Should Posters be Allowed to be Anonymous?

The issue of allowing posters to be anonymous has raged at this forum for quite some time, unlike on other forums that I’ve visited. It’s time to conduct a poll so we can ”give it a rest” and "get it over with."

While the forum is not a democracy because it is privately owned, a poll should settle the issue (at least for a while). If a majority of participants vote against anonymity, then I will step out of the closet (No Levicoff, not that closet ).

Vote your conscience.

 

Mannnn, it's "two to four" in the poll so far: against me.

If I lose, would that involve posting both a first and a last name, or just a first name? The criteria isn't established.

What am I thinking? This poll isn't over yet!

 

Poll's Statistics Have Changed!

• Yesterday, only 18% of the respondents said that posters should be allowed to be anonymous.
  
• In only 24 hours, that figure has more than doubled to 40% who believe that posters should be allowed to remain anonymous.

Isn’t it amazing how the interpretation of polls can be manipulated by the writer?

 

Well, 15 people have voted so far. According to the "About Us" page, there are 1,400 registered members (plus four times as many casual visitors to the board). So you're going to have to wait for a long time before you get a representative sample!

 

Not necessarily. That depends on how confident you want to be that the sampled proportion represents the population proportion, and how wide your confidence interval can be. I'll give you those numbers, if you'd like.

Of course, statistical inferences regarding a sample's ability to predict a population is based on random sampling, where each participant chosen has an equal chance of being chosen as any other member of the population. The kind of poll where the participants decide whether or not to join the sample do not meet this standard. That's why call-in and webpage-based polls are not considered "scientific." Still, I can predict the sample size necessary to acheive a given level of confidence and a given confidence interval. Not hard at all, even for a binomial.


Rich Douglas

 

For example, right now the numbers are 10 for anonymous posting and 13 against. That's 43% for and 57 % against. You can be 95 percent sure that the population is plus or minus 25.54% of 43% for. (17.46% to 68.54% in favor of anonymous posting.)

Of course, this would assume random sampling, which we do not have. Now, let's look at how large the sample would have to be to narrow the confidence interval from 25.54% to 5%: 377

If we lower the confidence level to .9 (90%) we'd need: 266

If we raise the interval to plus or minus 10% and keep the confidence interval at 90% (we're 90% sure the population parameter falls within the confidence interval): 41.

So, if your sample size is 41 (and the proportions stay similar), you can be pretty sure the population as a whole is pretty near there (plus or minus 10 percentage points, or .1). Unfortunately, because the two sides are so near, you really couldn't substantiate that there is a difference in the entire population: your sample wouldn't provide enough data and your confidence interval (margin of error) would be larger than the difference between the two sides.

I predict a result that is too close to call a statistically significant difference, unless the data moves wildly one way or another.

(Except that the data is skewed by the fact that only those motivated to participate will be included in the sample. You'll have to determine for yourself whether that has any effect on the result. Polls like these tend to get the extremes, but miss the "Silent Majority," to use a Nixon-era term.)

Rich Douglas

 

accuracy or inaccuracy?

• Yesterday, 40% of the respondents said that posters should be allowed to be anonymous.
  
• In only 24 hours, that figure has increased by two percent to 42% who believe that posters should be allowed to remain anonymous.

Wow, another increase! Two days in a row.

It's all in how the news is reported e.g. the respondents (thus far) have clearly voted against anonymity, but the writer (me again!!!!) has not pointed this fact out (until now). Instead, the writer has chosen to report the percentile increase in voters who want to remain anonymous, despite the fact that it was marginal. Words are everything! And the vote isn't over with yet.

I don’t know. This is a voluntary forum… a voluntary thread… Only those with an opinion will vote. The poll is unique to DegreeInfo.com and, subsequently, there may be some validity to the end result. And maybe not.

The last presidential poll comprised far less than 1% of the population. Conversely, there was less than a 1% margin of error (3% is the norm). How do you suppose this margin of accuracy was obtained when such a small sample was tested? (I know the answer, as told by one of my professors. The answer is absolutely facinating).

 

Re: accuracy or inaccuracy?

• 
  
  There are three variables to polling. You can adjust the number in the sample, the confidence level, or the margin of error. An increase in one will decrease a second one if the third stays constant. For example, if you keep the confidence level the same, you can decrease the margin of error by increasing the number sampled.
  
  So, what is your fascinating answer?
  
  Rich Douglas

 

The Trilateral Commission fixed the result up front.

 

Hmm, my guess would have been the Illuminati.




Tom Nixon

 

Re: accuracy or inaccuracy?

• 
  
  By the way, 1% of Americans comes out to 2.6 million or so. Even taken from the lower number that makes up registered voters, that's several hundred thousand. No pollster takes those kinds of numbers. The cost would be too high, and the benefits beyond a much smaller sample would be tiny.
  
  If you reduce the margin of error and keep the sample size the same, you have effectly decreased the confidence level. For example, you could poll 100 people (randomly selected, of course), keep your margin of error at 1%, and get a result. But your confidence level that the sample result accurately reflects the mean (plus or minus 1%) will be pretty low. In statistics, a confidence level of 90, 95 or 99 percent is commonly used.
  
  It's all in how you want to present it (and what information you chose to omit). You've not indicated that, nor revealed your "fascinating answer." Waiting....
  
  Rich Douglas

 

A report just came across my desk where a citizen called my agency and she told the responding officer that "electromagnetic waves" are being deliberately sent into her home and the waves are giving her a headache. She wants us to find out what kind of waves they are and she wants us to make it stop. There are people out there who really believe this stuff.

 

Tell her to wrap her head in aluminum foil, that's the most popular remedy among the mentally challenged around here.


Bruce (who once went to a call where a guy was lining his apartment with aluminum foil...50+ boxes worth)

 

Almost as bizarre as the people who really believe a degree mill credential is a valid legitimate degree, huh?

 

Re: Re: accuracy or inaccuracy?

Gee, it's a shame DL doesn't work--a guy could actually learn something from this threaded discussion. I have to assume that if Rich were telling us this stuff in person we would somehow develop a meaningful personal relationship with him and, thereby, learn it better.

Tracy<><

 

Re: Re: Re: accuracy or inaccuracy?

Yeah, like I'm going to have a deep, personal relationship with any of you guys!

The professors who talk about the value of face-to-face learning are the same ones who teach with their backs to their students while writing on the board. They're the ones who'd shove their students into a black hole in a New York minute if they could. They're the ones who can't figure out how to change the voice mail greeting on their phones, much less function in a computer-savvy environment. They're the one's trying to save not their jobs, but their comfort zones.

Screw 'em. DL lives!

Rich Douglas

 

Re: Re: Re: Re: accuracy or inaccuracy?

I'd swear that was quite an emotional transmission, were it not for the fact this medium precludes the possibility.

.

 

Thus far (4/12/02):

• 70% have voted to ban anonymous posters!?!?!
  
• 29% have voted to keep anonymous posters!

Gee. I wonder if anyone here voted more than once... Lots more than once!

 

There's no way to vote multiple times, well... unless you're Ken/Lewchuck/Scott Henley/etc.

So, looks like you're going to have to step out of the closet...

 

Yup, it does, doesn't it

We're all waiting....