Hi all,
Welcome to Proper Admin for June 2016.
If you need a concept refresher, click here.
Thanks for reading this past month.
Tim
Blog Admin
Commentary...
Nothing specific for this month- just got a couple of nice comments after my Bluets post and wanted to mention my appreciation for those.
What was all the chatter about? May 2016 blog posts...
*5/3- Life Changing Books, 2012
As a follow up to this post, I am going to go back and pull out any reading from this year that I recommend. I might have that ready in time for next week's May book recap.
*5/6- Proper Admin, May 2016 (Part One- general)
*5/8- Proper Admin, May 2016 (Part Two- Bluets)
In the ongoing saga regarding the nationality of my phone- I recently got the auto-fill to return 'Okehampton'- a small town within Devon in the southwest of England. A town of just over seven thousand, Okehampton does not even have a properly professional soccer club.
*5/10- Proper Admin, May 2016 (Part Three- books)
*5/17- More Or Less- TB test and Simpson's Paradox
I do not plan on turning this space into a statistics blog. However, in light of this post, I thought putting my book notes up for How To Lie With Statistics might be of some value. Please see below the footnotes if interested.
I first read this book for my high school statistics class. If I recall correctly, the point of reading the book was to learn how statistics are inherently deceiving. We then proceeded to spend all of junior year learning just how to deceive people with statistics.
I merged my TB test experience with the topic of that podcast because I thought it highlighted an oft-cited statistical anomaly about disease incidence- as testing becomes better or more frequent, observed cases of a disease will increase even if the rate at which the disease infects people does not change.
Learning how to respond to increased observance of positive test results is one of the many significant challenges facing healthcare in the coming decades. For the news consuming public, learning how to respond to sensational news headlines about exploding disease incidences is just a new form of an old challenge.
*5/20- Pre-2011 life changing books pt 1- Eat This, Not That + The Seven Habits of Highly Effective People
*5/24- Leftovers- Eat This, Not That (Dieting) + The Seven Habits of Highly Effective People (Peter principle)
The Peter principle is an example of how systems are assumed to run well until something exposes that this is not the case. From my perspective, the biggest challenge that will face any system's success comes when an individual acting in his or her best interest causes outright harm at the system level (or, at best, has no ability to bring the system any closer to its intended goal). After all, if an individual's impact to the system is negative, what is the point of remaining in the system at all?
Once an individual reaches this line of reasoning, the only logical outcomes are the impending collapse of the failing system or the removal of the individual from the system by those tasked with maintaining its smooth operation. Does this mean all systems are 'doomed' to run smoothly until an unforeseen complication alienates the individuals within it?
This pattern might be inevitable in some cases but I think there is plenty of room overall for improved designs. Unfortunately, too often I see system designs that weigh the details over the principles and the result is an endless parade of rules, restrictions, and exceptions. Keeping the simplest idea in mind at all times- individuals acting in their best interest should make things better overall- and making sure changes maintain this concept would go a long way toward solving some of the chronic issues that plague many systems.
*5/30 (Monday) This Post Took Forever
This post did take a long time to sort out and finalize. I probably took out all kinds of stuff, half-written or otherwise, and found a few other things that I wanted to go on about but failed to figure out a way to include it in the post. Some of it is on the back burner for future posts. I guess this is bound to happen from time to time.
One angle I wanted to explore more in this post was the concept of getting 'back to the basics'. I wanted to write about the idea because I've always found that going 'back to the basics' has left me feeling re-energized in times where I've felt sluggish. As the post developed, it became clear to me that this idea was only loosely related to the topic at hand and focusing on the impact of the layoff was a truer thread along which to describe the events.
This blog, in a certain way, represents something close to the 'back to the basics' concept. I've always found myself writing over the years in some form, no matter the specific situation. One great part of my old job was the platform to write meaningfully. In reflecting on this most recent post, I do not think it is accidental that, despite the long list of things I described losing track of post-layoff, finding a way to continue writing in some fashion was the first of those things I properly addressed.
What else happened in May?
We'll cover my usual reading recap separately in my post next Tuesday. However, just as a preview, here are three books that only if you already know that you like the author:
*Jane by Maggie Nelson
*Episode in the Life of a Landscape Painter by Cesar Aira
*Downtown Owl by Chuck Klosterman
One note about Jane. One of my least favorite things about a book is one that starts with a long introduction or preface that is not 'numbered' as part of the main text. Usually, these sections are numbered separately, perhaps using Roman numerals. What happens in these sorry publications is that you will read page after page, perhaps for an hour, before reaching page...three.
The first thing I noticed about Jane was that this book 'starts' on page 13, suggesting immediately that it was potentially, if not a literary masterpiece, at least a publishing milestone. As far as I am concerned, reading would be a better activity if all books correctly numbered their pages.
