First, you said that the claim about the 'likes' was correct because the sample size was several thousand. That is wrong. If the sample is not representative, then it won't ge you the right result no matter how large a sample.
It is also incorrect that a larger sample will exponentially get closer to the correct answer as the number increases. Take my example. You want to know the world population who speaks Japanese. I'm not sure what the percentage is, but let's say it's 5%. Now, you take your sample in Tokyo. Of the first 1,000 people you choose, 99.5% speak Japanese. So you take another 1000 - and still have 99.5% speaking Japanese. Now you ask a million Tokyo residents and 99.5% of them speak Japanese. No matter how large your sample, you're not going to get close to the correct number - because your sample is not representative. It has absolutely nothing to do with sample size - and certainly this argument does nothing to establish that the self-selected sample you cited has any relevance to the population as a whole.
You're not disproving my point. What you're saying is that someone could take a horrible sample and get data that doesn't represent a population. In your example the survey taker is trying to determine the worldwide population of Japanese speakers while only sampling in a single country. There is an obvious error in the sampling. In the case of this video the sample is of people who came to watch the video. The population is also people who came to watch the video. There is no obvious error in the sampling. YouTube didn't only take reviews from women, men, 18 year olds, disabled veterans, or any other particular group. The sample is likely fair and unbiased.
You're not disproving my point. What you're saying is that someone could take a horrible sample and get data that doesn't represent a population. In your example the survey taker is trying to determine the worldwide population of Japanese speakers while only sampling in a single country. There is an obvious error in the sampling. In the case of this video the sample is of people who came to watch the video. The population is also people who came to watch the video. There is no obvious error in the sampling. YouTube didn't only take reviews from women, men, 18 year olds, disabled veterans, or any other particular group. The sample is likely fair and unbiased.
Of course there's an obvious error. Several, actually:
1. People who come to see a particular video may have a particular mind set.
2. People who bother to respond may be (and probably are) different than the population as a whole.
What you have is a self-selecting sample - and self-selecting samples are almost NEVER representative of the population as a whole.
Once again, you really need to get at least a basic understanding of statistics before commenting.
Of course there's an obvious error. Several, actually:
1. People who come to see a particular video may have a particular mind set.
2. People who bother to respond may be (and probably are) different than the population as a whole.
What you have is a self-selecting sample - and self-selecting samples are almost NEVER representative of the population as a whole.
Once again, you really need to get at least a basic understanding of statistics before commenting.
Firstly, belittling me and my knowledge of statistics does not further your argument and in fact makes you less believable to anyone else reading your post. Intimidation tactics fill in the gap from missing knowledge. But don't worry about my opinion of you as I already know you're spouting off things that you know nothing about. You do it all the time. I, on the other hand, am currently taking my fourth statistics class so unless you have a degree in statistics please refrain from lecturing me.
Anyway, I'll rebut your self-selecting sample idea. It's true that in many cases a self-selecting sample is biased. When a surveyor calls a home a woman is statistically more likely to answer the phone. Bias. When a surveyor is in the mall asking for opinions an extrovert is more likely to respond. Bias. When a survey is long a person with a less demanding job is more likely to respond. Bias. In the case of this video respondents are anonymous, have no reason to be of a particular demographic (at least no more particular than the population of video watchers), and the “survey" takes less than a second to complete. The only way that bias could cause problems in this case is if I tried to apply my knowledge of the sample to a different population, such as the entire US.
Firstly, belittling me and my knowledge of statistics does not further your argument and in fact makes you less believable to anyone else reading your post. Intimidation tactics fill in the gap from missing knowledge. But don't worry about my opinion of you as I already know you're spouting off things that you know nothing about. You do it all the time. I, on the other hand, am currently taking my fourth statistics class so unless you have a degree in statistics please refrain from lecturing me.
Anyway, I'll rebut your self-selecting sample idea. It's true that in many cases a self-selecting sample is biased. When a surveyor calls a home a woman is statistically more likely to answer the phone. Bias. When a surveyor is in the mall asking for opinions an extrovert is more likely to respond. Bias. When a survey is long a person with a less demanding job is more likely to respond. Bias. In the case of this video respondents are anonymous, have no reason to be of a particular demographic (at least no more particular than the population of video watchers), and the “survey" takes less than a second to complete. The only way that bias could cause problems in this case is if I tried to apply my knowledge of the sample to a different population, such as the entire US.
