Add research workflow guide

This commit is contained in:
yila-research-bde9 committed 2026-10-22 21:49:32 +01:00
1 parent 0ffe3bbe0f
commit fd4dc950b4
1 file changed
+56 -2
+56 -2
View File
@@ -1,3 +1,57 @@
# research-log-habit # Keeping a Research Log You Will Actually Use
How to keep a research log you will actually use, with source-linked reading notes via https://yila.ai Every researcher has had the experience of opening a file months later and having no idea why a decision was made. Why was this variable recoded? Why was that paper excluded? Which version of the dataset produced the figure in the draft? Memory is a poor record of research, and the cost of forgetting grows with the length and complexity of a project. A research log is a simple remedy: a running, dated record of what you did, what you decided and why. This guide explains what to record, how to structure the log so that it stays useful, how to make logging a habit rather than a chore, and how a source-linked assistant such as [Yila](https://yila.ai) can help you capture reading decisions with references to the passages that informed them.
## Why a log matters
A research log serves several purposes at once. It supports reproducibility by documenting the steps that led from raw materials to results. It supports transparency by recording decisions and their rationale, which reviewers, examiners and collaborators may later ask about. It supports thinking by giving you a place to write down half-formed ideas, puzzles and hypotheses before they disappear. And it supports continuity, letting you pick up a project after a break without hours of reconstruction.
Laboratory scientists have long kept notebooks for these reasons. Researchers in fields without that tradition, including the social sciences, humanities and computational work, benefit just as much from the same discipline.
## What to record
The core of a research log is a series of dated entries. Each entry should capture what you did, what you found or produced, what you decided and why, and what you plan to do next. Not every entry needs all four, but the habit of asking these questions keeps the log useful.
Record decisions especially carefully. Decisions are the points where a project could have gone differently, and they are what people most often need to explain later. A good decision entry states the options considered, the choice made and the reason. "Excluded the 2019 survey wave because the question wording changed; see codebook page 14" is far more useful than "Dropped 2019".
Record problems and dead ends as well. An analysis that did not work or a source that proved unreliable is valuable information, both for your future self and for anyone who tries to build on your work.
## Logging your reading
Reading generates many small decisions: which papers to include, how to interpret ambiguous findings, which definitions to adopt. Recording these in the log, with references to the passages that informed them, makes your literature work far more defensible. When you decide that a study does not meet your inclusion criteria because its population differs, note the passage that describes the population.
Yila can help here. When you ask a question about a paper and receive an answer linked to a passage, you can paste the answer and its reference into your log. Later, when you need to justify the decision, the evidence is already there.
## Structure for findability
A log is only useful if you can find things in it. Use consistent dates at the start of each entry, ideally in a format that sorts correctly, such as year, month and day. Add short tags for the main topics or project components, such as data cleaning, analysis, literature or writing. If you work on several projects, keep a separate log for each or tag entries by project.
Some researchers keep a single chronological file; others use a folder of dated files or a notebook application with search. The tool matters less than consistency. Choose something you will use every day and that will still be readable in five years. Plain text and widely supported formats are safer than proprietary ones.
## Link to files and versions
Entries should point to the materials they describe. When you mention a dataset, a script or a draft, include its file name or location and, if you use version control, the relevant commit or version number. This creates a trail from the log to the actual work. When a reviewer asks which version of the code produced a table, you can answer precisely.
For computational work, consider logging the software environment as well: the versions of key packages and the operating system. Small differences in versions can change results, and recording them costs little.
## Make it a habit
The biggest obstacle to keeping a research log is not knowing what to write but remembering to write it. Attach logging to existing routines. Write a short entry at the end of each working session, before you close your files. Write an entry whenever you make a decision you might need to explain. Start each session by reading the last entry, which reminds you where you left off and reinforces the habit.
Keep entries short. A few sentences are enough for most sessions. A log that demands long entries will soon be abandoned. You can always add detail when something important happens.
## Use the log when writing
When you write up the research, the log becomes a primary source. Methods sections draw directly on recorded decisions and procedures. Limitations sections draw on documented problems and dead ends. Response letters to reviewers draw on rationales you recorded at the time. Writing from a log is faster and more accurate than writing from memory.
The log also helps with project reports, progress reviews and thesis meetings. Skimming recent entries gives you a clear summary of what has been done and what remains.
## Share appropriately
Logs are usually private working documents, but parts of them can be shared. Collaborators benefit from access to decision records, especially in team projects where several people make choices. For open science, some researchers publish cleaned versions of their logs or decision records alongside their data and code. If you plan to share, write with that possibility in mind, avoiding personal comments or sensitive information that you would not want others to see.
## Review periodically
Every few weeks, read back through recent entries. Look for patterns: recurring problems, decisions that seem inconsistent, ideas you meant to follow up. Summarise the period in a short entry. These reviews help you see the shape of the project and catch issues before they become serious.
A research log is one of the simplest tools in research and one of the most powerful. It costs a few minutes a day and repays that investment many times over, in clearer methods, stronger justifications, smoother collaboration and fewer hours spent trying to remember what you did and why.