Publications guide

An analysis piece presents an argument driven by ACLED data, drawing upon visuals. It provides a medium for readers to engage with ACLED data and understand how these offer unique insights into dynamics and patterns of conflict. 

This guide aims to provide a concise overview of different types and formats of analysis, outlining expectations, best practices, and relevant workflows authors should consider before and during writing.

Do’s and don’ts in analysis work

There is no single set of rules to produce good analysis work — whether you are writing a piece, assembling a presentation, or responding to questions. Authors bring different backgrounds, expertise, and styles that all contribute to making each output unique. This variety is one of ACLED’s strengths, allowing different outputs to be produced.

ACLED’s approach to conflict analysis rests on some key principles:

1. Underscore the importance of data-based evidence and its complementarity with other forms of evidence and context

2. Be clear about what data can and cannot say

3. Produce analysis rigorously, without explicit bias or prejudice

4. Present the data and evidence in an accessible and robust way

5. Examine the components of the conflict as thoroughly as possible

6. Present the harms and risks without obscuration or exaggeration

To incorporate these principles in your analysis work, please remember to apply the following:

Highlight ACLED’s added value: All analysis should leverage the added value of ACLED data to provide new original insights into political disorder. To this end, analysts are encouraged to showcase the richness of the dataset using the most appropriate indicators and not limit themselves to the most immediate measures of conflict severity, namely conflict events and reported fatalities. Analysts could instead integrate alternative indicators, including interactions, fragmentation, diffusion, deadliness, etc., or link ACLED data with compatible datasets to illustrate compelling trends.

Use data strategically: Conflict data are ACLED’s bread and butter. Analysts should be aware of the data’s limitations and biases and avoid drawing conclusions not adequately backed by evidence. Avoid cluttering a text or a presentation with disconnected data points without providing the reader with the necessary time and space to understand their significance. Analysts can instead develop a compelling data-driven argument by limiting the data points in an article to those relevant to support the key argument (e.g., reduce the noise) or highlighting a single key takeaway/data trend supported by other data points (e.g., zoom-in/zoom-out).

Understand your audience: Contemporary conflicts are complex. Our job as analysts is to untangle this complexity and make it accessible to a wide range of audiences. It is essential to understand whom you are addressing before producing any analysis and consider the trade-off between comprehensiveness and detail. A piece for a non-specialist audience or on a country/theme where ACLED’s work is limited may require providing some essential background information, which may not be needed in a high-level briefing or an in-depth analysis for a more specialist audience. In other words, always ask yourself whether your audience needs more context to understand what you say or write.

Bring clarity to crisis: Our job at ACLED is down to this key mission. Bringing clarity to crisis requires acknowledging the complexity of contemporary conflicts and using our data to bring light to key trends and dynamics. Analysts should avoid stating the obvious (“the conflict is complex”) and instead focus on distilling the key elements and incorporating the data your audience needs to understand the conflict. 

DON’T: Conflict continued in 2024 at heightened levels. (Focus on change preferably, or contrast stability with changing trends)

DO: Despite overall stable levels, an uptick in rebel activity threatens the survival of the regime.

DON’T: Violence against civilians increased by 5%. Battles grew in intensity, driving a 10% increase in reported fatalities. Likewise, rebel activity increased by another 7% from the previous year. (Adding seemingly disconnected data points does not help the reader grasp the essence of your analysis.)

DO: Nationwide, political violence increased by 2%. Yet, violence targeting civilians grew by 5%, driven by a surge in militant activity in the country’s northernmost border region.

DON’T: In December alone rebels seized four new cities from the government without fighting. (Why is this data point relevant? State the significance of these events to highlight the value of the data.)

DO: In December alone, rebels seized four new cities from the government without fighting. This underscores how their quick advances are finding limited resistance from government forces.

Analysis-led publications

Analysis-led publications are outputs approved and overseen by the Analysis department. They include a host of elective and grant-linked publications typically planned for in the annual Analysis agenda.

All Analysis-led publications require the appropriate use of ACLED data. Authors are encouraged to combine ACLED with additional data and evidence, including external data sources and qualitative insights (e.g., interviews, original research, etc.)

There are two main types of Analysis publications at ACLED: 

• Descriptive analysis is the type of analysis that helps describe, show, or summarize data points in a constructive way. In descriptive analysis, you can highlight averages, edge cases, and differences among groups, spaces, and trends, making comparisons and drawing inferences and conclusions. Data are front and center in your analysis.

• Explanatory analysis is the type of analysis that aims to answer a key question, identifying trends, patterns, and key contexts. Understanding the why of a given dynamic is the ultimate aim of this type of analysis.

