Chapter 1 Introduction

This book is tailored for revealing key concepts of intersections between “data science” and “monitoring and evaluation” for humanitarian sector. Do not question the language since most of these thoughts and conclusions are personal, and this book acts as a notebook of my experiences, also, I am horrible at writing stuff down. No shame, that is the golden rule of being monitoring and evaluation person. Shame triggers emotions, and as an outcome one hides stuff and reflects only bright side, and that reduces lessons learned capacity for sure.

Five years in humanitarian sector as data scientist with background of industrial engineering and management information systems, I have encountered gaps and common mistakes in the sector. For some parts, I will be questioning these gaps and unknowns for building a collective knowledge, hoping that it will help the future actions. Through the book I will take a guess that reader have basic knowledge of monitoring and evaluation in humanitarian sector so I will dodge going in-depth of the aspect, and the book will focus on data analysis and tips&tricks of designing operational planning.

Humanitarian sector actors has background of “monitoring and evaluation”. There are many individuals whom got mindset and experiences towards this aspect. Yet, in the latest versions of M&E, inevitably this aspects merges with data and knowledge. One of the most experienced individuals with humanitarian data Aldo Benini reflects in his notes; “Keep it simple. You may use R or Python, but if you want to be a voice for this spectacular community, stay with excel, since most of the people are more relaxed using this.” This conclusion of years of experience is still valid. Yet, I am observing two types of individuals. On one hand you have mid-class proficient excel users, whom can conduct (descriptive) analysis. On the other hand one has very skilled individuals whom can use R or Python with their amazing machine learning skills and so on. Thus, in the sector, we do not have one common ground but two.

As for data analysis for humanitarian sector, with the motivation of donors, we are observing more “evidence-based” actions. One of the old-school WFP manager once told in a VAM training: “Back than we were disturbing in-kinds as food, hygiene whatsoever, and then come back home. Nobody was asking question neither we did care for it. We believed we did good and that’s all. But now, there is data everywhere and logic is more stronger.” This is a clear vision of experience, telling us how data become more important during years.

For red pillar, IFRC is pushing more for IM - when they are losing sight of M&E analysts and mark this concept as PMER where these individuals mostly have no idea about data analysis. So it looks like quantitative staff is IM and qualitative staff is PMER, which is a terrible idea. A data scientist can become IM, but can not become M&E data analyst, since they have to have an experience for programme design-logframes-impact analysis and understanding concept of the programme. Frankly, we did have an issues with this mindset while working with IFRC. M&E Data Scientist and IM persons are not the same thing. Through the book, one will realize this more. ICRC does cover a different path which I would prefer as well. They call this phenomenon as “Analysis and Evidence Team”. Which does serve their structure well, as they are more enveloped and their work-groups are focusing on very different topics. Yet, they have intersections such as between Protection(unit) - ECOSEC or ECOSEC (economic security) - WATHAB (water habitat). But mentality of the unit is the same. It is like having a team that makes knowledge out of ones data or any secondary data. That team creates room for being more agile, moving with evidence, not with instincts (tho sometimes it’s better, we do not make data fetish here). Their team can be interpreted as M&E Data Analyst team, under covered. If you read their documents, they are pretty similar with any other M&E guidelines and handbooks. For some programme coordinators (usually happens in National Societies) I am seeing the attitude of taking data analysis, evidence-based decision making and M&E is a layer of quality of a programme instead of considering M&E as core foundation of any humanitarian programming which should shape other operational/programmatic components. This mindset is an eclipse in humanitarian sector. These thoughts are belongs to past, what WFP manager refers to. This eclipsed-mind still lives in 50 years ago, where only aim of a programme is to deliver assistance, without any monitoring, without understanding needs, without any targeting, without any reporting at all. If you see an environment and attitude like that, just leave the ship. Donors will not approve that either.

For someone who worked for ESSN, biggest cash programme of ECHO and in humanitarian sector, I have an understanding of how ECHO interprets monitoring and what they expect out of evaluation studies. ECHO does care about evidence-based management style, and M&E. They do respect their logframes. Thus, if one puts an indicator there can expect a set of questions towards its interpretation. Also in interim reports they are very interested in calculation of MEB, how indicators talks with each other and so on. So if you are an M&E data analyst with ECHO supported programme, expect series of questions and prepare good story-telling 3-pagers. It will be a challenge! I strongly recommend to read out DG Echo Thematic Policy Document No 3 - Cash Transfers, March 2022.

