Data collection and analysis tools for food security and nutrition

 Rigorous analysis tools should never be, or 
even appear as, “black boxes”, especially to those 
who will be called to action by the results of the 
analysis. 
Formalizing the rules to be followed in the 
analysis of data is one mechanism that reduces 
the risk of different conclusions being drawn 
by different analysts, who may be asked to 
answer the same question, using the same 
set of available evidence. The goal of explicit 
formalization is to increase the extent to which 
results from the analysis of data are objective 
and trustworthy, especially where data are 
scarce or where there may be lack of consensus 
around the constructs involved. This aspect 
is becoming especially problematic with the 
diffusion of automated, algorithm-based data 
processing systems, powered by 
artificial 
intelligence (AI)
and 
machine learning (ML)
. 
Use of these new systems in areas of immediate 
consequence for human health and well-being 
raises important concerns regarding how 
trust and transparency can be sustained. As 
noted by Burrell (2016) and discussed by Oliver 
(2021), algorithmic decision-making used in 

---

12 
]
DATA COLLECTION AND ANALYSIS TOOLS FOR FOOD SECURITY AND NUTRITION
data analysis can be opaque for three reasons: 
because of deliberate decisions by the owner 
or creator of the tool to protect intellectual 
property, because of a lack of technical skills 
and knowledge among the users of the tool 
and those who are bound to be affected by 
the decision, and because these systems 
may actually be “intrinsically opaque due to 
the characteristics of many state-of-the-art 
machine-learning methods” (Oliver, 2021 p.61).
DEFINING DATA GOVERNANCE
Data governance has been defined in different 
ways. In this report, we define it as a:
 Globally relevant set of principles, strategies, 
policies, regulations and standards developed 
by institutions to collect, manage, share and use 
data. 
By establishing rules and standards, data 
governance aims to enable the broadest possible 
data sharing, so that data can be used effectively, 
while ensuring the protection, integrity and 
transparency of data systems. To be effective, 
institutions responsible for developing data 
governance frameworks, including international 
organizations, national governments, academia, 
the private sector and civil society organizations 
(CSOs), must act in a coordinated manner. 
These institutions must reinforce collaboration 
to establish and maintain data systems that can 
inform the design of interventions and policies 
needed to address FSN challenges. We address all 
these issues in Chapter 5.
A CONCEPTUAL 
FRAMEWORK TO INFORM 
DATA COLLECTION AND 
ANALYSIS TOOLS FOR FOOD 
SECURITY AND NUTRITION
One of the main objectives of this report is to 
support more effective FSN decision-making, 
by providing guidance on the most appropriate 
ways to use and analyse data. In order to do this, 
it is necessary to, first, frame the many factors 
which influence FSN (and therefore determine 
which data are needed for decision-making) and, 
second, emphasize the data-informed decision-
making process from a conceptual standpoint 
Drawing on previous work by the HLPE-FSN 
and others (Bronfenbrenner, 1979; HLPE, 2017, 
2020), our conceptual framework illustrates 
how multiple levels of factors influence the food 
security and nutritional status of individuals and 
aims to help guide data collection and analysis. 
The conceptual framework can help define 
pathways to build evidence for decision-making 
through setting research priorities, with the goal 
of enhancing the FSN of individuals, households 
and communities.
One of the challenges for FSN is the complexity 
of the concept of food security. As noted by the 
HLPE-FSN:
 The concept of food security has evolved 
to recognize the centrality of