Archive | May, 2010

Louis CK “Learn to Appreciate Technology” Perspective

30 May

Information V Data – Meh

14 May

The other day I was reading a post on the Word Stumpet about when “Good is good enough” for you? This is actually a fascinating question, and there is an entire science devoted to it, although they are usually approaching it from another direction, such as its monetary value. This reminded me of a couple of saying I will throw in at this point because I cannot see a more appropriate location and I really want to include them.

“The perfect is the enemy of the good” (Voltaire)

“Better sooner is good enough.” This was a slogan of the Soviet Union in their military production, and we all know how well that ended.

And one I may have originated: “If you don’t have time to do it right the first time, what makes you think you have time to do it over?” (I’m going to totally take credit until and earlier reference than circa 1989 when I said it to a secretary comes forth.)

This is a soapbox lecture I give intermittently and spontaneously – and I really should have written this down a long time ago. Since I have not previously organized this, I expect it will meander like a conversation where I extrapolate on an idea and well beyond too. To begin will, a definition. The difference between data and information. This light bulb moment came from James Burke. Watching his series of “Connections 1 – 3” and “The day the Universe Changed” will entrance most nerds. About 25 years ago, while I was at the UC Davis Medical Center in Sacramento, and residing in the people republic of Davis, I attended a lecture sponsored by the local PBS station featuring – wait for it – you’ll never guess, James Burke. As I recall, the talk was about how the information age was going to change work. As an example of data, he said “If I told you that the person next to you, who is coughing, is sick that is just data-because it doesn’t change your behavior. If I told you that person had a highly communicable deadly disease, that would be information because it would dramatically affect your behavior.” What struck me about that analogy was that it nicely summed up the kind of work I do. I take data and change it into information. That information is provided to the clinician and should change their behavior in how they manage their patient’s treatment.

What factors determine the value of information? Information needs to be timely, and needs to be accurate, it needs to be clear and concise and easily comprehended. There may be some other factors, and if and when I think of them I will modify the post to include them. My usual example of the information source is the Wall Street Journal. An 100% accurate data that is old is not particularly informative. The value of knowing 100% accurate stock market results of 20 years ago is not particularly helpful. Data that would help you get the value of a stock or the direction of the market tomorrow that is 100% accurate is highly valuable. The presentation of the data is also important. One of the items that has been sorely lacking in medical reports is the graphic display that are common in many other forms, particularly PowerPoint, which did not exist at the time of the initial lecture. Many a time I will present data as a list of numbers and dates. This is not nearly as informative as a graphic presentation would be. Medical reports typically are alphanumeric text. When presenting imaging data, ideally a hyperlink between the report and the image would be preferable. Although this concept, presented by me approximately 15+ years ago, the method that is currently in use for most medical imaging still does not incorporate this approach and only now are medical records beginning to take advantage of technology present on the Internet for over a decade. The integrated electronic medical record is moving towards connecting data, but not too hyper-linking specific report information.

The value of the information depends on the individual weighting of each value. If timeliness is extremely critical, then a mostly accurate and understandable report is more valuable than any perfectly grammatical and 100% accurate report later. A graphic display of the data is clearer and more understandable than a listing of dates and data points. In the example that Charlotte used with her editor, accuracy and clarity are of a significantly greater value than timeliness, therefore the effort to correct and perfect the document would take precedence over getting it in sooner. For a blog, where the timeliness and clarity of the issue are more important, the grammatical accuracy and typographical errors, unless really gross, do not significantly removed from the value. (I keep telling myself this.) A blog unlike a newspaper or magazine article, can be updated and corrected in real-time.

A mathematical formula expressing this may look something like this:

V = Fn(nA, mT, oC)

Where V is equal to the value of the information.

Fn() Is a function that determines the value.

A Is the accuracy and n is the weight value.
T Is the timeliness and m is its weighting value.

C is the clarity and O is its weighting value.

The classic example of this is a graft showing Napoleon’s Russian campaign of 1812. In a visual form it clearly and easily demonstrates the horrendous losses. Imagine trying to understand this if it was listed out as a table of numbers, or worse, listed out as dates and numbers in paragraph form.

Click to enlarged.

Note: If you know of a source with a higher resolution of above graph, please leave a comment. I am disappointing in the clarity of the writing. Also I would like add links to where it could be purchased and how to analysis and present statistical information.


Adorable kittens on a slide

11 May

How to reduce stress and get out of difficult situations.

11 May

This product is AWESOME! Buy today!

What’s next?

