3.1.2 Procedures for how and when to record data
A. Background & Definitions
Electronic data: Output from an instrument and/or its software application. The information needed for review and/or analysis of electronic data may include metadata, such as machine settings, software type and version, etc.
Hard copy/paper data: Records may also be made on a non-electronic medium (e.g., paper notebook, chart recording).
Experimental Data consists of two components:
- Primary data (raw data): All original recordings that are the result of the original observations and activities in an experiment.
- Secondary data (derived data): outputs of the analysis of Raw Data.
The user should define how to record the data of each experiment. As far as possible, data should be recorded without delay after their generation.
B. Guidance & Expectations
- The creation of an experimental record should be concurrent with the initiation of an experiment.
- Ensuring quality and integrity of the experimental records will permit researchers / peers to appropriately evaluate and, if necessary, repeat the reported experiments. Thus, it is critical to ensure that experimental records are attributable, legible, contemporaneous, original, accurate as well as complete, consistent, enduring and available (commonly referred to as ALCOAC+):
- Attributable: The author(s), all individuals who participated and/or contributed to the experiment, including, where applicable, recorder(s) must be clearly identified, so that the data can be traced, by name and date to each individual s contribution.
- Legible: The record must be readable and recorded in/or on a permanent medium (paper or electronic).
- Contemporaneous: The newly-generated/collected data, as well as new scientific discussions and ideas, should be recorded at the time of the observation.
- Original: The first recording of the data should be retained.
- Accurate: The recorded observations must be true and accurate.
- Complete: Records should be complete to ensure their traceability, prompt and accurate retrieval, and to enable exact reconstruction or review of the work described. Therefore, each experimental record should have a unique identifier in accordance with the applicable procedure (e.g. SOPs) and all related experimental records must be cross-referenced.
- Consistent: The data should be chronologically arranged, including a date and time stamp that is in the expected sequence.
- Enduring: The material used to record the data should be in a manner which will last a sufficiently long duration of time without losing the readability.
- Available: Data should be accessible whenever needed, over the life of the data. Availability ensures the data meets it’s use, since it can be applied when the need arises.
Extra care has to be taken:
- All changes/alterations to data already inputted into an experimental record should be clearly described and explained.
- Subsequent changes to source data should not obscure the original data and should be explained. The changes must include identification of the person making the change and the date to make the change attributable/traceable.
- If changes are made in an open experiment, they may be made directly in the experimental record. If changes need to be made after an experimental record is closed, the experimental record can be reopened or another experimental record, citing the original experimental record, can be opened. It is recommended that both records are signed upon completion and witnessed.
PLEASE DO NOT FORGET
- To minimize chances of loss or damage to the experimental data, to the extent possible, they should be recorded directly into the data recording platform.
- For computer application(s) used to collect, analyze, plot, summarize, or otherwise characterize experimental data, provide the following information: Name, version and vendor of the application and where in the experiment or record-keeping the application was used (e.g., tabulating data).
- Processed original data may also be recognized as raw data when original observations cannot be stored for technical reasons, e.g.:
- a research tool conducts pre-processing of original observations (example: movements of a rat in an open field are recorded by means of the photobeam breaks; research software may present the raw data as a movement track or a calculated distance traveled rather than a sequence of photobeam breaks);
- a research tool records data in a specific format that may or may not be readable at a later time point (e.g. if the license to use this research tool expires) and therefore pre-processing supports long-term accessibility of the original observations;
- a research tool generate exceptionally large volumes of data that are technically difficult to store without pre-processing to reduce the storage volume (e.g. imaging data).
- Guidelines for SCIENTIFIC RECORD KEEPING in the Intramural Research Program at the NIH 
- Creating a digital signature in Microsoft Office applications
back to Toolbox