?

Log in

No account? Create an account
The Chronicles of Jack [entries|archive|friends|userinfo]
Jack

[ website | SoundStage 64 ]
[ userinfo | livejournal userinfo ]
[ archive | journal archive ]

A question about experimental methods [Mar. 19th, 2011|02:06 pm]
Jack
[mood |pensivepensive]

Been reading a book on experimental methods (nerd!), and been thinking about Jamie's degu and about the waste of having the first year bio students do a lot of dissections when it's not going to be relevant to where a lot of them go (even within Biology, so I heard from a Bio TA).

This has collected into a hunch about a way to make animal testing more efficient, in the sense that the same amount of testing could be done but at the cost of fewer animals.

------------------------
Set up:

Say experimenters A and B are both measuring the effects of different treatments on, say, the red blood cell count of mice. They both need their own control group (no treatment) of mice to compare their treatments against. But the control group in each experiment is, in a lot of ways, the same: a group of untreated mice which we take blood from.

If A and B find out about each other's work (likely), they can compare the conditions in which the mice were raised and sacrificed. If those conditions are sufficiently similar, they can compare the red blood cell counts from each experiment to see if the control groups had effectively the same counts.

-----------------------
The questionable leap:

Are experimenters A and B then able to pool their control groups together into a larger, blocked, control group? Could they both use this meta-group to get more statistical power in the comparison between the (meta-)control and each of their treatments?

If so, this would imply that each experimenter would be able to be more certain about the effects of their treatments without having to increase the sample size of either experiment.

But wait..
What if A and B published their raw data, specifically the data about the control groups and their methods of measurement, into a repository for animal testers?

Say experimenter C, who is measuring the effects of a different treatment but on the same variable, then could she...

- Match her measurement methods to what A and B did or...
- Use a smaller control group and check that for equality with A and B's controls...

...and use the collective AB or ABC control group as the control for her experiment?

-----------------------------------------------------------------

I suspect there's already a lot of standardized procedures in place for such testing, so sufficient similarities could happen often.

But how often is the same variable measured in different experiments? Would the decay in statistical strength from applying control to a similar but ultimately independent experiment be too much to actually save any animals? Would experimenters toss outlier control groups in favour of the associated meta-control, and would this cause an under-reporting in the inter-block variance of the meta-control? Would experimenters alter their experiments in order to be able to use a published meta-control even if it was detrimental to the independent experiment?

I feel like there's potential here, and that experimenters working independently on similar experiments are wasting information that could be shared. But I'm wary about the validity of using that information. Different experiments are different, the whole point of an experiment to try something unique. Would such a repository not validly be able to link control groups together? Would people try to anyway to make their experiments seem more significant? Then there's statistical assumptions like independence, what happens to that?

This needs more thought.

Your input?

I need to know more about: Bayesian Statistics, Resampling.
link3 comments|post comment

The daily grind [Feb. 10th, 2011|03:15 pm]
Jack
[mood |creativecreative]

What if there was a button hidden down by the docks of a city, I'm thinking of London, which if pressed at exactly between 4:51pm and 4:52pm each day would produce $100. Once it produces its first $100, it has to be pressed every day or else it will stop working forever.

Imagine someone activates this button and, by trying to recreate the conditions of payment, figures all this out and assumes the last condition (maybe because of something they read in the library).

This becomes the person's 'job', where they make their way to the button every day and press it in a way discrete enough that nobody figures out what she's doing. Maybe all the other jobs she finds require her to be there until 5, maybe this becomes a farce or commentary on the rigidity on labour when a lot of the jobs with this requirement have no obvious reason to have a schedule that includes 5pm other than the workday tradition (Transcriber comes to mind).

So the button becomes the job.

At first this seems fantastic. It's almost $3000 per month for doing no useful work. Sure it's a bit of a commune from where she lives, and close to rush hour too, but who cares?!

Then it hits. Every day. Every day she has to be there at the button regardless of weather and traffic and crowding on the underground. She decides to move closer into town to get to the button faster each day. Even then it takes almost two hours each day to get there and back and back to whatever she was doing. It's also expensive to live in town, which is eating into her button payments.

It's harder to see her friends now that she's moved into the city. She's having a hard time meeting new people because she doesn't have a job to go to where she has to interact with people. It's hard to go out and meet new people even though she' an interesting person because she's afraid of being asked what she does for a living.

