Anuman- An inference for helpful in diagnosis and treatment
Day 1 Session 7 Quisumbing_ Linking mixed methods
1. Designing your study: Linking gender
and nutrition through qual and quant
methods
Agnes R. Quisumbing
IFPRI/A4NH
Presentation at the A4NH Methods Gender-Nutrition Methods
Workshop, December 6-7, 2013, Nairobi
2. Overview
• There are many research methods to use for linking
gender, agriculture, and nutrition
• Need to remember the right tool for the right question,
but most importantly, need to have an appropriate
study design
• An ideal study design to understand gender,
agriculture, health, nutrition linkages must be able to
link and integrate:
– Qualitative and quantitative methods
– Social science and nutrition data
3. Why should the design take linkages into
account from the start?
Qual and quant methods
• Demonstrated gains from
various research programs
from using mixed methods
work (CPRC, CAPRi, IFPRI
intrahousehold, GAAP, etc.)
• Quant work allows you to
measure impacts, qual work
enables you to understand
why
• Both are important
Social science and nutrition
• Social norms underlying
decisionmaking processes
often determine the
allocation of resources
toward health and nutrition
• Among these are norms and
practices surrounding
gender
4. What to take into account in designing your study
Qual and quant
•
•
•
•
•
Integrated and iterative qual and
quant
Qual and quant researchers work
together to define research
questions, analyze and interpret
results
Qual study as diagnostic, help
frame the questions
Use quant sample to define
“frame” of qual study—then can
link qual study responses to quant
data
Quant sample typically uses larger
n, qual study can use small to
medium n, but be purposive
Social sciences and nutrition
• Have a “theory of behavior”
that links behavior to nutrition
outcomes
• Collect data on determinants
(gender, resources, etc.) and
on outcomes (nutrition,
health, education, etc.)
• Collect data on the same
individuals (same households)
for whom you are collecting
nutrition outcomes
5. Food-Based Approaches to Reducing Micronutrient Malnutrition
Intervention: Dissemination of improved agricultural
technologies (vegetables and fish)
Nutrition Objective: Improve micronutrient status of producer
population
6. Evaluating the long-term impact of agricultural
technologies in Bangladesh
Panel data set based on 957 households surveyed in 1996/7 and 2006/7
3 technologies/implementation modalities:
1. improved vegetables for homestead production, disseminated through
women’s groups (Saturia)
2. fishpond technology through women’s groups (Jessore)
3. fish pond technology targeted to individuals (Mymensingh)
Compare “early adopters” to “late adopters”
Page 6
7. Survey Design in 1996/7
-4 round panel 1996/1997
-Coverage of 3 major agric. seasons
-3 sites, 47 villages, 955 HHs
IN EACH SITE
HH type
NGO member
adopters
NGO member,
likely adopters
Page 7
Type of NGO village
“A” technology
had been
introduced
“B” technology
had not yet been
introduced
A (n=110/site)
Non-NGO members, C1 (n=55/site)
general population
B (n=110/site)
C2 (n=55/site)
9. Data collection efforts over the years
• 1996-97: 4-round quantitative household survey
• Qualitative work on gender conducted between rounds 3 and 4
(Naved 2000)
• 2001: Qualitative work and further quantitative analysis to look at
impact of new technologies on poverty, empowerment,
vulnerability in 2000 (Hallman, Lewis, Begum 2007)
• 2006-2007: Qual-quant chronic poverty study
– Focus groups (25% of sample villages)
– Quant household survey (all respondents and new households formed
from original)
– Life histories (based on poverty transition category; poverty status
computed from quant survey)
• 2010: Follow up on impacts of food price crisis
Page 9
10. Information collected at household and individual levels in 199697 and 2006-7 rounds
Household
•
•
•
•
Per capita expenditures (food,
nonfood consumption)
Household assets and landholdings
Household income, by source
Detailed production module
Individual
•
•
•
•
•
Note that dietary diversity modules were
relatively “new” at the time of the baseline
•
survey, but DD scores can be calculated from
the food consumption module at household or
Individual levels
Page 10
HH roster information (age, sex,
education, relationship to hh head)
Schooling, labor and employment
Land and assets
Individual food consumption, 24hour recall (and then converted to
nutrient equivalents), all individuals
Hemoglobin (via Hemocue), all
children and women up to age 65
Height, weight for all hh members