Personally, I am not experienced with blue pillar (UN). Yet WFP is leading this concept and they are disaggregating more into “division of labor”* doctrine. They have monitoring unit, sided by VAM unit (vulnerability analysis and mapping). Even there is a unit collaborated by UNHCR and WFP, just for targeting. Also they have an unit studies targeting. Worked with them for targeting exercises for camps, and I learned a lot from them! No wonder why they are in a leading role, as they care and dig more into these lands. Most of the red-pillar managers does not understand what targeting is, yet other hand has an unit for it. As mentioned, personally, I do have very little information of other UN institutions. Yet, WFP is a shining role-model for sure.

This is a one, dusty overlook to humanitarian sector and data. I am smelling a trend between cash based interventions and data-knowledge tree. Huge correlation is lurking, if you check ECHOs latest Large-scale cash programme guidelines, lots on monitoring and evaluation explained. Thus, if humanitarian sector evolves more into large-scale cash, data and M&E will become more and more important. In this book, I will not dig into if it is possible to do more cash-based interventions and infrastructure of banks or stakeholders’ potential. Yet we all can agree that cash still has a big future and a room in our daily lives as humanitarians and Ukraine crisis proves that right. So lets build more towards data analysis and M&E to increase accountability, evidence-based actions and to have better programmes. Note that in this paper, I am assuming that reader has basic understanding of M&E, and stresses more on the data dimension.

A small note on division of labor, which is an important aspect. Because, if we go all in with specialized skills, we have to divide tasks into small piece and assign them to specialists..

Division of labor, the separation of a work process into a number of tasks, with each task performed by a separate person or group of persons. It is most often applied to systems of mass production and is one of the basic organizing principles of the assembly line. Considering workload and standard reporting of the unit, current flow can be interpreted as assembly line. Pros for division of labor in unit:

  • Efficient mastery
  • Quicker training
    • Considering unit will hire more people and circulation of individuals in sector, this is a very important key point.
  • Productivity
  • Innovation

Cons for division of labor in unit:

  • Boredom of repetition
    • Roles will not assign to each site and each site will work horizontally. Thus we will eliminate this con.
  • Interdependence
    • Not a con for our unit. Collective work is a must.
  • Lack of responsibility
  • Better carrier chances and additional studies will eliminate this con.

Some possible unit ideas for the future:

  • IMER (Information management, monitoring, evaluation, reporting)
  • VANA (Vulnerability and Needs Assessment Unit)
  • MER (Monitoring, evaluation, reporting –We have PMER but planning should be separate set of skills and title)

1.1 Audiance

The primary audience for this book is data science / data analyst practitioners whom works for monitoring and evaluation units, from organisations directly involved in the design, implementation, monitoring, and accountability of projects using cash and vouchers to deliver humanitarian relief. For example:

  • Monitoring and evaluation data analysts
  • PMER staff with data skills
  • IM teams involved in humanitarian data analysis
  • MEAL specialists etc.

The secondary intended audience is other humanitarian stakeholders involved in advancing CTP monitoring and evaluation practices. Thus, may include TPM institutions as well. Please note that monitoring and evaluation is an expensive exercise, and things explained-commented in this doodle are mostly orients for large-scale programmes. Thus, even this mainly concerns big programmes with big resources, still it is a good-to-know for any analysts in the sector.

1.2 Why Is This a Book Needed?

As mentioned in introduction, data analysis within monitoring and evaluation is a growing aspect. Many new colleagues are taking part, some of them has skills towards data analysis and some grades up from field teams, some are just new to the sector or to data analysis. In this book, I will mention about M&E aspect and give some examples through operation, data cleaning, data analysis and basic examples of calculating humanitarian indicators (few important ones usually involves with cash-based interventions) with R. Thus, my belief is to make this book an honest starter for my colleagues, to read and learn from each other. Having no other hands-on guidance and written experiences was missing for me, so I am trying to fill that gap. Of course, every institution has their SOPs, but these SOPs are usually serves and limited with their current context.

1.3 Book Structure

This book mainly runs in two pillars; i) Monitoring and evaluation design in humanitarian sector ii) Data Science methods. In this book, for every section, I will set a scenario, a scene to reflect reality and then will take action with coding and explain how to implement studies to a real life situation.

Data sets will be random, the ones comes with packages. Anyhow, for different exercises I will be using random data sets from some packages but main concept will be same with humanitarian data. Here, mentality of tidy data comes into play, where every row stands for analysis unit and every column stands for variables.

Structure of the Book