Nervous System Studio
A private log for noticing which settings actually help you, and which
do not. You record where you were and how you felt; after enough entries
it compares settings against each other and tells you what it found.
It needs JavaScript, and it needs it for a reason
Every entry stays in this browser. Nothing is transmitted, there is no
account, and there is no server holding your logs. That means the
comparison has to run here, in the page. With scripts disabled there is
nowhere for it to run.
What it will and will not tell you
It only reports a comparison when the data can carry one. Four numbers
decide that. They are the difference between this and a horoscope, so it
is worth saying where each of them comes from rather than asking you to
take them on faith.
At least 4 entries before anything is compared
An unusually bad evening tends to be followed by a less unusual one
whether or not you change anything. That is regression to the mean, and it
is the reason almost any remedy tried at your worst moment appears to
work: the measurement had nowhere to go but back toward your average.
Barnett, van der Pols and Dobson set the problem out plainly in
Regression to the mean: what it is and how to deal with it,
International Journal of Epidemiology 34(1):215-220
(doi:10.1093/ije/dyh299).
Four entries is not a cure for it. It is the smallest number at which a
first extreme entry stops dominating everything the tool would say.
Each setting seen at least 2 times
A group of one is a single evening wearing a label. Nothing in the data
separates the setting from the day it happened on: the argument you had,
the sleep you did not get, the fact that it was a Tuesday. Requiring two
means a claim about a setting has at least been true twice.
A gap of at least 1.5 points on the 0 to 10 scale
Self-report moves on its own. Momentary ratings of one’s own state
carry substantial within-person variability from one report to the next,
which is much of the point of the method reviewed by Shiffman, Stone and
Hufford in Ecological Momentary Assessment, Annual Review of
Clinical Psychology 4:1-32 (2008,
doi:10.1146/annurev.clinpsy.3.022806.091415).
A difference of half a point between two groups is not a finding about
your environment; it is the scale breathing. 1.5 is set above that drift
on purpose, which means the tool will stay silent about real effects that
happen to be small.
At most 4 observations shown at once
Compare enough pairs and some will clear any threshold by chance alone.
This is the multiple-comparisons problem that Benjamini and Hochberg
addressed in Controlling the False Discovery Rate, Journal of
the Royal Statistical Society, Series B 57(1):289-300 (1995,
doi:10.1111/j.2517-6161.1995.tb02031.x).
A capped list is a blunter instrument than a false-discovery-rate
correction, and it is used here because it is honest about being blunt:
showing you the four strongest comparisons rather than every pair that
happened to clear the line.
Why the confidence grade stops at Moderate
The grades run Preliminary, Emerging, Moderate, and Moderate is the
ceiling. No amount of logging reaches a higher one. The reason is that a
log like this is a single-case design, and the standard for treating one
as evidence is demanding: the What Works Clearinghouse
Procedures and Standards Handbook, Version 5.0 requires
at
least three demonstrations of an effect at three different points in
time. Keeping a journal and comparing groups afterwards does not
produce that, however many entries accumulate. The ceiling is where it is
because the design cannot support a higher claim, not because the tool is
being modest.
The log is not a neutral instrument
One thing worth knowing before you start: monitoring your own progress
tends to change it. Harkin and colleagues found this across 138
experimental studies in Does monitoring goal progress promote goal
attainment?, Psychological Bulletin 142(2):198-229 (2016,
doi:10.1037/bul0000025).
So the act of keeping this log is itself one of the things acting on you,
and the tool cannot separate that from whatever your environment is doing.
That is not a reason to avoid logging. It is a reason not to read a rising
score as proof that the lamp you bought was the cause.
Each finding shows both groups, how many entries are in each, the
difference between them, and how fragile it is: the score one further
entry would have to fall below for the finding to disappear. If a single
additional evening would erase a finding, the tool says so on the finding
itself.
What it is not
It is not a diagnostic instrument, not medical advice, and not
evidence about anyone other than the person keeping the log. It compares
your own entries with each other. That is the whole claim.
RN Builds