Agriculture Reference
In-Depth Information
the iterative use of methods and the role they played in listening, dialogue and
action, as well as in overall analysis and problem solving.
Critics of PEAR suggest that bias arises when researchers come in close
contact with and empathize with individuals involved in research, thus losing
their supposed objectivity. An alternative view is that data are co-constructed
by the researcher and participants while situations change when we are
interacting with one other and our environments. In other words, reflection is
at the heart of socialized knowledge. Human values enter into the research
process during problem selection, instrument design, analysis and
interpretation, and reflection is required in order to better understand and
avoid the trap of subjective conclusions (Guba and Lincoln, 1989, cited in
Sohng, 1995; Cornwall and Jewkes, 1995). Validity (the extent to which one is
actually measuring what one wishes to measure) and reliability (the extent to
which a result or measurement will be the same every time it is measured)
remain important considerations. Techniques that were used in this work to
improve validity and reliability included the use of standard procedures for
data-collection tool development (e.g. in questionnaires), appropriate analysis
of different kinds of data (e.g. statistical software for quantitative analysis)
and triangulating results obtained through different methods (e.g. use of
secondary data sources on health problems). PEAR also considers whether
social action arises from the research process and whether the changes made
bring about the outcomes observed - so-called catalytic validity (Heron, 1988,
cited in Sohng, 1995).
Language was a key issue in interpretation and analysis. French was the
working language for this thesis and was a second language both for the lead
researcher (a native English speaker) and for the farmers (native Wolof
speakers). For data collection, questions were written in French and then
translated into Wolof. For the questionnaires, another Wolof speaker from
outside the research context translated what he heard into French to assure
the words were correctly interpreted. All interventions were first developed in
French and then translated into Wolof, and then translated into French for
me. The notes taken were a combination of French and English and were
eventually transcribed into English. Given the important role that language
plays in the interpretation of meanings, the several filters that developed
during the research might have influenced the analysis of findings.
R
ESEARCH
F
INDINGS
Health and environment impacts linked to urban agriculture
When asked directly what the main environmental problems were in the
region in the year 2004, farmers spoke of domestic waste, rotting dead
animals, mosquitoes and random defecation in the fields. Other studies have
suggested that Dakar's waste-production levels exceed the garbage collection
and recycling capacity of the city because most citizens live under precarious
