Background:
Ye Generic Breath Meditation:
1. breathe in for A seconds
2. hold breath for B seconds
3. breathe out for C seconds
4. hold breath for D seconds
There is a claim made on breath meditation that the in/out timing ratio (A/C) affects the effects of the breathing exercise - namely, that a higher A/C will create more invigorating effects, and a lower A/C will create more relaxing effects.
A normal timing that most people can do is 5/1/5/1 (A~D).
A claimed 'energetic' timing would be 5/0/2/0. A claimed 'relaxing' timing would be 2/0/5/0.
Query: Is this claim (of differential effect) true?
Proposed test method:
Subjects: Random people walking around downtown Berkeley
Equipment: A fingertip pulse oximeter (FPO), stopwatch, coin, two chairs, and clipboard w/survey & consent
Method:
1. Approach random person. Ask: "Would you like to participate in a research experiment about breathing? It'll only take 5 minutes."
2. If agreed, have them sit down comfortably in the chair, and read and sign a simple consent form (stating that it's a study to investigate the effects of different breathing methods, no deception involved).
3. Measure and record pulse and SpO2. (Measuring after they sit & read will help to get closer to a baseline level, i.e. not immediately post-exercise of walking up stairs or the like.)
4. Secretly flip a coin. "Heads" is the 'energetic' timing, tails the 'relaxing' timing.
5. Tell them to breath in A seconds, out C seconds, per the toss; have them count it out subvocally and keep it up for 3 minutes.
* Alternately to 4-5 (if experimenter blinding is an issue), have them flip a coin themselves and pick an index card with instructions with "heads" or "tails" written on the back.
6. Measure & record pulse and SpO2 again.
7. Have them fill out survey:
1. current perceived state (relaxed ... energetic, tick-the-line style)
2. perceived change in state since starting the exercise (ditto as delta)
3. prior familiarity with meditation generally (1-5)
4. prior familiarity with breath meditation (1-5)
5. age
6. gender
7. initials
8. Thank for participation, debrief explaining details as above, & answer whatever questions they have. Move on to next subject.
Discard any results with #3 or 4 rated higher than 2 to prevent tainting via priming.
If the claim is true, then there should be a significant correlation between the BPM, SpO2, and/or perceived current/delta state and which group they were assigned to.
If the claim is false (or clinically insignificant), then there would be no significant correlation.
The experiment may need to reperformed later using only a single measure that performs well in the first round, to prevent sharpshooter fallacy.
Suggestions for improvement?
Ye Generic Breath Meditation:
1. breathe in for A seconds
2. hold breath for B seconds
3. breathe out for C seconds
4. hold breath for D seconds
There is a claim made on breath meditation that the in/out timing ratio (A/C) affects the effects of the breathing exercise - namely, that a higher A/C will create more invigorating effects, and a lower A/C will create more relaxing effects.
A normal timing that most people can do is 5/1/5/1 (A~D).
A claimed 'energetic' timing would be 5/0/2/0. A claimed 'relaxing' timing would be 2/0/5/0.
Query: Is this claim (of differential effect) true?
Proposed test method:
Subjects: Random people walking around downtown Berkeley
Equipment: A fingertip pulse oximeter (FPO), stopwatch, coin, two chairs, and clipboard w/survey & consent
Method:
1. Approach random person. Ask: "Would you like to participate in a research experiment about breathing? It'll only take 5 minutes."
2. If agreed, have them sit down comfortably in the chair, and read and sign a simple consent form (stating that it's a study to investigate the effects of different breathing methods, no deception involved).
3. Measure and record pulse and SpO2. (Measuring after they sit & read will help to get closer to a baseline level, i.e. not immediately post-exercise of walking up stairs or the like.)
4. Secretly flip a coin. "Heads" is the 'energetic' timing, tails the 'relaxing' timing.
5. Tell them to breath in A seconds, out C seconds, per the toss; have them count it out subvocally and keep it up for 3 minutes.
* Alternately to 4-5 (if experimenter blinding is an issue), have them flip a coin themselves and pick an index card with instructions with "heads" or "tails" written on the back.
6. Measure & record pulse and SpO2 again.
7. Have them fill out survey:
1. current perceived state (relaxed ... energetic, tick-the-line style)
2. perceived change in state since starting the exercise (ditto as delta)
3. prior familiarity with meditation generally (1-5)
4. prior familiarity with breath meditation (1-5)
5. age
6. gender
7. initials
8. Thank for participation, debrief explaining details as above, & answer whatever questions they have. Move on to next subject.
Discard any results with #3 or 4 rated higher than 2 to prevent tainting via priming.
If the claim is true, then there should be a significant correlation between the BPM, SpO2, and/or perceived current/delta state and which group they were assigned to.
If the claim is false (or clinically insignificant), then there would be no significant correlation.
The experiment may need to reperformed later using only a single measure that performs well in the first round, to prevent sharpshooter fallacy.
Suggestions for improvement?
Comments
To save time, why not screen for familiarity with meditation before having them do it? And why not randomly assign subjects before testing? You'd probably also want to get a baseline mood survey immediately before starting to test, not afterward.
Definitely screen for COPD, other respiratory stuff, and circulatory problems, particularly in smokers, older folks, and women. Don't use kids, adolescents, or the elderly because of different respiratory rates.
In general, the "random people walking around" doesn't seem like a very efficient way to get subjects. Most of the people I'd imagine would have time are likely to be too elderly or too young.
If you're looking to publish, you'd also want to get IRB approval. Are you human subjects-approved anywhere, or do you know someone who is?
Unless you're suggesting it'd be potentially harmful, why screen for respiratory variance? I'm not doing inter-subject number comparisons, just delta %ages; comparing raw data would obviously be invalid.
No IRB I know of will talk to non-affiliates, unfortunately.
Screening for respiratory and circulatory variance is important so that you weed out the people who'd throw off the pulse oximetry results.
Could you become an affiliate? It's not uncommon here and at a few other places I know for alumni to come back and hang around as research assistants. If you can't get IRB approval, especially for human subjects, publishing is going to be a big issue.
By the way, what stats would you run on this?
Why would respiratory/circulatory variance throw off the pulse oximetery results between that same person's before & after?
Stats: *shrug* simple correlations to the group choice. First round, see what correlates. Second round, do just the one metric, same deal. Should be pretty straightforward.
Ignoring all that, do you really want to not screen for circulatory/respiratory problems on a study involving manipulating respiration? I can't think of an IRB that would like the risks that creates.
*shrug* Fairly; I was just curious as to the particular procedures you'd run, and where you'd put your sig levels.
Would such questions bias/prime it? How old or young is too old or young?
As for biasing or priming, no. You're not doing anything that involves deception, and subject screening is a part of just about every study. Your procedure's going to be very clear to your subjects, and your hypothesis will probably be apparent to at least half. Besides, you could be asking about medical histories for any number of reasons; I doubt anyone would immediately make the respiratory history->meditation leap.
At the end, sure.
But yeah, I can just give 'em a page that lists disqualifying conditions and have 'em say whether they're in any of those categories.
Whitecoat effects can go either way, for what it's worth. Sure, you often get ones who want to help. And you get ones who don't. It's not something researchers lose a great deal of sleep over. Most studies don't involve deception, and don't screen for the subject guessing the hypothesis.
*shrug* Overall, I think your concerns about priming are overstated. They'll likely guess when they take the second mood survey, at latest. If you're really that concerned, you can always throw in some dummy surveys as well.