9 May

A reoccurring problem I have, and occasional conversation about, is the indeterminacy of the word next. It seems straightforward enough. Here is the definition from a dictionary.

next |nekst|
adjective
1 (of a time or season) coming immediately after the time of writing or speaking : we’ll go next year | next week’s parade.
• (of a day of the week) nearest (or the nearest but one) after the present : not this Wednesday, next Wednesday | [ postpositive ] on Monday next.
• (of an event or occasion) occurring directly in time after the present or most recent one, without anything of the same kind intervening : the next election | next time I’ll bring a hat.
2 coming immediately after the present one in order or space : the woman in the next room | the next chapter | who’s next?
• coming immediately after the present one in rank : building materials were next in importance.
adverb
on the first or soonest occasion after the present; immediately afterward : wondering what would happen next | next, I heard the sound of voices.
• [with superlative ] following in the specified order : Joe was the next oldest after Martin.
noun
the next person or thing : one moment he wasn’t there, the next he was | the week after next.
preposition archaic
next to : he plodded along next him.
PHRASES
next in line immediately below the present holder of a position in order of succession : he is next in line to the throne.
next to 1 in or into a position immediately to one side of; beside : we sat next to each other. 2 following in order or importance : next to buying a whole new wardrobe, nothing lifts the spirits quite like a new hairdo! 3 almost : Charles knew next to nothing about farming. 4 in comparison with : next to her I felt like a fraud.
the next world (according to some religious beliefs) the place where one goes after death.
what next an expression of surprise or amazement.
ORIGIN Old English nēhsta [nearest,] superlative of nēah [nigh] ; compare with Dutch naast and German nächste.

In computer programming there are segments of code called loops, and in the language called BASIC, they are called up for-next loops. An example of code would look something like the following (I have not written BASIC in decades, so this may not be exactly  syntactically correct).

10    For I = 1 to 10
20    Print I
30    Next I
40     End

This code would print out the numbers one through 10.

My problem with the word next is that in computer code if you stated, “turn right at the next street.” Whichever street immediately came up, you would turn right. In real life, if you are 5 yards from the street and your wife tells you to turn right at the next street, and you turn right she will snap at you saying,  “I said the next street not this street!” Now if you were a quarter of a mile away and was told to turn right at the next street, you’d be expected to turn right at that street, not drive through it and go on to the following street. There is some poorly defined distance where the immediately upcoming street is “this street” but not the “next street.” Once you go beyond that point  this street is the next street or the phrase  “this street” really should be “this upcoming street” because “the street” doesn’t make any sense. Confused?

It’s not just streets it’s also days. If it is Monday and I tell you next Tuesday, that really means a week from tomorrow, probably, not the next day-which by the way is not Wednesday but would be Tuesday. If it’s the previous Tuesday and I tell you next Tuesday, then it’s a week away. OK, so which day does next Tuesday become a week from the upcoming Tuesday as opposed to the immediately upcoming Tuesday?

The world may never know-Sort of like how many licks does it take to get to the center of a Tootsie Pop.

By the way the comment section actually does work.

What are the Chances?

2 May

A while back Judy on Zebra Sounds linked to a “Brain Test” What Sex is your Brain. While interesting, it is certainly not nearly long enough to be statistically significant, and like most pop psychology, it simplifies things to the point of absurdity. For example, being empathetic and recognizing the signs that would trigger empathy are two distinct items, as, I think, anyone with Asperger’s Syndrome or who have dealt with them would know.

There was one section, where my score I found particularly odd – that is identifying emotions with eyes. I got 5 of 10 correct. There were only 2 or 3 that I looked at and my response was represented by one of the 4 choices. This lead me to the question “What is the probability of getting 5 right by chance?” followed by getting 5 right given 2 or 3 known. I am only going to address the first question.

If this was a binomial problem with a 50:50 outcome then is because a simple combinational problem. For example what are the chance of getting 5 heads in ten tosses of a fair coin?

Letting k be the number of heads and n be the number of tosses the formula is:

n!/(k! * (n-k!))

since we don’t care about the order. The ! is factorial, meaning 3! equals 3 x 2 x 1; and by definition (IIRC) 0! = 1

The following table shows the different number of combinations:

Combination
Number wrong
Combinations
0
1
1
10
2
45
3
120
4
210
5
252
6
210
7
120
8
45
9
10
10
1

The sum of all the combinations is 1024, so the chance of tossing 5 heads in 10 flips is 252/1024 = 0.24609 or approximately 25%.

For this test, it is more complicated since there is a 75% chance by guessing of getting the wrong answer; that is getting all 10 wrong would be 0.75^ 10. Getting 5 right and 5 wrong would be 0.75^5 x 0.25^5 x 252. There is only one combination where they are all right or wrong but 252 where half are right and half are wrong.

Probability
Number wrong
Combinations
Individual Prob
Combination Prob
0
1
9.537E-07
9.537E-07
2
45
2.861E-06
2.861E-05
3
120
2.575E-05
3.090E-03
4
210
7.725E-05
1.622E-02
5
252
2.317E-04
5.840E-02
6
210
6.952E-04
1.460E-01
7
120
2.086E-03
2.503E-01
8
45
6.257E-03
2.816E-01
8
45
1.877E-02
2.816E-01
9
10
1.877E-02
1.877E-01
10
1
5.631E-02
5.631E-02

Interestingly, the chance of getting 5 right by randomly picking, apx 6%, is very close to getting them all wrong.

I will delve into the probability of getting 5 answers correct given 2 or 3 answers are know later.

Note: I see WordPress removes the center align tags in the tables – Nice.

The Muppet Show – Julie Andrews

1 May

Lonely Goat Herder