And it never hurts to give a reader credit for finishing twelve pages after the first syllable (*).
Anything else new?
My hair is now officially long enough to have a ponytail. This happened on May 28.
I see...
I went bowling again this month for the first time in ages. Bowling is fine. Sometimes I do OK, sometimes I don't. It is usually a good time for a group.
What I cannot figure out is how this game came into existence. Bowling, I learned after several minutes of focused research, is hundreds of years old. Its history is somewhat colorful- I enjoyed the story of how ten-pin bowling was invented after gambling on nine-pin bowling was outlawed by the state of Connecticut in the 1800s. The Simpsons episode where Homer buys Marge a bowling ball for her birthday is a minor classic.
But I could not find an answer to my question of why anyone would invent a game that required someone at the far end to stack the pins back up and return the ball to those playing at the other end. Perhaps doing this was considered a plush job among 12th century servants or maybe it was 'first to nine'. Who knows? I stopped googling.
There are some loosely related cousins of the game, such as bocce, whose objective is to throw a ball close to another thrown ball. This one makes sense to me- no extra setup required. I could see this being popular while everyone was waiting for the ink to dry on the Magna Carta.
Did you leave the apt at all?
The start of the month marked my annual appearance at the Walk For Hunger. Unlike famous or important people, I use the world 'appearance' literally here for I appeared at it, technically volunteering, and noticeable to anyone with functioning eyes. This marks the eleventh or twelfth year since my first time there (when I walked) and perhaps my seventh or eighth time actually there.
Although the event is basically the same thing every year, there does always seem to be something that happens which makes a given year memorable. This year, a chain of events led me to ride the #1 bus along its entire route- Dudley Square to Harvard Square. This highly pointless accomplishment was, I believe, notable enough for being the first time I have ever ridden an MBTA bus from start to finish.
A few days later, I attended a book reading for When Breath Becomes Air. The memoir was written by a young surgeon named Paul Kalanithi while he was dying from brain cancer. Although he passed before the manuscript was entirely finished, it was completed by his wife, Lucy Kalanithi. She was the one who did the actual reading and fielded questions from the crowd at the Cambridge Main Library.
I've only attended a handful of book readings and I always come away wondering why certain questions get asked. In this case, a lot of the questions asked were technical ones about death, dying, and cancer. Some of the others were about her own feelings or experiences after the publication of the book.
At the time, some questions seemed to me like they were missing the whole point of the book. The questions I thought mattered were those addressing the concerns of meaning, of purpose, of how to live courageous within the process of dying (1).
It did occur to me after the reading, though, that this reaction says a lot more about me than about anyone else in the audience. Like most books, When Breath Becomes Air means something different to each person who reads it. For some, the book really is about love, family, and relationships. For others, it is about facing mortality. A doctor might find this book useful as a tool to improve her job performance while a reader like me might find it very helpful in the grieving process.
This thought hit home when I was discussing the reading a couple of days later. One person remarked how they really liked the title of the book- the idea that breath becoming air beautifully captured the essence of what they found remarkable about the book. Others more or less seemed to agree with that idea.
I can understand where that thought is coming from but I'll admit here that I gave the title little thought until that remark. At the time of reading, it had no impact on me. My breath always becomes air, I suppose. But the comment did underline for me how the book will mean something different for different people. I think the points of universal agreement about this book are the great truth found in its writing and the courage of the author in openly describing his own experiences.
The other angle about book readings is that I always come up with good questions to ask five minutes after the event ends. In this case, I would have liked to hear Lucy Kalanithi elaborate more on her remark that end of life care is the biggest question facing health care in the 21st century.
In terms of all-time book reading rankings, I probably would still slot it just behind Chuck Klosterman's appearance a few years ago at Porter Square Books. This was memorable for his late arrival (fifteen minutes past seven) which he blamed on airlines. He then launched off on a five minute speech about how there should be a government option for airlines so that even the lousiest private carriers would seem luxurious by comparison.
On a final note, I made another trip to the Museum of Fine Art. What I realized on my most recent trip is how much I enjoy art that captures power. This might mean a painting of a massive storm, figuratively portraying the size and force of nature, or a fifty foot tall statue towering over you that must have required great physical exertion to complete.
I also find fascinating anything that makes me think differently of the way other people live. The MFA has a couple of replica rooms that you can walk through. These are set in a specific place and point in time- ancient China, Victorian England, etc- and the rooms show the furniture and art of the time in a way that comes across more directly to me than is managed by a standard exhibit layout.
