• Access may be blocked
• Privacy may inhibit access
• Selectivity, availability
Content analysis · Able to capture meaning
and terminology through assessment of words, sounds and images
• Focuses on communication and thus social interaction in an unobtrusive manner
• Cheap to administer
• Easy access to material
• Systematic and reliable
Source: Adapted from Yin (1993) in Tellis (1997)
· Biased selectivity
• Reporting bias, author bias
• Difficult sometimes to assess context appropriately
• Can lack a theoretical underpinning
• Time consuming, especially relating to coding and inputting of data
And, do not forget, as Vicky Heap and Jaime Waters detail in Chapter 5, mixed methodologies and methods are also appropriate means through which criminologi- cal research can be undertaken; the key benefit of using a mixed approach is that by delivering qualitative and quantitative methods alongside each other, the problems of one method can be ameliorated by the use of another.
Access and sampling
Blaxter et al. (2006) identify that in choosing and then refining your research approach during the planning stage you will need to give some thought to issues of access. You may require access to:
· people: in the community, their place of work, the home, particular groups
· organizations or institutions: government/public departments, public institutions, private organizations
· documents: held in public or private institutions, libraries and archives
· data: as detailed above, not all secondary data will be made available to you and when it is, it may not be delivered to you in a form that you want.
|
|
|
Access may prove difficult to secure for you as a researcher. With regard to data or documentation, they may not be made available to you. If available, it may involve negotiating access in a format that is suitable and appropriate. It is also important to be aware of issues relating to data reliability, accuracy and availability. Just because the data has already been collected does not mean that it will be accurate or reliable. Certainly, you will have to think about how you will ensure the reliability and valid- ity of the data that you wish to use. Access to people and organisations will often need to be negotiated, primarily through face-to-face meetings and written correspond- ence, which can be time consuming and not always successful, and in some cases additional security clearance may well be required. And you will need to think of the impact your research may have on people and organisations and how you will miti- gate that risk. Moreover, for all forms of access, you will need to think carefully about how long you will need access for, and why, and in doing so, you should think about whether you will be able to secure access in the time that you have. That is, securing access can sometimes be a very long and drawn-out process.
Access also involves making decisions about sampling, and selecting a sample will
be dependent on your research design:
· A quantitative approach determines that a statistical level of confidence is required. It is crucial in maintaining confidence and rigour in the findings and you should think carefully about the number of respondents you can attract and pro- vide adequate time to analyse the data that you secure. You may wish to consider having between 50 and 150 respondents for an undergraduate dissertation.
· If your approach is qualitative, then a richer understanding of the issue is required and a small number of respondents is acceptable. In qualitative studies, a relatively small sample can produce great diversity, detailed information and rich descrip- tions. Lincoln and Guba (1985) estimate that a relatively small sample can reach saturation point (where the greatest possible amount of data or information is obtained). You may wish to look at having between 4 and 10 respondents for an undergraduate dissertation.
There are various approaches to sampling, some more appropriate to a quantitative approach and some more to a qualitative approach, and as part of planning your research, and connecting your research questions to how you will do it, you must design and develop a sampling strategy. Kumar (2005: 169) considers the aims of selecting a sample to be:
· ‘to achieve maximum precision within a given sample size
· avoid bias in the selection of your sample’.
Sampling strategies can be identified in two broad categories: probability sampling which involves randomization; and non-probability sampling that involve non- randomization. Box 2. 7 identifies some of the more familiar sampling strategies.
 |
|
|
|
