Why Do We Randomize in Experiments?

Curious people experiment.  We want answers we can't find on Google.  Unfortunately, without some training, experimental results may not be valid.

One of the first things to know about experimenting is the need to randomize. First, randomization helps avoid the pitfalls of bias. Bias is the tendency for the data to all move one direction due to something other than what we are interested in investigating.

Second, statistical analysis makes some assumptions. We assume all errors will gather about an average and approximately cancel to zero. Without that assumption, many statistical tools are useless. So we randomize to justify making that assumption.

Design of experiments is a complex subject. But learning some simple principles will help you get started with everyday experiments that answer YOUR questions.

Happy experimenting!


123112

Book Review -- Probability: An Introduction

Probability is the basis of not only statistics, but of almost all science, including mathematics, physics, chemistry, biology, engineering, marketing, etc.

In general, we rely on a number of principles of randomness.  Entropy is a measure of disorder or randomness.  An example of entropy is our reliance on air molecules spreading themselves randomly.  This is critically important to our ability to breathe.  If air typically arranged itself in a pattern that excluded part of a room, some people would suffocate.  Fortunately, nature is somewhat random.

This foundational subject is covered in a nice introduction by Samuel Goldberg.  Probability:  An Introduction was originally published in 1960, but has been republished by Dover since 1986.  The current cover price is only $16.95.  You get a lot of bang for your buck with this book.

Although the style is stodgy, the information is solid.  The book contains five chapters, which cover sets, probability in finite sample spaces, sophisticated counting (which is combinations and permutations), random variables (which are functions), and binomial distributions.

After reading this book, you will understand the theoretical basis of an introduction to statistics.  The book covers the theory behind expected values (means), variance and standard deviations, z values in a normal distribution, as well as much more.  This is the best explanation I have ever received on combinations and permutations.  Overall, reading the book has been very satisfying.

If you can get past the style, this book is worth the time and money.

Happy reading!

Keep Your Own Medical Records - Part 2

If the only thing stopping you from collecting and keeping you own medical data is knowing where to start, read on...

Obtain a permanently bound notebook. A regular composition notebook will do. Write the date somewhere in each entry, preferably the top. Record the basic facts of each event along with any special notes. Include any advice you receive with names of those present.

If you decide to make control charts, you can create simple time line charts in your notebook. More involved charts will probably require software.

One website you will want to look at is when evaluating your medical record history is http://www.hhs.gov/ocr/privacy/hipaa/understanding/consumers/index.html.

More information on charting to come...

113012

Do Yourself A Favor - Keep Your Own Medical Records

No matter how mundane your doctor visit may seem, keeping your own record of each visit is a good idea.  When you think about it, you are paying the doctor to do three things for you: 1. Evaluate your condition. 2. Make a record of your condition. 3. Treat you when necessary.

Even though you are paying for a record to be made, you do not have control over that record. The doctor can note whatever they want, even if it is not really what happened. You do have the right to dispute your record, but without a solid contradictory record, you will have a hard time doing that.

Besides being able to write whatever they care to write in your record, the doctor can redact information at a later date. So the record is subject to change, again, outside of your control.

Why should you care about what the doctor does with the record they make, own, and keep outside of your control? Sometimes what happened previously really matters. Some possibilities are insurance disputes, malpractice events (which you think will never happen to you, but may), recommendation of procedures requiring a second opinion, and ability to donate blood or organs.

Perhaps the best reason to keep your own record of your medical care is your health. You spend more time with yourself than anyone else. You have the opportunity to track data related to medical visits as well as everyday events.

One extreme example of tracking everyday events is diabetic care. Diabetic people must continually evaluate blood sugar levels. Relying on a doctor to do that is not a realistic option for maintaining good health. In this case, use of control charts would provide valuable data on which events/foods/etc. affect blood sugar levels the most. Lifestyle adjustments could be made based on the data collected.

Another example of event tracking would be control charts for weight. Exercise and diet could be included for additional decision making ability.

Just tracking your own immunizations could keep you out of the office for the occasional tetanus shot that you can't remember if you need or not.

Control your own data. Control you own life.

Analytic Heros - You'll Be Suprised!

What do Sherlock Holmes, Adrian Monk, and Shawn Spencer have in common?  Well, yes, they are all fictional detectives, but they are also analytic experts.

Thinking about the definition of analytics (the discovery and communication of meaningful patterns in data), it makes sense that detectives are at the top of the hero list.

My favorite quote from Sherlock Holmes is "...when you have eliminated the impossible, whatever remains, however improbable, must be the truth."  Here Sherlock demonstrates the depth of his understanding of analytics.  The heart of the matter is believing the data will lead you to correct conclusions.

Adrian Monk also puts profound faith in data by way of evidence.  In one TV episode, he proclaims the murderer is a man who has been in a coma.  While everyone one around him disbelieves due to improbability, Monk follows the data.

Finally, Shawn Spencer uses analytics to promote his psychic powers.  He boldly states a dinosaur has killed a victim despite the improbability of that event.  In the end, the data proved him correct.

Do you find you believe the data (evidence) in front of your senses?  In general, we tend to ignore the improbable, despite the evidence.  The very act of estimating improbability makes us all statisticians.  It takes conscience training to overcome our skepticism of improbable data and believe the logical conclusions before us.

Happy sleuthing!

Who Needs Analytics?

Everyone uses analytics, whether they know it or not.  Wikipedia says "Analytics is the discovery and communication of meaningful patterns in data." (http://en.wikipedia.org/wiki/Analytics)

Toddlers discover patterns in sound data to learn language, identify food vs. non-food object data to eat, find repetitive patterns in behavior data to play games, and generally categorize action data in order to interface with other people. 

Learning to recognize analytics in everyday activities empowers.  Government agencies, medical providers, retail stores and their marketing agents, as well as other organizations collect and analyze data about us around the clock.  When we collect and analyze our own data, we level the playing field.  When we act on that analysis, we're ahead of the game.