Hi, First of all, thank you for sharing this valuable dataset. I just wanted to clarify something about the timestamps. In the data dictionary page for sleepiness checker, the timestamps are stated as "Date & time of survey completion in UTC". And I wanted to confirm this for a couple of reasons: 1. The format for many of them appear to be local time(e.g.: '2016-07-21T23:14:53-04:00'), while comparatively fewer are actually in UTC /Zulu (e.g.: '2018-11-11T12:18:00Z' ). With the first format, if it is UTC, I would presume (according to ISO 8601 standards) that local time is derived as datetime - the offset, which is given as 4 hours. But I suspect in this case the datetime itself may already be local time, not UTC. 2. When looking at the times in light of the demographic data, it appears that what is presented is in fact local time, and not UTC. For instance, the wake time distribution for those who work 9-5 peaks at 6am when treated as local time, which to me makes sense. However, if I treat the timestamp as UTC and account for the associated offset value, then the distribution peaks at 1am, which is inconsistent with what we'd expect. So could you please confirm for me if timestamps with this format ('2016-07-21T23:14:53-04:00') may in fact be local times with added time zone info (i.e. -04:00) and not actually UTC sampled? Worth noting that the KSS data fits the expected time of day variation when the timestamp is treated as local time, not UTC. Regards, Brook

Created by Brook Shiferaw brook.shiferaw
Hi Brook, First of all, I apologize for the extremely late response to your question. I can confirm that the timestamps in question are local with added timezone information, and not actually UTC. As such, the accompanying descriptions are incorrect, and I am going to change them right now to eliminate further confusion. Thank you very much for catching this and bringing it to our attention. Sean

Question on the timestamps page is loading…