Data collection and analysis tools for food security and nutrition

 In most societies today, the way in 
which citizens interface with the local food, 
health and environment systems – and thus 
contribute to determining their own food 
security and nutrition outcomes – is through 
personal, household, group and community-
level decision-making and actions, all of which 
are conditioned by the data and information 
they have access to. 
FSN drivers encompass macro level constructs 
made up of many fundamental elements which 
FIGURE 1:
FRAMEWORK FOR A SYSTEMIC VIEW OF FSN TO GUIDE DATA COLLECTION AND ANALYSIS
Source: Figure inspired by the HLPE-FSN Sustainable Food Systems Framework (HLPE, 2017, 2020), the UNICEF conceptual framework of the determinants of 
malnutrition (UNICEF, 1990, 2021), the socio-ecological model (Bronfenbrenner, 1979), and the Sustainable Livelihoods Framework (SLF)(DFID, 1999).
Macro le
vel determinants:
en
vir
on
me
nt,
te
ch
no
log
y a
nd 
inn
ova
tion
, edu
cation
, infrastructures, economic, socioc
ultura
l, poli
tical 
and
ins
titu
tion
al s
ettin
gs
Cro
ss-
cut
tin
g d
ime
nsio
ns: a
vailab
ility, access, utilization, stabilit
y, agen
cy, su
stain
abili
ty
Meso 
level determinants:
foo
d s
yste
m, he
alth system, environment 
syste
m
Mic
ro lev
el determinants
:
pe
rs
on
al,
ho
us
eh
old
, gr
oup a
nd community-le
vel d
ecis
ion
-m
akin
g 
Individual
food security
and nutrition outcomes:
diet adequacy, nutrition 
status, well-being

---

[
15
1 
SETTING THE STAGE
can be grouped into the following categories: 
trends in FSN outcomes, as well as trends that 
are occurring in other domains that are drivers 
of food system change, including: biophysical 
and environmental; technology, innovation 
and infrastructure; economic and market; 
political and institutional; socio-cultural; and 
demographic. (HLPE, 2020). Environmental, 
sociocultural and economic determinants, 
including their externalities, are explicitly 
included in the macro level because these 
drivers are only implicit in the “sustainability” 
dimension of FSN. Taken together, macro level 
drivers contribute to shape the more proximal 
food, health and environmental systems at 
the meso level, which jointly determine the 
enabling environment – comprising availability, 
affordability, proximity, knowledge and practices 
related to food – for people to become agents of 
their nutrition. (HLPE, 2020; UNICEF, 1990).
Deciphering the micro level or immediate 
determinants of FSN requires further discussion. 
For individuals to benefit from the flow of 
locally available goods and services related to 
FSN, decision-making must take place both 
individually and in groups, in coordination with 
their families and communities. At the micro-
level, specific and different territorial settings 
exist within national and regional levels (e.g. 
rural areas with livestock, fishery areas), which 
present highly diverse potentials. Together with 
individual and collective agency, diverse areas 
shape the possibility to achieve food security 
and nutrition for those who live or work in 
these areas. It is at this immediate interface 
between the individuals, their local food and 
health systems and their local environments, 
that people’s food security and nutrition is 
determined through myriad types of decision-
making. This is still nested within, and thus 
influenced by, the ever-present macro level 
determinants in a given society.
Finally, cutting across the four interrelated 
levels of our conceptual framework for FSN 
are the six dimensions of FSN: agency, stability, 
sustainability, access, availability and utilization 
(HLPE, 2020).
One notable challenge in the design of 
conceptual frameworks like the one discussed 
here is to incorporate the complexity deriving 
from the existence of competing views of life. 
Many Indigenous Peoples, for example, have 
a 
biocentric view of life (DesJardins, 2015) 
that differs radically from the 
anthropocentric 
approach conveyed in the conceptual 
framework, where human separation from 
nature is high, and human intervention is 
justified to actively attempt to regulate inputs of 
energy, nutrients, water and/or temperatures 
to favour production. In a biocentric view of life, 
ecosystems and their human and non-human 
co-inhabitants are intrinsically connected. 
Biocentrism underpins Indigenous Peoples’ 
traditional knowledge, culture, language, 
values, spirituality and cosmogony, as well 
as their food systems (FAO, 2021), informing 
practices of food generation and production 
and natural resource management. The 
inclusion of the environment as a proximal and 
systemic determinant of FSN in our conceptual 
framework is intended to accommodate this 
and to allow for considerations related to how 
biocentrism underpins Indigenous Peoples’ 
food systems.The conceptual framework 
provides guidance as to the different topics or 
themes data should encompass for a system 
approach to FSN data collection and analysis 
tools. Thus, the framework can be used to 
visualize potential impact paths and indicators 
at each societal level for FSN outcomes. The 
concentric circles in the conceptual framework 
are inspired by the aforementioned systems 
approach related to the socioecological model, 
and are 
not designed to convey a top-down 
approach that could result in overlooking the 
needs of local populations, including indigenous 
communities. On the contrary, the focus 
on the decision-making sphere at different 
levels grounds the framework in the human 
right approach, including for example the 
consideration of the right of Indigenous Peoples 
to self-determination, which includes the right 
to food as per the conjunction of the United 
Nations Declaration on the Rights of Indigenous 
Peoples (UNDRIP) (A/RES/61/295) and the 
International Covenant on Economic, Social and 

