UNITS OF ANALYSIS
    Individuals
    Groups
    Organizations
    Social Artifacts
    Units of Analysis in Review
    Faulty Reasoning About Units of Analysis: The Ecological Fallacy and Reductionism

    Most simply but uselessly put, units of analysis are the units we analyze.  Typically, these are individual human beings.  They can be the individual voters surveyed in a political poll.  They may be inidividual college students in research on student study habits.

    If we did a survey to discover the percentage of households with an internet connection, however, a group--the household--is the unit of analysis.  If we want to learn what leads to a successful (or unsuccessful) marriage, then married couples would be the units of analysis, not individual husbands and wives, even though we would probably talk to them and get information about them.  If we want to compare gangs from different ethnic groups, then the gang--a group--would be the unit of analysis.

    Sometimes formal organizations are the units of analysis.  We might ask whether e.com companies are more democratically run than conventional ones.  Are Protestant or Catholic churches more likely to sponsor day-care centers?

    Finally, social artifacts are sometimes the units of analysis: books, songs, poems, paintings, secret handshakes, etc.

    A discussion of the ecological fallacy and reductionism will illustrate the importance of being clear about your unit of analysis or show how unclarity can lead you into error.  For example, if cities (the unit of analysis) with the most sociologists have the highest crime rates, that doesn't mean sociologists are committing the crimes.  Maybe they are drawn there for research purposes.