How Old Are You (Poll)

[Infrasonic]

Originally Posted by Stazy

Doesn't look like a dolly parton distribution to me...

Post op...

 

[Don Carlos]

This is an interesting thread. I would not have figured that the board's population skewed as young as it appears to in the results display.

Granted, at 30, my definition of "young" is probably not everyone's definition.

Big thanks to the OP, btw, for putting 30 in such a generous slice of the categories. It's rare to find 30 at the tail end of a mid-20s category and not at the front end of a mid-30s category. Makes me feel good!

 

[Huntsman]

Originally Posted by Arrogant Bastard

This is an interesting thread. I would not have figured that the board's population skewed as young as it appears to in the results display.
Just what I thought -- I was expecting a peak around 21, and another in the mid-thirties. Admittedly, I am messing with the distribution by having my ranges of different sizes, but I find it challenging to put a 20 year-old in the same category as a 25 year old.

Originally Posted by Arrogant Bastard

Big thanks to the OP, btw, for putting 30 in such a generous slice of the categories. It's rare to find 30 at the tail end of a mid-20s category and not at the front end of a mid-30s category. Makes me feel good!
Hahah, that's probably more a function of how I started the distribution of ranges than anything else....or, of course, it is a Freudian slip that tells you something about my age... ~ H

 

[Egdon Heath]

455 in dog years. So maybe 65 or so. I know most people do the dog years thing only after knowing their true age, but since I was little my mother would only tell me how old I was in dog years so I have to figure it backwards, so, yeah 65, I guess.​

 

[Don Carlos]

Originally Posted by Huntsman

Just what I thought -- I was expecting a peak around 21, and another in the mid-thirties.
Admittedly, I am messing with the distribution by having my ranges of different sizes, but I find it challenging to put a 20 year-old in the same category as a 25 year old.
~ H

Yeah, I just noticed that, actually. Some of your categories/ranges are quite a bit smaller than others, which makes interpreting the distribution from the graph somewhat misleading. I hear you about how there's a big emotional/maturity difference among certain age ranges that is much more significant than the numerical age range would imply. But even still, I think it would be more statistically sound to have uniform ranges.

 

[Huntsman]

Originally Posted by Arrogant Bastard

Yeah, I just noticed that, actually. Some of your categories/ranges are quite a bit smaller than others, which makes interpreting the distribution from the graph somewhat misleading. I hear you about how there's a big emotional/maturity difference among certain age ranges that is much more significant than the numerical age range would imply. But even still, I think it would be more statistically sound to have uniform ranges.
Agreed. I thought (hoped) that by doing it this way I could milk the most information from the data -- sizing the small ranges so they can be summed into the 'appropriate' size for the statistically correct histogram. In Excel it is looking quite normal, which is not what I really expected (or desired). Self selection bias amongst the younger members? ~ H

 

[Don Carlos]

Originally Posted by Huntsman

Agreed. I thought (hoped) that by doing it this way I could milk the most information from the data -- sizing the small ranges so they can be summed into the 'appropriate' size for the statistically correct histogram. In Excel it is looking quite normal, which is not what I really expected (or desired). Self selection bias amongst the younger members? ~ H
I just think it's your survey design that's the issue. You should've designed a poll with uniform age ranges. It's not the poll-maker's job to tweak the data input mechanism in service of a hypothesis; it's the poll-maker's job to design a sound poll, and then interpret the data that comes out of it. Like I said, and as we seem to agree: there are a lot of sticky wickets here, because certain age differences feel a lot "bigger" than others. No one would argue, for instance, that there's a world of difference between a 20-year-old and a 25-year-old that isn't equalled in the difference between a 60-year-old and a 65-year-old. But that's the sort of interpretation or theorizing you shouldn't be trying to bake into a poll. All you want with this poll is a histogram of ages by uniform range. The analysis comes afterward. Anyway, keep on keepin' on. The poll results are interesting, even if they are a bit dicey.

 

[Huntsman]

Originally Posted by Arrogant Bastard

I just think it's your survey design that's the issue. You should've designed a poll with uniform age ranges. It's not the poll-maker's job to tweak the data input mechanism in service of a hypothesis; it's the poll-maker's job to design a sound poll, and then interpret the data that comes out of it. Like I said, and as we seem to agree: there are a lot of sticky wickets here, because certain age differences feel a lot "bigger" than others. No one would argue, for instance, that there's a world of difference between a 20-year-old and a 25-year-old that isn't equalled in the difference between a 60-year-old and a 65-year-old. But that's the sort of interpretation or theorizing you shouldn't be trying to bake into a poll. All you want with this poll is a histogram of ages by uniform range. The analysis comes afterward. Anyway, keep on keepin' on. The poll results are interesting, even if they are a bit dicey.

