PSC/SOC 340/JS 504: Social Science Research Methods

To decide whether survey research is a good design choice for you, consider the circumstances in which surveying respondents will be an effective way to address your research questions.

The following are among the most important criteria for deciding whether a survey is a good design for your research. Surveys are an effective research design when:

1. The data are best obtained directly from the respondents;
2. Your data can be obtained by brief answers to structured questions;
3. You can expect respondents to give you reliable information;
4. You know how you will use the answers; and
5. You can expect an adequate response rate.

Vogt, W. Paul (Author). When to Use What Research Design. New York, NY, USA: Guilford Press, 2012. p16.  http://site.ebrary.com/lib/spivalib/reader.action?ppg=39&docID=10539283&tm=1409086069301

When deciding whether to use the experimental option for your research, answering the following questions will be important:

1. Can cases, subjects, or participants be randomly assigned? 
2. Can the variables be manipulated? 
3. Are RCTs* effective for your research? 
4. Would an experimental intervention distort the object of the investigation? 
5. Is the research question more focused on internal or external validity? 
6. Is the research more concerned with causal processes or outcomes?

* RCT can stand for randomized control(led) trial, randomized clinical trial, and randomized controlled clinical trial. A variant is randomized field trials in which experiments are conducted in natural settings rather than laboratories.

Vogt, W. Paul (Author). When to Use What Research Design. New York, NY, USA: Guilford Press, 2012. p 51.  http://site.ebrary.com/lib/spivalib/reader.action?ppg=74&docID=10539283&tm=1409086185680

Vogt, W. Paul (Author). When to Use What Research Design.  New York, NY, USA: Guilford Press, 2012. p 47.  http://site.ebrary.com/lib/spivalib/reader.action?ppg=70&docID=10539283&tm=1409086134719

Observation is an effective research design choice when your research question leads you to:

1. Study social, cultural, psychological, or political processes as they unfold
2. Identify, develop, or refine sensitizing concepts or variables
3. Cultivate a rich or thick description within a particular context
4. Uncover or explore causal mechanisms or recognize interactive links between and among variables

Vogt, W. Paul (Author). When to Use What Research Design. New York, NY, USA: Guilford Press, 2012. p 69.
http://site.ebrary.com/lib/spivalib/reader.action?ppg=92&docID=10539283&tm=1409086232375

In order to answer your research question, do you need to generate your own data, or can you answer your question more effectively by using records or publicly available data archives? The answer depends on whether archival materials exist; detective work to find archival data must often precede data collection. If there is no way you could possibly generate data as useful for your research question as that which is available in archives, you select among data generated by others. While this in itself can be considered a kind of data generation, it is certainly very different from running experiments, conducting interviews, or surveying respondents.

Vogt, W. Paul (Author). When to Use What Research Design.  New York, NY, USA: Guilford Press, 2012. p 88.  http://site.ebrary.com/lib/spivalib/reader.action?ppg=111&docID=10539283&tm=1409086280416