Formats and bands

• 900-1,500 words

• 1,500-3,000 words

• 3,000-5,000 words (+ pdf)

• 1,200-1,800 words (observatories)

Workflows

Analysis pieces are approved by the Head of Analysis according to the general guidelines outlined in the Analysis agenda. The Head of Analysis may either assign or approve analysis pieces. 

Each year, regional analysis teams submit regional analysis proposals as inputs for integration into the annual Analysis agenda. These proposals can then be added to or adjusted throughout the year before quarterly regional analysis meetings. Proposed analysis pieces will then be discussed and approved during those meetings. Once a proposal is approved, the Analysis Manager will upload tasks onto Asana. 

Each analysis piece will go through five phases of development and review:

• Outline development phase

• Draft development phase

• Editorial review

• Final review

• Finalization and publication

Within each phase, a series of sub-tasks will need to be completed. These sub-tasks will be automatically generated when an analysis task is added to Asana. Sub-tasks from draft submission through to publication will be automatically assigned dates based on the proposed publication date. Earlier sub-task deadlines covering the outline and draft development phases are to be manually filled out by the author in discussion with their primary editor. 

The first phase in the analysis development process is the outline development phase. During this phase, authors are expected to carry out their research (including any supplementary research, such as interviews) and analysis work to develop a comprehensive outline of the piece. The author should use this template in developing the outline. Once submitted, the outline will be reviewed by the Analysis Manager with support from the publications team, before being reviewed and approved by the Head of Analysis. The outline will form the baseline for the final draft of the report, with any substantial deviations from the outline during the draft development phase shared with the Analysis management team.

During the review phases, please use “Suggesting mode” for all edits. Remember to address or respond to reviewer comments before passing the document back to the reviewer. Please accept edits or leave a comment. The reviewer is responsible for resolving comments once it has been addressed and should ensure the document is void of any pending edits or comments when it is moving between reviewers.

NOTE: Work must be produced to firm assigned deadlines. Please remember that while you may conduct most of your work independently, ACLED is a team effort and delays in submitting your work may result in organizational and reputational consequences. It is therefore essential that if an author expects not to be able to meet a deadline, they inform the Analysis Manager as soon as possible.

 Rules for Using Asana

• All analysis outputs are required to be represented in an Analysis Asana pipeline. 

• All major draft milestones are represented with sub-tasks and must be completed.

  • Draft and review-related sub-tasks will be automatically created when tasks are added to the pipeline.

• For each phase, the reviewer must provide their approval before a task can be moved along to the next section.

• Senior Analysts are responsible for the ongoing maintenance of their regional Asana pipelines.

Some essential writing resources

Authors should follow the ACLED house style. The ACLED Style guide is available on ACLED Net.

The Analysis and Writing tips hubs on ACLED Net include important resources to help you avoid some common mistakes when analyzing data and writing.

The Glossary lists several commonly used terms that authors should use appropriately at all times.

The Visual style guide aims to ensure consistency across all visual outputs. It provides an overview of ACLED’s brand identity policy and introduces guidelines for planning and selecting visualizations and best practices to keep in mind when building graphic assets.

Authors should avoid plagiarism at all times. Plagiarism refers to “the use, without acknowledgment, of the intellectual work of other people, and the act of representing the ideas or discoveries of another as one’s own in written work submitted for assessment.” It includes copying sentences, phrases, or expressions or paraphrasing without adequately acknowledging the source. A source must be provided for copied or paraphrased content, while verbatim quotations must be reported in inverted commas, directly acknowledged in the text, and a footnote citation to the source must be provided. All authors are expected to refrain from plagiarism and are encouraged to use the following resources:

• A Guide to Plagiarism and Paraphrasing

• Steps to Avoiding Plagiarism | UAGC Writing Center

• How to Paraphrase | Step-by-Step Guide & Examples 

Self-check list prior to final review

• Can I summarize the argument of my piece in one sentence?

• Is my argument supported by ACLED data?

• Is my argument outlined in the introduction and recapped in the conclusion?

• Are there sections in the piece that are not relevant to the overall argument made in the piece?

• Are there any inconsistencies or contradictions in my argument to correct?

• Does the piece meet the formatting requirements?

• Is the use of measures and terminology consistent throughout the piece (e.g., actors, event types/sub-event types, time frame)?

• Has the use of these parameters been justified?

• Do the visuals illustrate and enhance the related analysis points/trends discussed?

• Have I submitted the graphics pack to the Research Analyst?

• Have I incorporated all visuals effectively into my analysis, using time frames and measurements in-text that correspond with those in each visual? 

• Have I provided all relevant references?

• Have I consulted and followed the relevant guides, including this one and the Style guide?

Further readings: 

• Style Guide 

• Regional Overview 

• Glossary 

• Visual Style Guide