proposal for testing narratives: early-wake up with single narrative prompt

[contralto]

Background:

• External stimuli that are dream relevant seem to make their way into (and thus manipulate) dreams more frequently, although this is highly variable across studies/people (Salveson et al. 2014)
• Wake-back-to-bed (WBTB) is a method where people are awoken for 30-120 min. Often mixed with MILD (Mnemonic Induction of Lucid Dreams), where people basically rehearse experiencing a dream and becoming lucid before going back to sleep; seems like PM naps were most effective for this according to this 2012 review of lucidity induction methods by Stumbrys and colleagues: https://www.sciencedirect.com/science/article/pii/S1053810012001614?via%3Dihub#s0150
• Benefits are people early are more likely to 1) dream, 2) lucid dream, and 3) dream about whatever you prime them with during this wake.
• For our purposes I want to try something that's less cumbersome in the morning and doesn't require a nap later, but we can think about where to go if this first design doesn't seem to work. The brief dream summary (see below) isn't founded in the literature, so we can try it but I'm sure there are many reviews out there with alternative, tested methods

Proposed design:

• ~5-6 hours post-sleep start, try to gently wake participant (music from meditation? wake-up script?)
• have them confirm they're awake (maybe some ongoing screen tapping requirement so we know they don't fall asleep immediately?)
• during this brief wake period, have a brief synopsis (maybe a minute max) of the dream they should try to experience – this would be a primer for the actual narrative that we play during the final sleep cycle
• allow them to go back to sleep
• monitor heart rate, start cuing ~15 min after the priming period, if heart rate spikes, stop and try again 3-4 minutes later (or whatever is typical)
• upon wake, collect a voice memo of any general dream content and/or specific details they can recall – emphasize that no detail, feeling, or broad takeaway is too small to report, even if it's fuzzy or they're uncertain
• would also be nice here to collect some mood data and see if it aligns with how we've categorized the data? in this case we'd likely want to select the prompt ourselves so participants aren't biased toward having a specific emotional outcome (e.g. feeling calmer) because they chose a calming prompt
• ask if they were lucid during any dream they can recall

TODOs:

• implement timing-based cuing with heart rate monitoring
• wake confirmation
• 30s-1min narrative summaries

Other thoughts:

• might need them to listen to the narrative first? this could look like before bed with some audio cue and then we could play that audio cue during sleep, this could guide a MILD intervention?
• with this narrative structure it seems we want to maximize the amount of content that actually makes its way into dreams. one of our thoughts is lucid dreaming is better for this, so we may also need to implement some lucidity induction training (but we can think about that after we see whether people experience the narratives during this period)
• because dream interventions are so unpredictable, we'll likely need a pretty large sample size once we have a design we're confident in