4 ms·

> Most people believe that they are above average, a statistical impossibility It's perfectly possible for the majority to be above the mean. If most people be

> Most people believe that they are above average, a statistical impossibility It's perfectly possible for the majority to be above the mean. If most people believed to be above the median, that would be a statistical impossibility. I think there is another reason for people to believe their results in some areas are better than those of others: because of their own world view. A teacher who focusses on controlling the classroom may see many colleagues be worse at that and decide they're not as good a teacher as him; his colleague who prefers to leave the studying behaviour of his students may think his controlling colleague is not as good a teacher as him because he doesn't leave students enough freedom. Whether this applies to beauty I don't know - I'd expect it to apply only to aspects that they chose to modify (i.e. someone with dyed hair will probably have dyed their hair in a colour they believe to be more attractive and then believe they're more attractive).

redwood 13y ago
Your stats are true when outliers massively move a mean.. but that wouldn't be the case in a distribution like this one... I don't think. I mean it's more appropriate to imagine a normal distribution here with the mean equal to the median because the measurement is totally subjective anyway.
- gizmo686 13y ago
  Theres a huge difference between something being statistically impossible, and something being statisticly impossible, assuming it is a normal distribution.
  - bthomas 13y ago
    Might just be tired but I don't get this...care to explain?
    - HelloMcFly 13y ago
      If you assume a normal distribution, it is statistically impossible for more than 50% of people to consider themselves above the mean on a given metric. However, if the distribution is severely negatively skewed (i.e., the "bulge" of the distribution is moved to the left and there are many outliers below the mean) then more than 50% could be above the mean because the outliers on the left end of the distribution drag the mean down significantly.
    - gizmo686 13y ago
      In a normal distribution, the mean and median are the same value, so only 50% of the data points can be above average. Without this assumption, you are not guaranteed to have only 50% of your data points above average. For example, what percent of the population do you think has an above average number of legs?
      - colomon 13y ago
        Last line is the Best Example Ever of median and mean not agreeing. Did you come up with it yourself?
mistercow 13y ago
Technically, "average" is not a precise term, and can also be used to refer to the median or the mode. In a situation like attractiveness, the median and mean are likely to be approximately equal, so the distinction isn't very important.
azakai 13y ago
And, the article went right on to talk about the median, in fact :) > Most people believe that they are above average, a statistical impossibility. The above average effects, as they are called, are common. For example, 93 percent of drivers rate themselves as better than the median driver.
lotharbot 13y ago
You're exactly right. When I try to make myself look attractive, I aspire to my own standards (and my wife's). When others try to make themselves look attractive, they aspire to their different standards. We modify a lot of things about our own looks, which may not be as blatant as hair dye: clothing, hairstyle, makeup, facial hair, eyewear, and of course our own facial expressions. This means others have a lot of opportunity to boost themselves according to their standards, and penalize themselves according to mine. Likewise, I'm more likely to have a "good life" than others by my own definition. I'm more likely to be "educated" than others in the fields which I consider education important. I'm more likely to find the "right balance" between work and family, or between present and future, or between hope and realism, because I'm aspiring to exactly the point which I consider "right". Generally, people tend to do better according to their own standards than others' standards, and are thus more likely to be "above average".

redwood 13y ago

Your stats are true when outliers massively move a mean.. but that wouldn't be the case in a distribution like this one... I don't think. I mean it's more appropriate to imagine a normal distribution here with the mean equal to the median because the measurement is totally subjective anyway.

gizmo686 13y ago
Theres a huge difference between something being statistically impossible, and something being statisticly impossible, assuming it is a normal distribution.
- bthomas 13y ago
  Might just be tired but I don't get this...care to explain?
  - HelloMcFly 13y ago
    If you assume a normal distribution, it is statistically impossible for more than 50% of people to consider themselves above the mean on a given metric. However, if the distribution is severely negatively skewed (i.e., the "bulge" of the distribution is moved to the left and there are many outliers below the mean) then more than 50% could be above the mean because the outliers on the left end of the distribution drag the mean down significantly.
  - gizmo686 13y ago
    In a normal distribution, the mean and median are the same value, so only 50% of the data points can be above average. Without this assumption, you are not guaranteed to have only 50% of your data points above average. For example, what percent of the population do you think has an above average number of legs?
    - colomon 13y ago
      Last line is the Best Example Ever of median and mean not agreeing. Did you come up with it yourself?

mistercow 13y ago

Technically, "average" is not a precise term, and can also be used to refer to the median or the mode. In a situation like attractiveness, the median and mean are likely to be approximately equal, so the distinction isn't very important.

azakai 13y ago

And, the article went right on to talk about the median, in fact :) > Most people believe that they are above average, a statistical impossibility. The above average effects, as they are called, are common. For example, 93 percent of drivers rate themselves as better than the median driver.

lotharbot 13y ago

You're exactly right. When I try to make myself look attractive, I aspire to my own standards (and my wife's). When others try to make themselves look attractive, they aspire to their different standards. We modify a lot of things about our own looks, which may not be as blatant as hair dye: clothing, hairstyle, makeup, facial hair, eyewear, and of course our own facial expressions. This means others have a lot of opportunity to boost themselves according to their standards, and penalize themselves according to mine. Likewise, I'm more likely to have a "good life" than others by my own definition. I'm more likely to be "educated" than others in the fields which I consider education important. I'm more likely to find the "right balance" between work and family, or between present and future, or between hope and realism, because I'm aspiring to exactly the point which I consider "right". Generally, people tend to do better according to their own standards than others' standards, and are thus more likely to be "above average".