No one intimidated you or belittled you. I simply pointed out that you were wrong - and that you continued to be wrong when you kept repeating the same nonsense.
You're STILL wrong. A self-selected sample (the people who choose to click on 'like') is not representative of the population as a whole. Period.
And I noticed that you still haven't responded to the evidence of inherent bias in your method. Look at videos at random on YouTube. Likes outnumber dislikes by at least 10 to one in most videos. That's the sign of a self-selection bias.
No one intimidated you or belittled you. I simply pointed out that you were wrong - and that you continued to be wrong when you kept repeating the same nonsense.
You're STILL wrong. A self-selected sample (the people who choose to click on 'like') is not representative of the population as a whole. Period.
And I noticed that you still haven't responded to the evidence of inherent bias in your method. Look at videos at random on YouTube. Likes outnumber dislikes by at least 10 to one in most videos. That's the sign of a self-selection bias.
You're correct. Most people that dislike a video probably just close it than click the dislike button. Now it'd be a different story if one had to click like/dislike.
You're correct. Most people that dislike a video probably just close it than click the dislike button. Now it'd be a different story if one had to click like/dislike.
That's one source of bias. Another source of bias is that people who are Samsung fans are probably more likely to go to the video than people who don't care about Samsung.
In addition, considering that it was presented on a couple of sites as an Apple bashing video, all the iHaters would be more likely to go to that video than the average citizen.
As a general rule, self-selected examples are almost never statistically valid.
No one intimidated you or belittled you. I simply pointed out that you were wrong - and that you continued to be wrong when you kept repeating the same nonsense.
You're STILL wrong. A self-selected sample (the people who choose to click on 'like') is not representative of the population as a whole. Period.
And I noticed that you still haven't responded to the evidence of inherent bias in your method. Look at videos at random on YouTube. Likes outnumber dislikes by at least 10 to one in most videos. That's the sign of a self-selection bias.
That videos often have more likes than dislikes doesn't prove I'm wrong. It just means that the population of viewers on a given a video are likely to like that video because they chose to watch it in the first place. Once again I'm not saying that the sample is representative of any population other than the one that watched the video.
That videos often have more likes than dislikes doesn't prove I'm wrong. It just means that the population of viewers on a given a video are likely to like that video because they chose to watch it in the first place. Once again I'm not saying that the sample is representative of any population other than the one that watched the video.
The sample isn't even representative of the people who watch the video. As long as there's a propensity for people to be more likely to click 'like' than 'dislike' (or vice versa, for that matter), the results are not representative. And considering that the people who really hate it are likely to shut it off rather than clicking dislike, it's even less representative.
Once again, please stop embarrassing yourself. Pick up a book about statistics and see why self-selected samples are almost never valid. And it's certainly not meaningful among the population as a whole - which was what you were implying initially.
Oh, and in case anyone cares, here's the full superbowl commercial. Better than the teaser, but still not very impressive.
IMO it wasn't better than the teaser. This one is just rambling.
I thought the teaser was funnier overall, perhaps because of its brevity and focus, but this one had plenty of funny moments. Paul Rudd suggesting Seth Rogan is there to see a guy named Sam Sung, and LeBron James saying how he can do a cameo on a tablet while doing a cameo on a tablet were on the mark. And unlike the teaser they incorporated the product, its features and name often without making it feel forced. Finally, I think they hit a very wide demographic with the choice of celebrities without degrading the entertainment factor of the spot. I think they did a brilliant job.
IMO it wasn't better than the teaser. This one is just rambling.
True, but at least they showed some of the things you could do with their product. Of course, they skimmed over the only one that's unique to their product so fast that most people won't catch it, but at least the product was visible.
True, but at least they showed some of the things you could do with their product. Of course, they skimmed over the only one that's unique to their product so fast that most people won't catch it, but at least the product was visible.
Yeah, the part with S-Beam was so brief that if you weren't aware of it before you aren't going to wonder about it after it watching that ad. It should have had another second of two to at least make people think "What is that?" or "I really like that?"
