From: "Dr. Christy Gladwin"
Date sent: Wed, 31 Mar 1999 17:52:37 EST
Subject: RE: Jen Scheffee report changes needed
Copies to: Hildebrand
Glad also to see Bob and others on my side re: camp officer
training, the poor forgotten extension agents! YEs, I feel like a
louse! Maybe we can do 1-day training every time one of us visits
there, and it's up to the fac member about what the training is
about? (Like I could do interviewing techniques and qualitative
research methods, Clif something else, Rick leadership training,
etc.) Will that do?
I didn't know WV had asked to cut the UF budget back. Yes, they
definitely wanted to send us as little money as possible. But I
e-mailed Joe Mowunge anyway yesterday, but did not put his approval
contingent on our sending a revised proposal and budget to Soroko,
who can always say "no" if he doesn't like it. That's the worst he
can do, so we may as well try for more money. the contract is signed
and will soon be sent out to W Vision.
I will also ask Jim Dunn for advice today. I'd really like you to
keep your eyes and ears open about the starter pack report that is
now being written by Kate Longley of NRI, London, and others, and
talk to Harry Potter ofDflD who's in the British High commission in
Lilongwe who is pretty smart (he's commissioned this study of the
starter pack), and maybe e-mail me a copy of it?? I don't want you
to just listen to everything USAID says about the starter pack --
they are too negative about it. Harry is also a friend of Anne
Conroy, who you should also meet, whose thesis was about fertilizer
and maize, and knows agriculture up the gazoo but was working in the
finance dept. She swims at Capitol Hotel and works out of her house,
but RF (Barati Patel (Zambian lady pathologist) there) will know her
phone number, and you should give Barati a copy of your report too.
I think you can have real input into this assessment even if you are
a gofer, cause you have a voice and experience. I want you to talk
to this team! Also, they should visit with Steven Carr who lives in
Zomba (on the road up to the plateau and camp ground, you see a Carr
household sign, RF will know about it) and knows more agriculture
than I EVER will! You would love him.
Also, I promised to have final report to Param April 8th. OK, here
goes. I'mjust giving my impressions as I go; change only what you
agree with. I like summary p. ii where you separate out factors
which do and don't constrain farmers: good and clear. Good writing,
organization, clear, logical as ever, etc.
I didn't understand before why you wanted research into use of Tv
(SPell out!) as an inssecticide. Please expand in an e-mail (not in
report), also about fruit trees -- for research proposal.
Table XX has CR, IF, F not spelled out anywhere. Spell out here
on p. iv.
On p. 2, please expand on the explanation of the 4 year IF cycle,
and is Average Yield of maize from the 3rd year or 3rd plus 4th year
or 1st + 3rd + 4th year? Pete Hildebrand and Peter Matlon still
think the maize-yield benefits of Ss are too low for farmers to adopt
(cause maize yields are 0 in 2nd year, only 50% higher in 4th year,
may be normal or less in 1st year, = 12.5% increase in maize
yields over the 4 year cycle, right??), so this is an important
issue/question. Can't just report on the 3rd year maize yields, and
we need this up front here.
On p. 4, please add (See Figure 3) in first paragraph, Sampling
Techs. Replace XX in second paragraph (P). 1st sentence of third P
unclear.?? P. 7: "every womAn" = replace e with a. pp. 3 & 7:
Who's "we"? Ambiguous. Maybe explain in a footnote on p. 3 that you
and I interviewed 3 women/vlg to develop the tree in fig 1, then you
yourself personally (Jen Scheffee) did the rest of the interviews.
It's NB (important) for reader to know that you personally did these
interviews, not a camp officer or interviewer, but you yourself. I
understand if you don't want to change we to I cause in our jointly
authored papers, we'll have to change it back again and it's a
hassle, but here it should be "I". And your name can be on p. 1 of
report, as Bob and Robin did, as this is yours.
"Most female testers come from FHHs." Wild. Pie charts are good.
p. 12: Maybe add a sentence about how residence is patrilocal
here, ethnicity of 4 villages? Which is Chewa and which is Ngoni?