This isn't what she went to college for, and she's starting to feel a personal decay from her lack of self-actualization. She gets accustomed to drifting in neutral gear and it's eating away at her. It makes London, which has more to offer than almost any other city on earth to her, feel mundane and pointless. She needs a vacation, a vacation from doing nothing.

But she can't. The button must be pressed every day.

Does she trust someone else in on her secret, to 'buttonsit' while she's away? Would she have to pay someone to do what she's been doing for free for months?

Then the story is left unresolved because that's how reality works. We get stuck just like this and the narrator has no right nor ability to tell you how get out of your own ruts.

The title of the story would be "Nine to Five".
linkpost comment

Local Storage Idea [Aug. 27th, 2010|11:41 am]
Jack
I have to put a new bedframe into the storage locker downstairs today when I get home, but I know it's already full of books and My Little Ponies, so it's going to be a squeeze to get it in there, even disassembled. It would be nice, considering the three-to-five thousand people in apartment buildings within *thinks* a 500m bubble of where I live, for there to be some sort of independent storage company nearby. I suppose it makes sense for all the storage places to be way out at the northern tip of town where the land is cheaper, given that everyone who needs storage drives anyway.

'cept for me, of course.

Here's an idea:

Instead of these sprawling storage lockers which are little more than a collection of spare garages, how about a building which is mostly a metal frame full of wooden lockers that you can take out and roll where you need them. I mean a miniature trailer that you could slot into this building, where you would pay to rent both the slot in the building and the trailer.

You could even couple this with a handful of electric scooters to rent so that you could hook the locker-on-wheels up to the scooter, scoot the 1km or so to your apartment building, load up the locker like you would a trailer or a pickup truck, and scoot back. There, you've used much less power, and you've cut out hiring the driver to haul your stuff all the way to the end of town and back entirely.

Also, if you need something specific and small from the locker, you can walk over the storage location, and pick it out, assuming it's not packed deep in there.

What about the increased space cost in town? Two solutions:
- First, people in apartment buildings tend to have less stuff and aren't going to be storing things like boats and ATVs, so each locker doesn't have to be as big, but you can charge more per cubic metre of space than if you were renting out in bulk.
- Second, for people intending to not fill their locker with bowling ball, I don't see any reason why couldn't slot your locker into a lift and have it stacked on top of other lockers (all held in place in the metal frame of course). For bonus sustainability points, such a lift would likely be used infrequently enough that solar panels could supply a reasonable portion of the energy needed to operate the lift.

There, done. Now you've taken the space of about 20 parking spaces or one house lot, and turned it into a one-person rental office with, room for 6 scooters and 128 such lockers (192 if you stack three high). 128 Rentals at $25 a month is $3200, and another if each scooter is rented out 1 hour/day on average at $5/hr (a lot more during moving days, many days not at all), that's another $900/mo. $4100 a month, let's see your twenty parking spaces do that?
linkpost comment

N [Aug. 13th, 2010|12:57 pm]
Jack
New Jersey Devils
Boston Bruins
Carolina Hurricanes

New York Islanders
Buffalo Sabres
Florida Panthers

New York Rangers
Montreal Canadiens
Tampa Bay Lightning

Philadelphia Flyers
Ottawa Senators
Washington Capitals

Pittsburgh Penguins
Toronto Maple Leafs


Chicago Blackhawks
Calgary Flames
Anaheim Mighty Ducks


Detroit Red Wings
Colorado Avalanche
Dallas Stars

Nashville Predators
Edmonton Oilers
Los Angeles Kings

St. Louis Blues
Vancouver Canucks
Phoenix Coyotes

San Jose Sharks
San Jose Sharks
San Jose Sharks
San Jose Sharks

5 MORE points if you can guess why I posted this
link1 comment|post comment

1999-2000 Season [Jul. 21st, 2010|02:15 pm]
Jack
New Jersey Devils
Boston Bruins
Carolina Hurricanes

