Chapter 10: Unobtrusive research
Unobtrusive research: methods of studying social behaviour without affecting it. There are 3
types of unobtrusive research methods.
1. Analysis of existing statistics (Durkheim’s study of suicide was an example of it)
2. Content Analysis (researcher examines a class of social artifacts that are usually written
documents like newspaper editorials)
3. Historical and comparative research (a form of research in which the main resources for
observation and analysis are the historical records).
Analyzing Existing Statistics:
In this method, we look at ways of drawing on data analysis reported by others, such
as government agencies.
It differs from secondary analysis in which you obtain a copy of someone else’s data
and do your own statistical analysis.
Nonetheless, much overlapping and limitations in both of these two methods.
the limitations and constraints of available measures and lack of knowledge
about how these measures were collected and coded.
Lack of ethical approval in both methods.
Existing statistics can often provide the historical context within which to locate your
For example, Sosteric did participant observation study in a nightclub in a
small Canadian city. In support of his claim that the club’s workers made
above average incomes, he reported data from statistics Canada on industry
averages for full and part-time workers.
Existing stats can also provide the main data for a social scientific inquiry. An example
for it the classics study of Emile Durkheim.
He wanted to discover the environmental and social conditions that encourage
or discourage the suicide patters.
He noticed that the predominantly protestant countries had 190 suicides per
million population; mixed protestant Catholics 96, and predominantly catholic
Thus he suggested that the suicides are a product of anomie, or a general sense
of social instability and disintegration.
During times of political strife, people may feel that old ways of society are
collapsing and they become more stressed and demoralized. Suicide is one
answer to this severe discomfort. Seen from other direction, social integration and solidarity reflected in
personal feelings of being part of a coherent enduring social whole offer
protection against depression and suicide. Here the religious difference comes
in. Catholicism is more structured and integrated system, which gives a
greater sense of coherence than the loosely structured Protestantism.
Units of Analysis:
The unit of analysis in the study of existing statistics is often not the individual.
Durkheim for example worked with geographical-political units like countries, regions,
states and cities.
This can sometimes cause a problem. As in Durkheim’s case, he wanted to know whether
Catholics or Protestants were more likely to commit suicide. The difficulty was that none
of the records available to him indicated the religion of those people who committed
suicide. Because protestant countries, regions and states had higher suicides than Catholic
countries, regions and states, so he drew the obvious conclusion.
There’s a danger in drawing these kinds of conclusions. It’s always possible that patterns
of behaviour at a group level may not reflect corresponding patterns on individual level.
Problems of validity:
Two characteristics of science are used to handle the problem of validity in analysis of
existing statistics. - Logical Reasoning and- Replication. (How do you know if the
measurement is an accurate reflection of what you are measuring?)
Durkheim’s study provides the example of logical reasoning. Although he could not
determine the religion of people who committed suicide, he reasoned that most of the
suicides in a predominantly protestant region would be protestants.
Replication can be done by repeating or retesting your measures and seeing if you get the
Problems of Reliability: (is it an accurate report of what it claims it reports?)
This problem can also be solved through logical reasoning and Replication.
Sources of existing statistics:
Statistics Canada: responsible for Canadian census, labour statistics, official stat crimes
and stats on health and welfare, finance, agric