Residence same or different? Need this here, not just in appendix.
p. 13: Think you need to add, "as compared to very poor men" to
"more very poor women who test the tech adopt it." Interesting!
p. 20: this year's harvest = 98/99? Clarify. 73% 84% hhs are
not food self-sufficient?? Wow. How much off-farm work is there?
(p. 20). Did you get data on that? Forget. If residence is
patrilocal, why do women hold land at all? (Table 11). Do married
women hold land, or only FHHs? IfFHHs, clarify. On p. 20-21, did
you do the pie charts testers/non-testers for men/women with fallow
land vs. men/women without fallow land? This would be good.
p. 22: is manure cow or goat as in Malawi? Only 35% use manure:
interesting. 30% stopped fertilizer use in early 1990s! Can you say
why... "when fertilizer prices increased from to or
devaluation of ? If not, OK.
p. 23: small bags applied to rape,... and maize ... in dimba
production? So on average, men got 4 bags total, and women 2 bags
total offert. Good on credit and fert. What do you feel -- in
gestalt sense -- of these FHH -- NTs? Are they like the shy young
married woman we interviewed whose picture I took? From p. 25 table,
they won't do anything!! Are they drunkards? Young marrieds? Too
poor? What? Any feel for that? Do we need more research on this?
p. 28: IGA section good. Any chance of frequency counts to go
with IGA 1-25? Pie-charts of testers/non-testers for Men/women with
IGAs? Ready to kill me?
p. 29: Insert in 3.2.1: According to "Figure 1 (p. 5-6 above)"
cause they won't know what you mean by "first dec tree". I'd use
"should" rather than "would", or "the model in Fig 1 says that"....
Then in 3rd P at bottom of page, you need to tell reader that some
questions on the questionnaire were specifically designed to test
this model in Figure 1, to see if it accurately predicted what the
farmers did re adoption of IFs, in order to correctly identify the
motivations of the IF-testers and the constraints of the
IF-nontesters. This section is about "testing the dec tree model to
test an IF", rather than testing the IFs -- that's not perfectly
clear from your heading. P. 30 top is pretty cryptic: tell reader
that the errors (e.g., on the left-hand path of figure 34) help you
revise the decision tree so that it can more correctly identify the
reasons farmers adopt. E.g., error rate of 47% after criterion "Are
you worried about animals getting into an IF?" tells us that
the risk of animals grazing in an IF is either not an important
criterion to farmers who adopt -- or this criterion is not specified
correctly, i.e., the farmer-non-tester is more than worried; s/he is
convinced animals will destroy any benefits of an IF.
p. 32: "second dec tree" is unclear; use "dec tree in Fig 35 is
the model of the choice process for the 91 farmers who have passed
Fig 34 because they are motivated (by at least one reason) to try IFs
and so they proceed on to the list of constraints in Figure 35." The
dec tree in Figures 34 and 35 is one decision tree model rather than
two, so say "Obviously, the dec tree model in Fig 34 and 35 was
fraught with errors." Is the overall error rate of this model 47%?
Say so in this paragraph.
p. 35: (I hope these page numbers are the same; Kim might have
changed them when she made the paper size 8 1/2 by 11): In
the revised decision tree in Figures 36 and 37" -- please replace
XXs which are very confusing here. In figure 38, there's typo
unlesS. GOOD paragraph following fig 38, "Following the logic of
"the revised model in Fig 38, if you have seen the benefits.... But
you now need to include a figure 39 with 117 cases going to it, cause
the third model includes the list of constraints too, with an error
rate of this model marked clearly. Where's figure 39? What's the
error rate of this third tree? This is important, as it is the one
you'll want to further test in the baseline survey with World Vision,
no? So let's see the disaggregation of gender data on this one, not
necessarily all the TE, NT stuff but FHH, M MHH, F MHH
disaggregation -- cause this is the one we'll test further. (I
still would like the fertilizer wodula? question tested in that
baseline survey, as I'm still now convinced of the "one bag of
fertilizer" question, which might be too narrow a phrasing of that
I've got to go home now, the rest of it (pp. 36-46) I'll go over
tomorrow, hopefully. Hope this helps and is not too burdensome,