---

16 
]
DATA COLLECTION AND ANALYSIS TOOLS FOR FOOD SECURITY AND NUTRITION
Cultural Rights
8
, by virtue of which Indigenous 
Peoples freely determine their political status 
and economic, social and cultural development. 
Consistent with the United Nations Declaration 
on the Rights of Peasants and Other People 
Working in Rural Areas (UNDROP), adopted 
by the UNGA in 2019 (A/RES/73/165), the 
framework provides for the inclusion of rural 
peasants and other local food system actors 
as important agents in FSN-related policy 
decisions.
8 Adopted on 16 December 1966 by UN General Assembly 
resolution 2200A (XXI).
DATA-INFORMED DECISION-
MAKING CYCLE
Because the conceptual framework does not 
provide insights on how priorities for data 
collection are decided upon, another critical 
conceptualization involves recognizing the steps 
needed to ensure more effective and efficient 
data-driven decision-making.
To this aim, we have adapted the data value 
chain from the Nutrition Data for Accountability 
and Action Framework (Piwoz 
et al., 2019) to 
illustrate six critical stages in the process of 
data-informed decision-making for food security 
and nutrition (
SEE FIGURE 2
).
FIGURE 2:
DATA-INFORMED DECISION-MAKING CYCLE
Source: Adapted from Piwoz 
et al., 2019.
Effective data 
governance and 
inclusiveness across
all FSN related data 
systems
Start by defining/refining 
evidence priorities and 
questions
Identify data needed. 
Review, consolidate, 
collect and curate data
Disseminate, share, 
review, discuss results, 
refine insights and 
conclusions
Analyse data, using 
appropriate 
analysis tools
Use results, insight 
and conclusions to 
make decisions
Translate data into 
results, insights 
and conclusions
Effective data 
governance and 
inclusiveness across
all FSN related data 
systems
Start by defining/refining 
evidence priorities and 
questions
Identify data needed. 
Review, consolidate, 
collect and curate data
Disseminate, share, 
review, discuss results, 
refine insights and 
conclusions
Analyse data, using 
appropriate 
analysis tools
Use results, insight 
and conclusions to 
make decisions
Translate data into 
results, insight and 
conclusions

