My name is Peter McPherson. I am a fourth year Psychology student at
RMIT University in Melbourne, Australia. I am currently conducting a
research study investigating community attitudes toward online therapy,
and I am looking for individuals that may be interested in being
involved. Participants will be asked to complete a brief online
questionnaire, taking approximately 15 minutes. By completing the
questionnaire, you can also enter the draw to win a $200 Amazon gift
voucher. If you are interested, please follow the link below for more
information and an opportunity to participate.

http://www.dlsweb.rmit.edu.au/appsci...survey/online/ .

Many thanks,

Peter McPherson.

Online Therapy?
Sure, why not. I supply my own couch and the computer supplies the tissue
box with hankies!
Seriously, this is the wrong newsgroup for your study.
How about finding losers at alt.tv-desperate-housewives!?




wrote in message
ps.com...
My name is Peter McPherson. I am a fourth year Psychology student at
RMIT University in Melbourne, Australia. I am currently conducting a
research study investigating community attitudes toward online therapy,
and I am looking for individuals that may be interested in being
involved. Participants will be asked to complete a brief online
questionnaire, taking approximately 15 minutes. By completing the
questionnaire, you can also enter the draw to win a $200 Amazon gift
voucher. If you are interested, please follow the link below for more
information and an opportunity to participate.

http://www.dlsweb.rmit.edu.au/appsci...survey/online/ .

Many thanks,

Peter McPherson.

On Wed, 02 Aug 2006 16:28:20 GMT, "Herb Ludwig" wrote:

Seriously, this is the wrong newsgroup for your study.

Which newsgroup? :-) Maybe one of the other four groups is potentially
correct.

-m-
--

Herb Ludwig wrote:
Online Therapy?
Sure, why not. I supply my own couch and the computer supplies the tissue
box with hankies!
Seriously, this is the wrong newsgroup for your study.
How about finding losers at alt.tv-desperate-housewives!?

A survey made through ANY newsgroup can't yield valid
results. A valid survey must get responses from a
representtaive crosssection, not from a self-selected group.
I don't waste my time on these online surveys.

In article , Marvin
wrote:

Herb Ludwig wrote:
Online Therapy?
Sure, why not. I supply my own couch and the computer supplies the tissue
box with hankies!
Seriously, this is the wrong newsgroup for your study.
How about finding losers at alt.tv-desperate-housewives!?

A survey made through ANY newsgroup can't yield valid
results. A valid survey must get responses from a
representtaive crosssection, not from a self-selected group.
I don't waste my time on these online surveys.

All surveys are useless rubbish because the sample size is always
ridiculously small - and that includes the stupid "ratings" for TV
shows. The one exception is the Census that is SUPPOSED to include
everyone, but never really does. Surveys where the people conduting it
pick-and-choose particiapants are even more useless than usual, while a
"random" survey like this is slightly more uselful than that ... but
only slightly, so it's still a complete waste of time and money.

Anybody wrote:
In article , Marvin
wrote:

Herb Ludwig wrote:
Online Therapy?
Sure, why not. I supply my own couch and the computer supplies the tissue
box with hankies!
Seriously, this is the wrong newsgroup for your study.
How about finding losers at alt.tv-desperate-housewives!?
A survey made through ANY newsgroup can't yield valid
results. A valid survey must get responses from a
representtaive crosssection, not from a self-selected group.
I don't waste my time on these online surveys.

All surveys are useless rubbish because the sample size is always
ridiculously small - and that includes the stupid "ratings" for TV
shows. The one exception is the Census that is SUPPOSED to include
everyone, but never really does. Surveys where the people conduting it
pick-and-choose particiapants are even more useless than usual, while a
"random" survey like this is slightly more uselful than that ... but
only slightly, so it's still a complete waste of time and money.

But maybe the 'survey' is a survey on responses to taking a survey!!!
Oh, how devilishly clever in its in-trick-a-sy!

Anybody wrote:
snip

All surveys are useless rubbish because the sample size is always
ridiculously small - and that includes the stupid "ratings" for TV
shows. The one exception is the Census that is SUPPOSED to include
everyone, but never really does. Surveys where the people conduting it
pick-and-choose particiapants are even more useless than usual, while a
"random" survey like this is slightly more uselful than that ... but
only slightly, so it's still a complete waste of time and money.

Certain groups self-select out, such as illegal immigrants.
Others are hard for the census takers to locate and
question, such as homeless people with mental problems.
There are statistical procedures to adjust for these sources
of error, but any time they are used there is a political
uproar.

In article , Marvin
wrote:

Anybody wrote:
snip

All surveys are useless rubbish because the sample size is always
ridiculously small - and that includes the stupid "ratings" for TV
shows. The one exception is the Census that is SUPPOSED to include
everyone, but never really does. Surveys where the people conduting it
pick-and-choose particiapants are even more useless than usual, while a
"random" survey like this is slightly more uselful than that ... but
only slightly, so it's still a complete waste of time and money.

Certain groups self-select out, such as illegal immigrants.
Others are hard for the census takers to locate and
question, such as homeless people with mental problems.
There are statistical procedures to adjust for these sources
of error, but any time they are used there is a political
uproar.