New York Islanders
Buffalo Sabres
Florida Panthers

New York Rangers
Montreal Canadiens
Tampa Bay Lightning

Philadelphia Flyers
Ottawa Senators
Washington Capitals

Pittsburgh Penguins
Toronto Maple Leafs


Chicago Blackhawks
Calgary Flames
Anaheim Mighty Ducks


Detroit Red Wings
Colorado Avalanche
Dallas Stars

Nashville Predators
Edmonton Oilers
Los Angeles Kings

St. Louis Blues
Vancouver Canucks
Phoenix Coyotes

San Jose Sharks
San Jose Sharks
San Jose Sharks
San Jose Sharks

5 points if you can guess why I posted this
linkpost comment

Monty Haul Problem... for Goats! [May. 9th, 2010|02:14 am]
Jack
This is my explanation of the famous Monty Haul problem, which is a commonly used exercise in how probability. It's not intuitive, but if once you get it you'll have a much firmer understanding of probability.


Say you're on a game show, and you're at the final round where you get to select a prize. There are three doors to choose from; two of the doors have decrepit old vehicles that you have a hard time hauling away for scrap value. These cars are so old that it looks more like the show's crew is using the contestants as a means of hazardous material disposal. We're talking 1978 here, totally worthless. The other door, however, contains something far more valuable than a junky old car: a magical goat. This goat has the power to attract (human) members of the preferred sex to its owner, kind of like a siberian husky that doesn't need walks and you can milk it.

The message here is: Two lousy prizes, and one greatly superior prize.

The catch is that you don't know what door has what prize, and you can only take home the prize behind one door.

So you pick a door, any door. You have no idea what door has that awesome goat, so you just pick a door at random. You chance of getting that goat (and therefore all the ladies/men) are 1 in 3.

But before your choice is finalized, a twist is introduced. (What devilry!) The host reveals one of the doors that you didn't pick, and there's something behind it that might have been a Volvo before something filled it with pinecones and made a home of it. You're now given the option of switching to the other unrevealed door. What do you do, WHAT DO YOU DO??!?!

Your new situation is this: You've picked a door, and you know where one of the wrong doors is. You need to decide whether or not to switch doors and why.

What is the probability/chance of walking away with that awesome goat if you switch/don't switch.

Solution:
Spoiler - SolutionCollapse )
link1 comment|post comment

Goat Fodder [May. 2nd, 2010|03:16 am]
Jack
I thought this up in the tub, it was stuck in my head.

"That's really nerdy." - Jamie
"I don't know what you're talking about, 3am is a fine time to be tie-dying things!" - ...also Jamie

How banks work

Why does a bank pay you interest? They make it back on loans.
Why doesn't a bank offer huge % and charge hue %? Because the bank of Canada has a set amount it will offer to loan/borrow at that other people and banks could always go to.


What's the deal with this thing about banks making money out of nothing? If you count the money you have as the difference between what you have and what you owe then the banks don't make money out of nothing. However, if you could money as something you can spend in order to get goods and services then it's more complex.

Take this last question from the perspective of the bank, pretend for a moment you're buying a Savings Bond for $100 1. When you buy this bond, you exchange $100 cash for a ticket that says that someone owes you $100. From your perspective, you still have $100 in assets, it's just gone step more abstract than you're used to. From the perspective of the entity you lent the money to, their balance sheet still shows zero; they have $100 in cash, but they also have -$100 because of outstanding debt that will need to be paid off at sometime.

HOWEVAH!!!!!!!! The cash can be spent (exchanged for goods or services), but what about the bond? The bond just says "The bearer is entitled to $100 at (sometime soon)", and in our simple world1 that's as good as cash. Surely the bond can be spent as well! So this means, by some clever financial innovation, that $200 in spendable capital now exists where only $100 did before. Both the bond and the cash represent the same $100, and this only holds together because everyone involved in the transaction knows that at sometime the bond will be traded in for cash and then cease to exist, leaving us with the original $100.

We didn't make wealth, we just made the wealth that already existed more liquid (able to be poured into goods or services). This isn't really a new concept. You likely have said something in the past along "I have $200 in Magic Cards" or "This My Little Pony is worth $60", which means that your collections are in a sense an abstract form of money because you could, with some difficultly, turn them into cash.

The only difference is that in the case of the bond, we created something temporary that has a more definite value (that is, its value is agreed upon more closely than bits of cardboard with art and little plastic horses) but is significantly less fun to play with on a day off. This is how abstract money works.