The sample isn't even representative of the people who watch the video. As long as there's a propensity for people to be more likely to click 'like' than 'dislike' (or vice versa, for that matter), the results are not representative. And considering that the people who really hate it are likely to shut it off rather than clicking dislike, it's even less representative.
Once again, please stop embarrassing yourself. Pick up a book about statistics and see why self-selected samples are almost never valid. And it's certainly not meaningful among the population as a whole - which was what you were implying initially.
Oh, and in case anyone cares, here's the full superbowl commercial. Better than the teaser, but still not very impressive.
I don't always agree with jragosta but his comments are never a joke. A tenable rebuttal to his comment would be a better defense than a personal attack.
I don't always agree with jragosta but his comments are never a joke. A tenable rebuttal to his comment would be a better defense than a personal attack.
His comments are not a joke. They're a sad window into the mind of a person who I suspect is lonely and bitter. How can someone who is always "right" have a meaningful relationship with another human. My cousin is the same as jragosta. He's (my cousin) an intolerable person who people cannot wait to see exit a room because he tries to correct people constantly and argues endlessly even when clearly wrong.
And of all the personal attacks that you see on this forum, this one, which was really an observation, is the one you decide to call out?
His comments are not a joke. They're a sad window into the mind of a person who I suspect is lonely and bitter. How can someone who is always "right" have a meaningful relationship with another human. My cousin is the same as jragosta. He's (my cousin) an intolerable person who people cannot wait to see exit a room because he tries to correct people constantly and argues endlessly even when clearly wrong.
And of all the personal attacks that you see on this forum, this one, which was really an observation, is the one you decide to call out?
Funny, but I see you still haven't managed anything even resembling a logical argument in support of your position. I calmly explained why you were wrong (even repeating my explanation several times when you refused to listen) and never called you a name. Nor were any of my comments even remotely personal attacks.
Your post, however, reads very much like someone who lost the argument and has no hope of a rational rebuttal and therefore decides to resort to personal attacks.
I would think that if you yourself recognize that what you're saying is a personal attack, you wouldn't say it. At least not on the forum proper.
There's a very small but active number of forum members here that must not consider what they write to be personal attacks, tho to an average reader it would be obvious many of their posts are just that.
Comments
You're not disproving my point. What you're saying is that someone could take a horrible sample and get data that doesn't represent a population. In your example the survey taker is trying to determine the worldwide population of Japanese speakers while only sampling in a single country. There is an obvious error in the sampling. In the case of this video the sample is of people who came to watch the video. The population is also people who came to watch the video. There is no obvious error in the sampling. YouTube didn't only take reviews from women, men, 18 year olds, disabled veterans, or any other particular group. The sample is likely fair and unbiased.
Of course there's an obvious error. Several, actually:
1. People who come to see a particular video may have a particular mind set.
2. People who bother to respond may be (and probably are) different than the population as a whole.
What you have is a self-selecting sample - and self-selecting samples are almost NEVER representative of the population as a whole.
Once again, you really need to get at least a basic understanding of statistics before commenting.
Firstly, belittling me and my knowledge of statistics does not further your argument and in fact makes you less believable to anyone else reading your post. Intimidation tactics fill in the gap from missing knowledge. But don't worry about my opinion of you as I already know you're spouting off things that you know nothing about. You do it all the time. I, on the other hand, am currently taking my fourth statistics class so unless you have a degree in statistics please refrain from lecturing me.
Anyway, I'll rebut your self-selecting sample idea. It's true that in many cases a self-selecting sample is biased. When a surveyor calls a home a woman is statistically more likely to answer the phone. Bias. When a surveyor is in the mall asking for opinions an extrovert is more likely to respond. Bias. When a survey is long a person with a less demanding job is more likely to respond. Bias. In the case of this video respondents are anonymous, have no reason to be of a particular demographic (at least no more particular than the population of video watchers), and the “survey" takes less than a second to complete. The only way that bias could cause problems in this case is if I tried to apply my knowledge of the sample to a different population, such as the entire US.
No one intimidated you or belittled you. I simply pointed out that you were wrong - and that you continued to be wrong when you kept repeating the same nonsense.