---

[
17
1 
SETTING THE STAGE
Our data-informed decision-making cycle 
consists of six components. At the centre of 
the cycle is effective data governance and 
inclusiveness across all FSN-related data 
systems, as a fundamental prerequisite to 
carrying out each of the components in the 
cycle for improved data collection and analysis. 
Data governance is central because multiple 
sectors and stakeholders are needed across 
the steps, but these may vary from step to step 
and depending on the nature of the specific 
issue(s) being addressed. Many users of this 
data-informed decision-making cycle will be 
inclined to follow only some of the components. 
However, the components are presented in a 
cycle in order to clearly and concisely illustrate 
the sequence of steps from collecting raw data 
to ultimately using the information gleaned from 
these data to guide decisions related to FSN. 
We distinguish between defining and refining 
evidence priorities and questions as an essential 
first step in the cycle, regardless of how many 
components are subsequently performed. The 
data-informed decision-making cycle for data 
collection and analysis for FSN is, precisely, a 
cycle, rather than a linear sequence, because 
previously published evidence should be used 
to facilitate this first step of defining or refining 
evidence priorities and questions. Thus, as a 
crucial starting point, prior to any data collection 
or analysis, it is important to define a clear set 
of evidence priorities, in line with effective data 
governance and inclusiveness, and to identify 
focused questions with clear linkages to said 
evidence priorities. The evidence priorities and 
related questions will serve as a clear guide 
for the subsequent steps in the data-informed 
decision-making cycle.
The other components of this cycle are: the 
identification of the required data and the review 
and consolidation, or collection, of primary or 
secondary data and the curation of the data; the 
analysis of the data, using appropriate analysis 
tools; the translation of data into results, insights 
and conclusions; the dissemination of the data 
in order to share them, review and discuss the 
results, and refine insights and conclusions; and 
the use of the results, insights and conclusions 
to make decisions. Depending on the different 
actors or stakeholders involved, they may 
perform one or a few of these components, or 
they may perform all of them sequentially.
Ideally, evidence priorities should be defined 
though a democratic process involving the 
decision-makers and the beneficiaries. To guide 
the process, several considerations, including 
the concepts in the conceptual framework, 
should be considered, especially when working 
with Indigenous Peoples. Both decision-making 
and prioritization are influence by external 
factors beyond research agendas, such as 
the quality of existing data. Once the evidence 
priorities and related questions are clear, 
the next step is to review and consolidate any 
existing data on the topic. If necessary, new 
primary or secondary data can be consolidated 
with existing data and used with analysis tools. 
In many cases, it may not be necessary to collect 
new data, as it may suffice to organize the 
existing data in way that it is useful to answer 
the questions. In other cases, it may be more 
appropriate to either re-define the questions or 
use sound proxies based on available data. Once 
the existing data are organized, if the questions 
are still not satisfactorily answered, a plan to 
collect new data can be made. In order to ensure 
that the data can be effectively translated into 
results and that those results can be effectively 
disseminated (these being the next two steps in 
the cycle) it is extremely important to plan the 
collection of new primary or secondary data in 
accordance with both the evidence priorities and 
the FSN conceptual framework. Locally adapted 
data and information reduce the risk of one-
size-fits-all solutions that can be detrimental 
to the well-being of Indigenous Peoples and the 
sustainability of their food systems, as have been 
observed during the COVID-19 pandemic (FAO, 
2020, e-consultation).
Both existing data and newly collected data, be 
they quantitative or qualitative, will likely need 
to be transformed (at a minimum, cleaned) 
and analysed using analysis tools to address 
the initial objective and related questions. 
Both new and existing data should be entered 
into a data management platform to be stored 
and subsequently organized and analysed in a 

---

18 
]
DATA COLLECTION AND ANALYSIS TOOLS FOR FOOD SECURITY AND NUTRITION
sensible fashion. Ideally, results and conclusions 
should be tested for accuracy and taking 
corrective measures; including field-testing 
when necessary and feasible and, if needed, 
corrective measures should be taken. Field-
testing aids the ascertainment of the accuracy of 
the data and the research methods used. Ideally, 
all stakeholders should also be involved in field 
testing the results and related conclusions prior 
to disseminating the results and conclusions.
In the translation step of the data-informed 
decision-making cycle, the results are translated 
into meaningful information, designed with a 
specific audience or user in mind, in way that 
can eventually be disseminated and used. Data 
translation is necessary for data dissemination, 
but it is a separate step. This is why these 
steps are conveyed as separate components 
of the data-informed decision-making cycle. 
Translated data in the form of results should 
disseminated, reviewed and discussed to provide 
opportunities to refine insights and conclusions 
prior to the results being freely used and 
perhaps, misinterpreted. This process should 
include recognizing the strengths and limitations 
of the analysis. Therefore, dissemination is an 
important step in the cycle, which should be 
conducted before the results and conclusions 
are used by diverse stakeholders (including civil 
society and participants in the data collection 
process) for decision-making and new, related 
endeavours.
Ideally, the data, both in its crude and translated 
formats, and the related analysis tools, should 
be made available to users at any level of the 
conceptual framework, from the macro level, 
through national-level policymakers, to those 
involved in the meso and micro levels, in group 
and individual level decision-making. The 
resulting data and results can, in turn, be used 
as input for subsequent, new, evidence priorities 
and related questions. It is for this reason that 
we refer to the data-informed decision-making 
process as a cycle and have conveyed it as such 
in Figure 2.
The data-informed decision-making cycle has 
a simple design; such that it can be adapted to 
any FSN objective and data type. It is important 
to acknowledge that this simple design does 
not negate the inherent complexities, such as 
bottlenecks, constraints and biases, in such a 
process. One of these complexities is the need to 
respect the traditional and customary institutions 
of Indigenous Peoples, who often maintain their 
own traditional governance systems. Under the 
framework of the right to self-determination, 
stakeholders who establish data-selection 
priorities must involve the traditional governance 
systems in the process of data collection and 
subsequent policymaking and implementation. 
The practical reality of utilizing a data-informed 
decision-making cycle such as the one presented 
here, includes a contested nature of data, the 
empirical reality of decision-making, and critical 
issues of power and voice in decision-making. 
The importance of strong coordination between 
actors in addressing bottlenecks and constraints, 
throughout the data-informed decision-making 
cycle, will be addressed in subsequent sections 
of this report.
USING THE CONCEPTUAL 
FRAMEWORK AND THE 
DATA-INFORMED DECISION-
MAKING CYCLE TO ADDRESS 
ISSUES RELEVANT FOR FSN
To demonstrate how the theoretical guidance 
presented in the conceptual framework (
FIGURE 
1
) can be overlaid with the methodological 
guidance presented in the data-informed 
decision-making cycle (
FIGURE 2
), we have 
constructed a matrix (
FIGURE 3
) showing how 
both of these can be used together to address 
evidence priorities and related questions relevant 
for FSN; with a particular aim of parsing the new 
data collection that would be necessary for such 
work. In this matrix, the column headers are 
the five main components of the data-informed 
decision-making cycle, after defining and 
refining the evidence priorities and questions, 
while the row headers are the levels in the 
conceptual framework.

