Re: At the Population Level


"Mentalguy2k8" <mentalguy2k8@xxxxxxxxxxxxxx> wrote in message
news:mN74l.32197$er1.30363@xxxxxxxxxxxxxxxx

"Norman Wells" <no-one@xxxxxxxxxxx> wrote in message
news:gG74l.54521$Ci.8166@xxxxxxxxxxxxxxxx
judith wrote:
I frequent that rather odd place uk.rec.cycling - one of the phrases
which is used time and time again is above - as in:

"No-one has ever proved that cycle helmets make cycling any safer at
the population level"

"population-level studies that show little or no positive
effects of large secular changes in helmet wearing.

despite asking what people think the phrase means no one has said.

Guesses - suggestions?

Can the "population" be any size you choose?

In statistics, the term 'population' means the entire aggregation of
items from which samples can be drawn. So, for example, if all of the 60
million people living in the UK are possible cyclists then the relevant
'population' for any statistical survey is 60 million.

However, no survey can possibly cover that number of people so smaller
samples are taken, and the results from that sample are multiplied up to
give a conjectured result for the whole relevant 'population'. Depending
on the size of the sample taken, the result for the entire population can
be stated statistically within what are called confidence limits. So,
for example, if 100,000 people are surveyed as to whether they are
cyclists, and 20,000 say yes, it might be stated that the survey shows
that 11-13 million people in the UK are cyclists at the 95% confidence
level.

Whether those posting to uk.rec.cycling have any understanding of this at
all is an entirely different question.


You're talking about a group of people who can't tell a red traffic light
from a green one, and a road from a pavement. At the population level, of
course.

And the person doing the talking (Norman) doesn't know anything about
anything at all as he frequently demonstrates to this group (though rather
less so of late). He clearly knows nothing of statistics because the whole
point of using a sample to predict something about the population as a whole
is that the sample can be deliberately biased in a number of ways so that
the prediction is in fact what you want to prove rather than any meaningful
analysis. The confidence level is thus nearer zero than anything in the 90
percent range.

But of course the OP didn't ask what 'population' meant but what 'population
level' meant.

Now watch Norman demonstrate that he doesn't understand the concept of a
killfile either as he directs a probably wrong answer at me in spite of the
fact that I won't see it (unless someone who hasn't killfiled him is stupid
enough to respond).


.

Relevant Pages

  • Re: At the Population Level
    ... "No-one has ever proved that cycle helmets make cycling any safer at ... So, for example, if all of the 60 million people living in the UK are possible cyclists then the relevant 'population' for any statistical survey is 60 million. ... At the population level, of course. ... He clearly knows nothing of statistics because the whole point of using a sample to predict something about the population as a whole is that the sample can be deliberately biased in a number of ways so that the prediction is in fact what you want to prove rather than any meaningful analysis. ...
    (uk.legal)
  • Re: At the Population Level
    ... "No-one has ever proved that cycle helmets make cycling any safer at ... In statistics, the term 'population' means the entire aggregation of items from which samples can be drawn. ... So, for example, if all of the 60 million people living in the UK are possible cyclists then the relevant 'population' for any statistical survey is 60 million. ...
    (uk.legal)
  • Re: At the Population Level
    ... "No-one has ever proved that cycle helmets make cycling any safer at ... So, for example, if all of the 60 million people living in the UK are possible cyclists then the relevant 'population' for any statistical survey is 60 million. ... At the population level, of course. ...
    (uk.legal)
  • Re: [fw-wiz] Re: Flawed Surveys [was: VPN endpoints]
    ... >>So long as they're flawed approximately the same way from survey to ... >>We often don't need absolute metrics, ... > I am sorry, Paul - if you believe the statement you made above, you ... > risk requires real metrics and statistics. ...
    (Firewall-Wizards)
  • Re: Take my survey?
    ... She goes to a very well known school as well. ... she is doing her survey. ... there can be many that do read crime blogs and many that don't. ... http://en.wikipedia.org/wiki/Sampling_ (statistics) ...
    (alt.true-crime)