Observational Studies and Document Analysis

Document analysis

measuring things unable to be measured through personal interviews or direct observation, lab/field experiments. Gathering data through existing documents

what are the advantages of document analysis?

-Distant in time (used what was written in that time)
-Distant in place (data that other nations produce and compile it)

Written Record

allows us to measure things that happened long ago, can be inexpensive if collected properly, big sample size, low instrument reactivity

The Episodic Record

Data not going through an ongoing, systematic, datakeeping program (diaries, manuscripts, memoirs)

The Running Record

Produced by organizations rather than private citizens (NYT archive, data carefully stored and easily accessible)

What are Case Studies

-Collecting data through the written record (can also do through observational study)
-Examining a single or very few cases of a phenomenon in considerable detail

Advantages of Case Studies

Small sample size

Disadvantages of Case Studies

-If you’re studying a single case, you might get a lot of info but may not be generalizable to other cases
-Bias (studying 1 case can get you Stockholm syndrome)

What is instrument reactivity?

-individuals affected by the instruments in the research
-Stress of being observed (stalking)

What is cross-sectional analysis and why must a cross-section be generalizable? Why is cross-sectional analysis so popular in the social sciences and what are the advantages and disadvantages of cross-sectional analysis?

-Gathering data on the cross-section of the population and then trying to isolate the IVs and the DV at that one point in time

-Sacrificing internal validity (control over the experiment) for external validity

-It is very popular in the social sciences because it is very realistic

-Limitations: a form of nonexperimental design, no control group to compare it to, no pre-test, gathering data at one point in time means you don’t know if something has changed, no way to control the effects of history, a bad cross section means results are useless

What is time-series analysis?

-Cross-sectional design over time
-Asking the same DV at multiple points in time, but to a different cross-section of the population (seeing if the DV has changed)
-represents both pre-test and post-test

What are the limitations to a time-series analysis?

It’s a different group of people so we can’t identify why individuals have changed

What is panel-study design?

Gather time series data, exact same questions and answers from the exact same people over a period of time, you can track individual change

What are limitations to panel-study design?

expensive and difficult to collect