The Kellogg Pilot
Readers of Phatlalatsa will know that S&T has been providing monitoring and evaluation services to the WK Kellogg Foundation Africa Programme for two years. This has included designing a multi-country baseline survey; S&T partner Nobayethi Dube describes the pilot phase, which she managed.
PILOTING THE INSTRUMENT
As part of the continuing support that Strategy & Tactics (S&T)
has been giving to the Kellogg Foundation, S&T helped design a baseline
survey questionnaire and managed a pilot of 100 households in the Mount
Mount Frere is a rural area, 98 km from Umtata. The pilot was undertaken
with the help of Isinamva Community Development Centre, a Non-Governmental
Organisation (NGO) which is funded by Kellogg Foundation. This organisation
facilitated access to the communities we visited, namely Mbodleni
and Mhlokwana. They also recruited the twelve young people who were
used as fieldworkers during the pilot.
The Kellogg Foundation invited four of their District Facilitators
from Lesotho, Limpopo and Zimbabwe to observe the training and assist
in the field, because the baseline survey itself will be implemented
in all the countries where Kellogg is implementing its integrated
rural development programme.
Training for the pilot took place over two days. The training was based
on how to administer the questionnaire as well as on sampling methods.
The first day was spent going through the questionnaire in detail and
addressing each question carefully. We then staged role-playing exercises
with interviewers using scenarios that may come up in the field. The purpose
of role-play was also to come up with translations for some of the concepts
used in the questionnaire.
The second day was spent training fieldworkers in sampling methods - this
was important since this was a dual-level survey which included household
information from the head of the household, and attitudinal and other
information from a randomly sampled household member.
To select the household, we used a public place (schools, shops)
as a starting point, moving towards the concentration of houses.
Because the villages that we visited had very scattered homesteads,
we administered the questionnaire at every third homestead/stand.
The second stage of sampling involved identifying the head of the
selected household. For the pilot, the head of household was identified
by asking simply who headed the household, and in their absence
(where it was long-term) we followed up by asking who made important
financial decisions for the household and could respond to questions
about the household. This became the respondent for Section A of
the questionnaire that covered household-level information.
The third stage of sampling involved identifying the second respondent
- who answered questions in Section B of the questionnaire relating to
attitudes, values, gender & youth, and related matters - for which
we used the birthday rule. This simple randomisation rule selects the
person whose birthday is next in the household as respondent. It should
be noted that in some instances the head of household could also respond
to questions in Section B, by virtue of being the one who celebrates their
The second day also involved going out to the nearest village to administer
the questionnaire - to pilot the pilot, as it were. The purpose for this
exercise was to see how long the questionnaire took to administer, if
respondents understood the questions, and if there were any other problems.
This also allowed us to have a feedback session where we further developed
The actual fieldwork began on the third day. The fieldwork team was welcomed
by the community but there were the usual problems where potential respondents
felt that 'filling these papers is a waste of time'. In some instances
we could not get access to respondents because they were working in town,
or conducting their daily chores such as fetching water from the river.
This meant that we had to arrange for second or third visits. The exercise
also highlighted that rural communities have busy lives and are not just
readily available. The pilot took four days to be completed.
S&T will analyse the data in order to assess the extent to which the
survey measures Kellogg's indicators. Nobayethi will draft a fieldwork
report, fleshing out the issues raised above. Thereafter, Kellogg and
its partners will decide how to move ahead with the baseline survey.