This bond is one of the more straightforward types a financial derivative that exist, which where all the hubbub about fractional reserves and Illuminati start. You see, with financial derivatives, people can have more spending and saving power without a government having to print more money (which dilutes the value of the money that's already out there; remember, gold is valuable because it is rare) This cool because companies can get what they need to operate and grow more smoothly, which often involves hiring people. However, this sucks too because it means that as more financial derivatives are built then more and more spending power is loaded onto each actual real dollar.

When it got to the point where roughly $25 of spending money was loaded onto each $1 of cash (because complex derivatives can be redeemed for simpler derivatives, we can get a ratio of more than 2 to 1), a few of those derivatives couldn't be redeemed because the cash (or asset like a house) that it represented was no longer there (or worth the stated value) and people quickly became less certain how much all these derivatives were really worth. Since these derivatives were only as valuable as people agreed they were (read: were willing to buy them for), their value really did go down. This is how a market crash works.

1 For simplicity this bond is completely certain to be paid off, can be redeemed at some set time soon after like a week, and that the bond pays no interest. Adding factors like interest, risk, and discount rates would be dull and just won't fit in a single Math For Goats episode.





How to make fractals (from what to install to what to punch in)

It bothers me that through all of university I was never given the chance in class to learn how to make a fractal. Sure in a couple of classes we worked through a bit of the theory involved in them, but we never had a lesson on/ assignment for making a pretty fractal. In this episode, we're going to install all the necessary software to make fractals, as well as write the code in with a minimum of theory and have some rendered fractals that the viewer can experiment from and fiddle with to start exploring this field on their own.


Statistical Significance / Correlation

Turbo Burrito - Makes you go really fast! (500 sold)
Turbo Burrito - Food the way you want it, caffenated! (501 sold)

Statistical Insignificance: That extra 1 sold could easily be just random variation.
Practical Insignificance: A difference of 1 sale (0.2%) is nothing you can make a new adveritising campaign around.


A campaign for depicting babies as screaming and nasty in January-June, during which period pregnancy rates go down.

Correlation: Preg rates when down from the start of the campaign.
...is not Causation: ... but we have no way to say that this is BECAUSE of the campaign. As it turns out people get pregnant less in the summer when it's hot to f***.
linkpost comment

Life Ensurance Part 2 [Nov. 21st, 2009|08:56 am]
Jack
(Continuation of Part 1 from last night)

I was thinking about this some more and realized that funding research would be the hard way to increase life expectancy. Sure, it would work, but if you want the most life per dollar, then focusing on preventative measures would be better, at least up to a certain point.

For example, it would be in the best interest of life insurance companies to lobby for a tax on unhealthy food. There's already been talk about this from nutritionists, but they don't have the monolithic financial backing to make it happen that Sunlife, Manulife, and insurance wings of big banks have. That would again require cooperation between competing companies because a tax on junk food helps all the companies and not just the ones paying to make it happen.

There are some incentives that could only be offered to policy holders of participating companies though. What about fitness, swimming, saftety, food-safe, and first-aid classes with a for-profit or break-even price to join for people that are not policy holders and a discount for those that are. What about a discount card for participating grocery stores for healthful food like fresh vegetables that only policy holders had? How about a coupon deal with participating healthful fast-food franchises like Booster Juice and Subway?

This is doubly effective in reducing financial risks from life insurance: First, it encourages current policyholders to live a more healthy lifestyle and therefore live longer and pay into the policy longer before cashing out. Policies like this are more attractive to people that already live healthier than average lifestyles, and would be willing to either pick up a policy or switch policies when they wouldn't have before because of the added incentive of savings to their lifestyle. Bringing someone into your policy who's already living healthy may result in a larger risk reduction than making someone already in your policy more healthy, so this might be a major selling point or focus of such a program.

If it works well, other companies might do the same and the companies that don't will be left insuring the high-risk fat lazy slobs. Sounds like a good way to lose money in a hurry unless you can get a lot of large premiums out of them first.

Furthermore, the cost to the life insurance company for promoting other companies through coupons and specific discounts could be quite small if the life insurance companies shopped around for retailers willing to chip in for a portion of the cost. The incentives would also be a fair excuse to increase premiums slightly to cover the incentives ($50/year increase for a potential $300 savings?)

