The pitfalls of questionnaire surveys
Practical Action
Conclusion
So when gathering information it seems we must consider our options and the trade-offs of
using standardized, rigid and extractive versus diverse, flexible and empowering methods.
Robert Chambers has noted the following which is worth considering and hopefully will
facilitate further discussion: A reasonable rule of thumb is that conventional questionnaires
should be used only if no participatory alternative can be devised, or should be used in a light
and quick manner for confirmation and triangulation with other methods. There is reversal
here of mental state and reflex. When numbers are needed, participatory approaches,
methods, and behaviours replace questionnaires as the standard approach that first comes to
mind. (Chambers 2008: pp 128)
The following is a table that has been started to spark discussion on what methods can be
used when approaching a particular project whether it is in the health, energy, shelter and
water and sanitation sectors. Please add to the table with your comments and suggestions
adding the methods you might use if you were to embark on a project.
Information required
Population
Family Size
Literacy
Health eg. diarrhoea
Water quantity and
quality
Income and wealth
Food security
Community profiling
(land size, land use,
ownership, skills)
Method
Participatory tools
Participatory
measuring,
community census,
participatory
mapping
Participatory
mapping
Pile sorting,
Seasonal calendar,
time trends, semi-
structured interviews,
proportional piling,
matrix scoring
Participatory
measuring
Wealth ranking and
wellbeing grouping,
flow diagrams,
Livelihood analysis
diagrams,
Seasonal calendar
Visual, workshops,
mapping, transect
walks, role play,
seasonal calendar,
Questionnaire
Survey
Can be used to get
statistical data.
Can be used to get
statistical data.
Not appropriate
Not appropriate
Not appropriate
Not appropriate
unless wanting to
know the extent of
the problem. Keep it
short and focused
and use in
conjunction with PRA
tools.
Not appropriate
unless wanting to
know the extent of
the problem. Keep it
short and focused
and use in
conjunction with PRA
tools.
Not appropriate
Comments
4