---

[
19
1 
SETTING THE STAGE
FIGURE 3:
HOW TO STRUCTURE A DATA-INFORMED, DECISION-MAKING PROCESS MATRIX
Completing a data-informed decision-making 
process matrix that combines the levels in the 
conceptual framework with the phases in the 
data cycle is not an easy exercise, but it is a 
necessary one. This is because too often policies 
and programmes and related objectives (e.g. 
updating food-based dietary guidelines) are 
designed without considering all the levels of the 
conceptual framework, or used to initiate new 
data collection without a sound understanding 
of the existing data. As the six cross-cutting 
dimensions of food security are still new, related 
FSN polices, programmes and objectives have 
not yet been purposefully framed with these six 
dimensions in mind. When thinking about FSN 
policies, programmes and related objectives it is 
imperative to think about how the different types 
of users of data and analysis tools might interact 
with or be aware of such initiatives. It is for these 
reasons that we recommend those interested 
in designing or updating FSN initiatives work 
through the challenging exercise of completing a 
data-informed decision-making process matrix 
(
FIGURE 3
).
First, by identifying what data are available (and 
from where) at the macro level (for instance, 
national-level data on environment, technology 
and innovation, infrastructure, economic, socio-
cultural, political, and institutional settings, 
education) one can identify which of the six 
cross-cutting FSN dimensions are captured, or 
not, in the identified databases. It is important to 
note that, for the matrix, 
environment is defined 
in the broad sense. Thus, environmental data 
can include data related to food availability (e.g. 
national fruit and vegetable yields) as well as 
national climate-related data that one might 
view being more related to the FSN dimension of 
sustainability. The proposal is to aim to identify 
data from as many of the macro drivers as 
possible, while also aiming to capture as many 
of the six cross-cutting dimensions as possible 
within these newly identified existing databases, 
if possible – all six of them.
Then we suggest continuing by reviewing and 
identifying the existing data for the subsequent 
levels in the conceptual framework – in other 
words, filling in the rest of the first column 
of the matrix. Thus, at the meso level, within 
the country (i.e., at the provincial, district or 
municipal level) food system data related to 
availability and access that varies throughout 
a given country should be identified, especially 
because this variability will affect the 
sustainability of a given objective within a given 
country. Regional and municipal-level databases 
regarding health and environmental systems 
should also be identified, which may include 
identifying existing food- or health-related 
policies (such as food-based dietary guidelines). 
At the micro level, identifying data related to the 
agency dimension (such as data on community 
Data cycle phase
Review, 
consolidate, 
collect, curate, 
data
Analyse 
data using 
appropriate 
analysis tools
Translate data 
into results, 
insights, and 
conclusions
Disseminate, 
share, review, 
discuss 
results, refine 
insights and 
conclusions
Use results, 
insights, and 
conclusions to 
make decisions
Level
Macro (distal)
Meso (proximal)
Micro (immediate)
Individual (outcomes)