Hrm. Am a little confused. Bear with me: I am an engineer, not a statistician! Or even a pollster for that matter. Are you saying that the configuration of the ranges is affecting the outcome of the poll (by outcome I mean affecting the randomness of the sample), not merely the presentation of the results by SF? SF is, of course, displaying a misleading distribution, but the correct one is constructable without interpolation:

The original point was just to see what the distribution was like, and if it was, as I suspected, not normal. Then as I was filling in the ranges the concern about differentiation amongst the younger members came up and I figured I could split the lower ranges into summable smaller ranges and still preserve the integrity of the sample while getting more data. While I can sum data into the standard 5-year range, I can never split it out if it is not differentiated when I get it. I didn't think this graph would be intrinsically affected by the choice of ranges, unless somehow I am attempting to get too sophisticated for my knowledge! ~ H

 

[Don Carlos]

When you collapse the categories and make the ranges uniform, as you seem to have done with that histogram, we start to see the real story playing itself out. And it's all good. My only point was that this is more work than you really needed to do. If you had just used uniform-range categories to begin with, you wouldn't have had to collapse categories after the fact. In this case, the problem was easily fixed. But it's possible to design a survey in which uneven ranges within categories can cause problems. Especially when the categories are more subjective, and/or have connotations associated with them. For instance, if I designed a survey asking people how many pounds they gained last month, there's a big difference between using: a) 1-2 pounds; 3-5 pounds; 6-10 pounds and b) 1-5 pounds; 6-10 pounds In this case, the survey respondent might not have gained an exact figure, so much as a moving target that roughly centers around one number. Let's say I gained about 2.5 pounds. In survey "A," I'd be inclined to make myself feel better by rounding downward to 2. In survey "B," I'd have no choice but to select the first range. This can be problematic, because now the survey results will look very different depending on which ranges you used when designing the survey. Simply collapsing or reorganizing the categories in survey "A" to match those of "B," after the fact, will not provide the same results as having designed survey "B" in the first place. In your case, like I said, you're probably fine just collapsing the categories. That's because pretty much no adult thinks in terms of fuzzy grey areas in between ages. If someone is 21.5 years old, they think of themselves as definitively 21. Hell, people tend to cling to 21 until the day before they turn 22, even though they're numerically closer to 22. That's why your age survey can be condensed without negative consequences on data integrity. Bear in mind that I'm not a statistician, either. But like you, I work with numbers for a living. I doubt either of us is an expert in statistics, so for all I know, some statistics PhD can pop up and ***** slap us both here.

 

[summit1986]

26-30

 

[TheDarkKnight]

Originally Posted by Arrogant Bastard

you're probably fine just collapsing the categories.

It would be interesting to see the age range in different sub forums.

My subjective feeling is that in the Men's Clothing sub forum, the age range would be much older, with most 30 - 45.

This would also have the effect of not making us look so old AB

Though I'm sure you will agree, 30 is really the start of adulthood for a man these days (subject to the responsibility someone has this may be younger or older)

 

[Don Carlos]

Originally Posted by TheDarkKnight

It would be interesting to see the age range in different sub forums.

My subjective feeling is that in the Men's Clothing sub forum, the age range would be much older, with most 30 - 45.

This would also have the effect of not making us look so old AB

Though I'm sure you will agree, 30 is really the start of adulthood for a man these days (subject to the responsibility someone has this may be younger or older)

Yeah, I'd also love to see how this poll plays out across the different subforums. Men's Clothing would definitely appear to skew older, at least if the pics in the WAYWRN thread are any indication. But then again, maybe there's just a high correlation between older people and people who post pics on that thread. At any rate, I definitely get the impression that it's younger here and older there.

As for the "30 is the new 20" thing, yeah, it's becoming true for many reasons beyond the emotional these days. The tanked economy is going to have serious, long-term repercussions for young people growing up right now, so the whole late-bloomer thing is going to become the norm even more than it already has.

 

[TheDarkKnight]

Originally Posted by Arrogant Bastard

As for the "30 is the new 20" thing, yeah, it's becoming true for many reasons beyond the emotional these days. The tanked economy is going to have serious, long-term repercussions for young people growing up right now, so the whole late-bloomer thing is going to become the norm even more than it already has.


Most definitely, and an interesting conversation to be had.

First of all with too many in university education, most whipper snappers don't leave education until 21 or 22. Given that a lot of degrees are frivolous, and an extended beer-a-thon, this really extends childhood further.

Add in as you say the economy and lack of credit and lack of jobs means buying property is prohibitive in the UK until most are 30, and house sharing university style is now very common because of affordability.

There is also the huge social upheaval of less marriage, and most young women seeming to be against having children until later and later.

In fact the average age for having kids and getting married is about 29 for women in the UK, 32 for men I believe.

As you say with less money and less job security, this will put these bastions of responsibility ( and adulthood) back further.

I am also very interested in the social effect this recession will have - if you look at large social upheavals like the 1st World war, and also after the Great depression, there was a large social backlash each time. One of hedonism, the other of facism.

Given the UK certainly had its hedonism in the credit bubble, I wonder what backlash we will see in 5 or 6 years?

Certainly there is a long period of austerity ahead, and I look forward to the change of government in May.

Interesting times!

One thing's for sure - I have a firm eye on moving if the going gets worse!

 

[either/or]

Originally Posted by RSS

^^^Late 50-something here. But I still remember my first 39th birthday ... and the 5th 39th ... and the 10th 39th ... and the 15th 39th ...

Originally Posted by JLibourel

I just voted. I'll be turning 68 on Thursday.
-you guys need to go hexadecimal. I'm 30, btw.

 

[Jerome]

In my mid-30s- the best age!!