You're STILL wrong. A self-selected sample (the people who choose to click on 'like') is not representative of the population as a whole. Period.
And I noticed that you still haven't responded to the evidence of inherent bias in your method. Look at videos at random on YouTube. Likes outnumber dislikes by at least 10 to one in most videos. That's the sign of a self-selection bias.
You're correct. Most people that dislike a video probably just close it than click the dislike button. Now it'd be a different story if one had to click like/dislike.
That's one source of bias. Another source of bias is that people who are Samsung fans are probably more likely to go to the video than people who don't care about Samsung.
In addition, considering that it was presented on a couple of sites as an Apple bashing video, all the iHaters would be more likely to go to that video than the average citizen.
As a general rule, self-selected examples are almost never statistically valid.
That videos often have more likes than dislikes doesn't prove I'm wrong. It just means that the population of viewers on a given a video are likely to like that video because they chose to watch it in the first place. Once again I'm not saying that the sample is representative of any population other than the one that watched the video.
Shamesung is at it again.
The sample isn't even representative of the people who watch the video. As long as there's a propensity for people to be more likely to click 'like' than 'dislike' (or vice versa, for that matter), the results are not representative. And considering that the people who really hate it are likely to shut it off rather than clicking dislike, it's even less representative.
Once again, please stop embarrassing yourself. Pick up a book about statistics and see why self-selected samples are almost never valid. And it's certainly not meaningful among the population as a whole - which was what you were implying initially.
Oh, and in case anyone cares, here's the full superbowl commercial. Better than the teaser, but still not very impressive.
IMO it wasn't better than the teaser. This one is just rambling.
EDIT: The ratio of likes to dislikes compared to the teaser ad ratings proves it.
/s
I thought the teaser was funnier overall, perhaps because of its brevity and focus, but this one had plenty of funny moments. Paul Rudd suggesting Seth Rogan is there to see a guy named Sam Sung, and LeBron James saying how he can do a cameo on a tablet while doing a cameo on a tablet were on the mark. And unlike the teaser they incorporated the product, its features and name often without making it feel forced. Finally, I think they hit a very wide demographic with the choice of celebrities without degrading the entertainment factor of the spot. I think they did a brilliant job.
True, but at least they showed some of the things you could do with their product. Of course, they skimmed over the only one that's unique to their product so fast that most people won't catch it, but at least the product was visible.
Yeah, the part with S-Beam was so brief that if you weren't aware of it before you aren't going to wonder about it after it watching that ad. It should have had another second of two to at least make people think "What is that?" or "I really like that?"
You're a joke.
I don't always agree with jragosta but his comments are never a joke. A tenable rebuttal to his comment would be a better defense than a personal attack.
Quote:
Originally Posted by SolipsismX
I don't always agree with jragosta but his comments are never a joke. A tenable rebuttal to his comment would be a better defense than a personal attack.
His comments are not a joke. They're a sad window into the mind of a person who I suspect is lonely and bitter. How can someone who is always "right" have a meaningful relationship with another human. My cousin is the same as jragosta. He's (my cousin) an intolerable person who people cannot wait to see exit a room because he tries to correct people constantly and argues endlessly even when clearly wrong.
And of all the personal attacks that you see on this forum, this one, which was really an observation, is the one you decide to call out?
1) Absolutely! You said he was a joke, you did not say his comments are a joke. Your anecdote solidifies your intent for it being personal.
2) You made no valid rebuttal. You only made the personal attack thus is stood out.
Funny, but I see you still haven't managed anything even resembling a logical argument in support of your position. I calmly explained why you were wrong (even repeating my explanation several times when you refused to listen) and never called you a name. Nor were any of my comments even remotely personal attacks.
Your post, however, reads very much like someone who lost the argument and has no hope of a rational rebuttal and therefore decides to resort to personal attacks.
I would think that if you yourself recognize that what you're saying is a personal attack, you wouldn't say it. At least not on the forum proper.
Quote:
Originally Posted by Tallest Skil
I would think that if you yourself recognize that what you're saying is a personal attack, you wouldn't say it. At least not on the forum proper.
There's a very small but active number of forum members here that must not consider what they write to be personal attacks, tho to an average reader it would be obvious many of their posts are just that.
Pompous is the new black.