---

20 
]
DATA COLLECTION AND ANALYSIS TOOLS FOR FOOD SECURITY AND NUTRITION
groups that support community gardens) is 
important for the subsequent steps in the data 
cycle that are represented as columns in this 
matrix. The final step will be to determine which 
databases may be used to identify the individual 
outcomes related to the objective at hand (such 
as national health and nutrition surveys).
The subsequent columns in the data-informed 
decision-making process matrix are likely more 
challenging to fill out, but are important to 
think through, and thus, to complete, prior to 
executing a new policy or programmatic objective 
that would be inextricably linked with a budget or 
financial ramifications.
The second step in filling out the data-informed 
decision-making process matrix is to identify the 
most basic, or simple, analysis that can be done 
from the databases curated in the first column 
and to identify the appropriate analysis tools in 
which to do so. The process of identifying which 
analysis and analysis tools are necessary for the 
objective at hand serves two main purposes: 1) 
to identify which elements from the previously 
identified databases must be analysed (such as 
data on fruit and vegetable yields) and which 
analysis tools can be used for this purpose (e.g. 
any statistical software package that can rank 
values); and 2) to identify which analyses need to 
be conducted for which there are no existing data 
in the first column of the matrix (for instance, 
average distance between farmers’ markets 
and homes), which may or may not serve as an 
impetus for new data collection, depending on 
how essential the data are with regard to the 
objective at hand.
The third step is to identify how the data 
identified in the first column of the matrix 
can be translated into results, insights and 
conclusions. At the macro level, how will the 
data identified in the first column be used. For 
example, will the data be used to inform a policy 
brief, inform new FSN intervention designs, 
or will it merely be used to identify important 
national gaps (for instance, in farmers’ market 
coverage). At the meso level, what types of 
variability can be identified in the data and, 
similarly, how is this variability related to the six 
cross-cutting dimensions? For example, does 
fruit and vegetable production vary regionally 
within a country and how will that affect the 
sustainability of the objective over the long term? 
Or how will global conflicts affect the national- 
or regional-level stability of food systems that 
has proximal impacts and ramifications for 
local food systems? At the micro level, how 
can the existing databases be used to infer or 
anticipate municipal-level FSN actions (e.g. how 
should school feeding programmes incorporate 
regional, or more local, fresh products)? And, 
finally, what individual level outcomes might be 
set as targets if the objective can be addressed 
according to plan (for instance, can individual-
level fruit and vegetable consumption increase?). 
Just as the six cross-cutting dimensions need 
to be colour-coded in the first column, they 
should also be colour-coded in the subsequent 
columns, after the specific content in each 
column has been completed. Doing so will help 
the practitioner more easily identify whether the 
objective at hand really does tackle the six cross-
cutting dimensions across levels and actors in 
FSN.
The task in the fourth column of the data-
informed decision-making process matrix 
is to identify the actors related to both the 
six cross-cutting FSN dimensions and the 
originally identified objective. Thus, who are the 
stakeholders within the specific sectors related 
to macro level determinants in the conceptual 
framework (i.e., environment, technology 
and innovation, infrastructure, economic, 
sociocultural, political and institutional settings, 
education); such as key stakeholders in the food 
system at the national level, as well as trade and 
industry and the education sector. At the macro 
level, these key stakeholders may be ministers, 
while at the meso level the key stakeholders 
might come from related areas at the regional 
level. At the micro level, it is important to 
identify the key stakeholders at the municipal 
level. Finally, in terms of individual outcomes, 
when thinking about how to disseminate, share, 
review, discuss results and refine conclusions 
and insights, think about how population 
disaggregated data, for example, might be used 
to propose new programmes aimed at improving 
individual outcomes, and how agency can be 

---

[
21
1 
SETTING THE STAGE
incorporated into user-centred design processes 
to improve individual outcomes related to the 
previously defined objective.
The last step in completing the data-informed 
decision-making process matrix is to identify 
how such (anticipated) findings in the third 
column (or data translation step) might be used 
to make related decisions. While the previous 
column focused on the related actors, the final 
column in the matrix focuses on what types of 
decisions (i.e., content) might be made based 
on the previously anticipated results and key 
stakeholders at each level. Thus, at the macro 
level, national-level opportunities for innovation 
could be made or national-level procurement 
programmes could be modified. At the meso 
level, such decisions could include regional 
supply chain adaptations, industry incentives 
and penalties, and improved health sector 
messaging. At the micro level, decisions are 
made at the local level, such as through the 
local health sector and local schools. And, at the 
individual level, decisions should be aimed at 
advocacy and coalition building.
The rest of this section will highlight one 
example of how such a matrix can be used 
to guide data collection and analysis in both 
a comprehensive and simple fashion. It is 
important to recognize that what follows is only 
one example to illustrate the utility in using both 
the conceptual framework and data-informed 
decision-making cycle to guide data collection 
and analysis tools. In the supplementary 
material we have included three additional 
examples that revolve around the following 
evidence priorities and questions:
•
To identify needs for humanitarian food 
assistance for districts in Haiti using IPC as the 
data analysis tool.
•
Does the existing evidence support a national 
school feeding programme mandated through 
policy that includes 10% of school food to 
include fish/seafood products from small-
scale fisheries (SSF)?
•
How to assess a sustainable healthy diet within 
a given local context?
Although, for the purpose of this report, the 
examples are being presented from one