That's the entire problem with surveys - "statistical procedures" are
no better than guessing and you can "massage" the real data to show
almost anything you want via statistics.

The ones that really really annoy me are the silly medical "studies"
that get reported in the media. They make big announcements that "doing
/ not doing X will mean you're 30% less likely to have cancer" (or
something), but if you actually listen to the details or read the small
print you find that the ridiculous "study" took place over a few weeks
and was based around only 100 people - 100 out of a world-wide
population is ridiculously stupid and won't give you any reliable
results at all.

Anybody wrote:

All surveys are useless rubbish because the sample size is always
ridiculously small - and that includes the stupid "ratings" for TV
shows. The one exception is the Census that is SUPPOSED to include
everyone, but never really does.

By very definition, a statistical survey is not a complete collection.
A complete collection cannot be a sample. Everything would be 'significant'.

Statistical measures are very helpful to those who understand them.

"Anybody" wrote

That's the entire problem with surveys - "statistical procedures" are
no better than guessing and you can "massage" the real data to show
almost anything you want via statistics.

Anyone can lie. True. However, there are certain statistical procedures
which are hugely useful.

John wrote:
Anybody wrote:


All surveys are useless rubbish because the sample size is always
ridiculously small - and that includes the stupid "ratings" for TV
shows. The one exception is the Census that is SUPPOSED to include
everyone, but never really does.


By very definition, a statistical survey is not a complete collection.
A complete collection cannot be a sample. Everything would be 'significant'.

Statistical measures are very helpful to those who understand them.

I fully agree. As a scientist, I've seen many successful
applications of statistics - when the appropriate
statistical procedure is used for a specific case.

Anybody wrote:
snip
People see a % character and suddenly believe that what it says is
completely true and can't possibly be wrong ... the reality is how it
can EASILY be wrong or "mis-interpreted (either by accident or on
purpose) and usually is.

Several years ago, the U.S. Weather Bureau included
estimated % chance of rain in forecasts. I thought it waas
a big advance over predicting yes or no on rain, or "a
chance". But most people got confused, and complained
instead of trying to learn, so this form of forecast was
dropped.

More recently, some pols in NY State prevailed on the state
health department to generate a table of annual incidence of
breast cancer by zip code. It was useless data for several
reasons. There were few cases in most zip codes. There was
no attention to whether the women who had breast cancer
lived in the zip code for years or recently moved in. At
the 95% confidence level, 5% of the zip codes would be
tagged as unsafe to live in, just by chance. Of course, the
pols issued press releases, and the public got alarmed, over
nothing. The same story for cancer near power lines.

Figures don't lie, maybe, but liars misuse statistics.

Good posts! The public-in-general cannot interpret statistics. Worse, humans
make poor decisions in general. They haven't a method. Most can't even tell
the liklihood of a list of coin-flips; shown two columns of flips, one real
and another made-up by a person, they are inclined to think the human-made
series is real.

--
another Flawed Human

In article , Marvin
wrote:

Anybody wrote:
snip
People see a % character and suddenly believe that what it says is
completely true and can't possibly be wrong ... the reality is how it
can EASILY be wrong or "mis-interpreted (either by accident or on
purpose) and usually is.

Several years ago, the U.S. Weather Bureau included
estimated % chance of rain in forecasts. I thought it waas
a big advance over predicting yes or no on rain, or "a
chance". But most people got confused, and complained
instead of trying to learn, so this form of forecast was
dropped.

More recently, some pols in NY State prevailed on the state
health department to generate a table of annual incidence of
breast cancer by zip code. It was useless data for several
reasons. There were few cases in most zip codes. There was
no attention to whether the women who had breast cancer
lived in the zip code for years or recently moved in. At
the 95% confidence level, 5% of the zip codes would be
tagged as unsafe to live in, just by chance. Of course, the
pols issued press releases, and the public got alarmed, over
nothing. The same story for cancer near power lines.

Figures don't lie, maybe, but liars misuse statistics.

Actual figures can't lie, but the problem is that percentages can and
do mislead, especially when you don't know what sample size their based
on - that is a large part of the problem. Reporting results as just
percentages is extremely misleading and is one of techniques employed
by those who want to confuse or lie to people. (Often they try to
disguise these percentages as pie charts.)

A "study" will "report" that 20% of people will get cancer, but then
you look at the silly "study" you find that that "20%" is really 20
people out of the 100 they interviewed ... NOT 20% of the entire
population. The error rate when using such ridiculously small sample
sizes (and all from the same area) is insanity and reporting such
"results" as meaning anything at all should be illegal.

If instead the reported "results" actually said that "20 people out of
the 100 in the study got cancer", then a lot more people would realise
how stupid the "study" actually was and how useless the "results" are
.... and what a total waste of time and money that "study" really was.

They aren't reporting actual facts, they're simply reporting guesses.

The exact same thing happens with TV "ratings".

Such surveys also try to confuse you by saying the error rate is only
0.5% for example. People see that and think that's too small to worry
about. In reality 0.5% of a population of 30million people is 150,000
people - the entire population of a small town.