Scientific Laboratory Notebook: Build a Traceable Experiment Record

Make the experiment reconstructable

A scientific laboratory notebook should let another qualified reader reconstruct what was planned, what was actually done, what was observed, which files and samples belong to the work, how calculations were performed, and where deviations occurred. Use dated entries, stable experiment and file identifiers, contemporaneous corrections, and links from raw evidence to later figures and report claims.

Copyable notebook entry fields

Use the format required by the laboratory or course. These fields form a traceability checklist; they do not replace institutional rules, witnessed records, electronic-notebook controls, or patent and legal advice.

Record enough context beside every raw file to reconnect it later: experiment and sample identity, instrument or software, settings, time, unit, and notebook location. A file name alone is fragile. A notebook sentence without the raw file is equally incomplete. The link between them is what allows a later calculation or report value to be audited.

• Date, experiment ID, objective, protocol or method version, and responsible person
• Materials, sample IDs, instrument identity, settings, environmental conditions, and actual sequence
• Contemporaneous observations, raw data and file identities, calculations, units, and software details
• Deviation or correction, original readable value, reason, time, and required sign-off
• Decision, unresolved question, follow-up, and links to report figure, table, result, or claim IDs

A traceable experiment-entry example

Entry EXP-024 records the objective and protocol version before work begins. During the run, the student records actual times, sample IDs, instrument file names, and an unexpected temperature change without overwriting the planned condition. Later calculations cite the same file identities and the report figure points back to the entry.

Record performed work as it happens

Open each experiment with its date, stable identity, objective, protocol or method version, materials or sample identifiers, and the decisions already made. Record actual work chronologically. Preserve original observations and attach file names or storage identities while they are created, so later analysis does not depend on memory or a folder whose contents cannot be connected to the notebook entry.

Trace in both directions. Pick each principal report value and find its notebook entry, raw file, calculation, unit, and any correction or exclusion. Then pick notebook outputs and identify whether they were analysed, excluded with a recorded reason, or left unused. Never repair a missing trail by inventing a measurement or retroactively rewriting what occurred.

• MIT OpenCourseWare: laboratory notebook guide
• MIT OpenCourseWare: lab notebook instructions

Preserve corrections, deviations, and raw evidence

Correct transparently rather than rewriting history. Keep the original value readable, add the correction, reason, date, and responsible person according to the course or laboratory's rules. Distinguish raw observation from calculation and interpretation. Unexpected conditions, failed runs, negative findings, excluded values, software settings, and deviations belong in the record even when they complicate the planned narrative.

Trace notebook evidence into the report

Before drafting the report, manually trace every reported value, figure, table, and material conclusion backward to the notebook and associated data. Then trace notebook outputs forward to show which were analysed, excluded with a recorded reason, or left unused. Once that evidence has been written into the report, LabCrux is limited to checking consistency inside the existing report; it does not inspect the notebook or raw files. Recordkeeping authority, scientific judgment, institutional policy, and any legal consequence remain with qualified humans.

Scientific and institutional limits

Consistency does not prove scientific correctness, and observations or results must never be invented or changed.

• Entries distinguish planned and performed work
• Raw observations and file identities are preserved
• Corrections and deviations remain visible
• Report values trace back to the notebook

Sources and discussion

• MIT OpenCourseWare: laboratory notebook guide (official)
• MIT OpenCourseWare: lab notebook instructions (official)

Community posts describe individual experiences and questions; they are not treated as universal proof.

Related resources

• /
• /blog
• /tools