Chapter11.ppt - University of Kentucky - Welcome to the University by mudoc123


									    Unobtrusive Research

Studying social life without affecting the it
              in the process
Unobtrusive      research    methods    can    be
quantitative or qualitative.

These methods allow researchers to study things
from afar without influencing the process.

   This can eliminate or minimize many of the
    threats to the validity of research, including
    reactivity and social desirability bias.
One common way to conduct unobtrusive research
is to analyze available records or events,
characteristics or behaviors. Three common
methods of using available records are:

 content analysis
 analyzing existing statistics

 historical/comparative analysis
Content Analysis
 Studying human social life through the examination of

 Common artifacts for analysis include magazine ads,
 books, television shows, bathroom graffiti, trash

 Same approach as with any research project:
    Definition of interest(s)       and   problem,   conceptualization,

    Development of means by which you’ll collect the data Sampling

    Data collection and analysis.
Manifest Content Coding
Analogous to survey data collection
Establish variables and attributes – record what is
present in each artifact – ready for data entry
Clarify units of analysis and observation
Establish a base of counting for comparison
   Generally code a sample of all editorials so you have a base
    for comparing to your findings about environmental
Latent Content Coding
Analogous to analysis of qualitative data

Look for the meaning underlying signs, symbols,
   subjective interpretation

You may begin inductively; looking for themes
Strengths of Content Analysis
 Time and money savings

 Safety net – can return to source to recheck data

 One mechanism for longitudinal analysis

 Unobtrusive – research won’t contaminate research
Weaknesses of Content Analysis
 Limited to recorded communication

 Potential for excellent reliability

 Check and recheck source

 Validity varies based partially on whether you are
 coding latent or manifest content
Analyzing Existing Statistics
 As backdrop to a study
 As the main data to a study
    Have to make due with the types of data provided; logic
     and replication help with dealing with validity problems

    Depend on the quality of the data itself
    How was the data collected?
Historical/Comparative Analysis
Much of the classic sociological work involves
historical/comparative analysis
    Marx, Weber, Durkheim, Comte, Martineau, Du Bois

Sources of data; usually a qualitative data analysis
process, but not always
    Existing historical accounts and analysis

    Raw data such as diaries, newspapers, magazines,
     public documents
Validity of Data

Question the accuracy of documents; triangulate
Is the source of your data perhaps biased in some way?

To top