A lot of this also applies to healthcare insurance, whereas my first post on life insurance companies funding research didn't necessarily spread like that. That means that an even larger (MUCH larger in the States) part of the corporate world could either get together or start competing with each other to make us healthier.

We live longer and better, and they save a lot of money. If it hasn't been done yet, there has to be a flaw right? So where is it? I imagine that people don't want to be controlled by their insurance companies, but how invasive are discounts?
linkpost comment

Institutional Inefficiency, Actuarial Edition [Nov. 20th, 2009|09:22 pm]
Jack
[mood |contemplativecontemplative]

So I asked my Actuarial Sciences (the math/stats of death and insurance), if a life insurance company or collection of companies could sell enough policies that it would be worth their while to invest in life saving research in order to collect premiums and avoid payments longer.

He said, "Oh absolutely, ICBC (Insurance Corporation of British Columbia) puts a lot of money into road safety and even commissions and recommends changes in traffic patterns and road setups just to avoid more car accidents."

Why doesn't that happen for the health of our bodies as well as our cars? Because ICBC is a monopoly. Every car registered in BC has to be covered, by law, by one of their policies. This means that anything they do to make driving less risky benefits them and not any other competing car insurance company. (There are other car insurance companies here, mostly at the retail level; they all work through and alongside ICBC.)

If a life insurance company (let this be Company A) was large enough that they decided that a grant towards cancer research would end up working to the company's financial benefit because it could collect life insurance longer before having to pay it off, then that research would benefit all life insurance companies. The other companies could then reduce their premiums now that the risk of death was reduced, but Company A must keep its premiums the same because it had the added expense of the research grant. This cost difference would eventually pull policy holders away from Company A, and thus costing them big time in the end.

So it's bad for a single life-insurance company to improve our health, but what if all life-insurance companies agreed or were forced to contribute the same portion of their earnings from premiums into health improvements. Each company on its own would be motived to find the most effective way to do this and they could collaborate in order to ensure the most health benefit per dollar (and therefore the most insurance payment savings per dollar) possible happened. Life insurance companies could go from insuring that the survivors are financially taken care of to actually ensuring that life continued.

I'm not the first person to think of this, and it must be obvious to anyone who's worked in the field for more than a year or two, but why hasn't it happened?

I can't understand people. These are supposed to be the most intellectual and financially savvy people in the world, and they can't organize themselves to save money together and earn some good will? What am I missing here? Is there some other reason why it doesn't happen, or are people (even the "best and brightest") really that irrational?

Perplexing.

Possible thesis topic?
linkpost comment

RIP Coralee Lane Davis [Sep. 7th, 2009|11:05 am]
Jack
During this summer, my grandmother's cancer was diagnosed as terminal.

For the records, it was breast cancer that had been dormant for more than 10 years, but had managed to spread to the brain. This is the second cancer related death my two-degree family in 2009, both in Edmonton no less.

She was quite lucid until the last few days to a week, which makes me glad that they didn't bother trying chemo or radiation therapy on an 80-year old woman. I'm also glad that palliative care (final care) is pragmatic like it should be. I think pragmatic is the right word. She got anything she wanted in her last few weeks, without any regard as to whether or not it was bad for her. Chips? Burgers? Beer and morphine together? What's the worst that could happen?

I got to have a Last Phone Call with her.

It's one thing to know that you're talking to someone for the last time before you part ways and head off to worldly adventures. It's something entirely different to know for certain that the person on the other side of the conversation's adventures will be far from worldly after you stop talking.

What do you say to your grandmother on the phone if you know she's dying?

We talked about birds.

I stretched the truth temporally and told her that I was going to start a Master's Degree in September. I would have felt pretty bad if that had turned out to be false. It made her proud and happy and that's what matters. Still, it would have been a jerk move to lie to her.

Mostly we talked about birds though.

What else was there? We had no qualms with each other, no unfinished business. I had nothing useful to teach her and she had no mystical final words of wisdom for me. It wasn't that she wasn't wise, but there was nothing that would have been more meaningful or good for both of us than to talk about how much better robins are than crows.

I guess that's a good sign. I miss her dearly, but I don't regret anything except spending more time with her, but I could say that even if I had.

Rest In Peace Cora Lee, you've earned it.
link1 comment|post comment

navigation
[ viewing | most recent entries ]
[ go | earlier ]