Methods for Literature Searches
Below is the protocol each person used while searching for papers containing mammalian sleep quotas.
- We found papers or articles that had EEG sleep data on a mammal. That meant that for this search we were not interested in papers on sleep in birds, reptiles etc. Nor were we interested in papers on sleep in cats, rats, mice, or humans as we already had those data. Finally, we were not interested in papers on sleep in a mammal unless it had some measure or estimate of what we were calling REM and NREM sleep “quotas.” Sleep quotas refer to the amount of time (out of 24 hours or out of total sleep time reported in the paper) that an animal spends in REM and in NREM sleep. Sleep quotas were derived from EEG or non-EEG (e.g., actigraphic) techniques. There were even papers where sleep quotas were measured simply with observational techniques (timing eye movements or unilateral eye closure). Because the latter two measures were less accurate than EEG measures, they received lower quality scores when the paper was scored for data quality. We, nevertheless, included data on these sleep quotas.
- We kept detailed notes (electronic format or in a written journal) of the databases that we searched and the specific searches that were completed. It was useful to print the search screen for a subset of different searches as a record of that particular search for later reference.
- We accessed specific databases like Biosis, Medline, PubMed, Web of Science, PsychLit, or some other zoologic database. We found it crucial to include the medical databases because they included the names of many useful sleep journals.
- We used the following key words (and selected ‘any word anywhere in abstract or title’): sleep, REM, NREM, EEG, electroencephalography, telemetry, actigraphy, active sleep, quiet sleep, spindles, K complex, slow wave sleep, delta wave, sleep rebound, paradoxical sleep. We limited searches to papers published since 1953 (the year REM was discovered).
- When the list of abstracts was returned, the abstract was scanned for indications that the paper contained sleep times/quotas. We looked for phrases such as “Time spent in active sleep was…”
and references to the use of an EEG.
- Whenever there was an indication that the paper reported numerical values/times for sleep states, we did the following: We saved the abstract, keywords, and full
reference in an electronic format for later use.