Indicators vs. Questions Events vs. Rates Rate is events in a defined group in a limited time period Rare events require special methods; How you collect often trumps what you collect (mortality)
Objective of Collecting Information Summarize and represent the responses of people: common views and variations Drivers and groupings of those responses
One Method Cannot get a Fair and Full Representation of a Community’s Voice Opinion vs Behavior Multiple Ways to Ask What matters is What changes over time What differences exist between subgroups What stays the same when everything else changes What differences are bigger than the imprecision we have in measuring them
How to seek information Systematic survey at point of use Water hole chat, commercial outlet sales record Meeting with community leaders ‘Snowball’ from one person to the next Remote sensing (new technologies) – Ushidi vs telecom
To Represent The Whole Random – Simplest, but often not the Best If not have a list of all members of community imitate it with Cluster in the last (geographic) stage If population not residential, not equal chance of being included (if 50% of people on the move, random is a joke) Then ‘random’ often false representivity
Systematic or Purposive Sample Seek most affected areas Find most affected and less affected groups (areas) Summarize before and after, high and low impact groups Represent the range of needs, not the average of the community Identify strata of key subgroups
Strongest Information From Triangulation Consistency of information across areas and levels of information collection Household and community interviews Opinions and quantitative representations
Sensitivity Representivity Valid Reliable (repeatable)