Good question. For most measurements, we have a digital or analog readout. For
subjective measurements, like well-being, we could use a (self-rated) scale of 1
to 10,
both before and after our morning walk. This may strike some people as being
unscientific.

Let's put things into perspective. There are 2 kinds of quantitation. The first
is counting,
which can be exact. Example: How many coins do I have in my wallet?

The second type of quantitation is measurement. Example: How much do I weigh?
This
may sound like a strange thing to say, but ALL measurements have some
uncertainty. You
measure the same thing 5 times, and you ALWAYS will get 5 different answers if
you have
a sufficiently high level of readout resolution. Measurements are NEVER exact.

There are 2 sources of measurement error. The first is random error. The second
is
method error.

Sometimes the thing that you want to measure is elusive; so you use whatever
proxy
variable seems most appropriate. Example: In a telephone survey about which
shampoo
people would use if price were not a consideration, some respondents hang up,
without
answering your question. It is fairly reasonable to assume that the people who
do NOT
answer your question have the same shampoo preference as the people who do
answer
your question, especially if the response rate is high. The data that you are
able to collect
is a proxy variable for the data that you would like to be able to collect. But
this
substitution results in method uncertainty.

In principle, many scientists feel that one should always report an estimate of
the
measurement's uncertainty along with the measurement itself. In the real world,
this idea
is honored more in the breach than in the observance.

This practice has led many lay people to believe that any number coming out of a
machine
is the word of God. It has also led to the horrible expression: 'exact' science.
In reality, the
only exact science may be mathematics, which is mostly independent of
measurements.
(Some mathematicians feel that very good measurements over very large distances
have
the potential to disprove the Parallel Postulate in Euclidean Geometry.)

Essentially, self-rating scales are no less scientific than other kinds of
measurements.
Some types measurement have larger uncertainties than other types. When the
random
uncertainty of an individual measurement is too large, you can do some Signal
Averaging.
Measure the same thing many times, and take the average value of all of your
measurements.

In the case of measuring well-being as a function of walking, if you are not
confident of
the results in AB testing, then do ABAB testing, or ABABAB testing, etc. until
your
confidence in the results is greater than 5 on a 10-point scale.

Regards, Larry



--- In SelfExperimenters@yahoogroups.com, gamma sync <gammasync@...> wrote:
>
> How do you measure well being?
>
>
>
>
________________________________________________________________________________\
____
> Don't get soaked. Take a quick peak at the forecast
> with the Yahoo! Search weather shortcut.
> http://tools.search.yahoo.com/shortcuts/#loc_weather
>