These displays reminded me of two exhibits I enjoyed at other museums. The first, at Chicago's Art Institute, was forty or fifty dioramas of American rooms from the 1600s through today. The other was at Boston's Science Museum and showed different dinner plates of people from around the world along with stories about the diners and the nutritional information of each meal.
The Mega Cities exhibit at MFA features a project showing the 'pigeon's eye' view of one of Asia's most crowded cities (New Delhi?). This was set in a 8x4x8 cube and designed so you could walk in and see, on three walls plus the ceiling, the densely packed slums of the city from the vantage point of a bird flying overhead.
Stunning is the right word to describe this one, I think, at least from the way it impacts a viewer. In terms of the nuts and bolts, I would imagine a determined eight year old could make something similar with a box of Lego blocks- provided those blocks were of two and three story aluminum-roofed buildings which housed several families, each block pressed right against (and occasionally, on top of) its neighbor.
Back on Tuesday with a reading recap for May, a little more about art, and a podcast I recently listened to.
Tim
Footnotes...
1. The other thing I should say about the book reading...
To be fair, since Lucy Kalinithi did not write most of the book, I recognize that it is probably challenging to come up with questions about the writing of the book itself since she did not actually write it. Still, I did enjoy some of the process nuggets that came through, such as how some of the writing in the book was not part of the original manuscript but rather was compiled from Paul Kalinithi's emails or papers.
OK- the post is over but if you want to read a bit more about statistics...
A few weeks ago, I suggested in a post that I would start sharing completed 'book notes' in some fashion on this space. Here are the notes I pulled together back in January for the statistics book I referenced above. There is no commentary or analysis here beyond what I did back at the time of reading- so feel free to skip if you are here just for the pure blog element of what I am putting up.
*How To Lie With Statistics*
January 16, 2016
*Core Purpose*
This book is a simple walkthrough of how statistics cannot simply be accepted for face value immediately upon their presentation. It describes numerous themes or types of statistical problems and provides some instruction on what to look out for.
Ultimately, knowing basic statistical techniques is vital given how much of our time is spent being asked to assess statistics. Outright rejection of such methods is not an option- people do not give up reading because some authors have been known to deceive with words- so improving one’s ability to interpret and best use statistics is vital.
*Using a sampling technique*
*The danger in sampling lies with how representative it is of the population
-> The less representative, the more likely that the results will be biased
*A good test to determine the quality of your sample- does every member of the population have an equal chance of being in the sample?
*Any product of a sampling technique comes with a statistical error- this represents how all the possible combinations of samples interact with the one that is being used
*Ways to imply causation…*
*Correlation is not causation- just because the changes in one measurement lead to changes in another does not mean that the change in the first measurement causes the changes in the second- it could just as easily have to do with a third factor to which the first two are linked
*When comparing groups, make sure they are comparable!
-> Rates of death in two groups might differ wildly if people more likely to die are also more likely to join one group over another
*Try to find out the denominators!*
*Using a small denominator can help sell a result because small changes can be reported as big percentage adjustments
*One technique to prove something you are unable to prove is to use a different figure and then try to relate it back to what you are trying to prove
-> A common way to accomplish this is to state changes in numerators without noting a proportional change in denominators- the observations are different but the percentage of occurrence is unchanged
*Dangerous ways to present the facts*
*Stating averages or medians when being less than those measures causes no relevant harm is dangerous because it frames the natural but perhaps slower than average progress of a perfectly normal person as being slow or a sign of defect
*Percentages lend an air of precision to which none exists- and percentages carried to multiple decimal places can often be more silly than informative
*A fifty percent pay cut is offset by a one hundred percent pay raise- this illustrates a type of problem that comes up when the base changes from two statistics that on the surface appear linked
*Five simple questions to ask of any statistic*
1. Who says so?
Essentially, before accepting the conclusions of any statistic, consider if the party reporting the statistic is invested in how the result is interpreted. Also consider how much the statistic has changed since the raw figures were calculated or recorded- the more analysis and reporting that takes place, the more you need to look for both conscious and unconscious bias within those analysis and reporting steps.
2. How do they know?
Focus on the compilation technique and determine if it influenced the result. Was the sample used biased or self-selected? Are the conclusions based on a large enough sample size?
3. What’s missing?
Is there a comparison statistic that needs to be presented in order to provide context? Oftentimes, a seemingly shocking observation is put into context when the same result is measured within entirely different groups or populations. It could also be that a seemingly significant percentage is rendered meaningless when the raw figures are presented (or vice-versa).
4. Did somebody change the subject?
This is perhaps the most important question if the target statistic is being used to represent something different- for example, reported cases of a disease can be used to represent incidence of a disease. A variation of the same issue is seen when the same question is asked in different contexts- people may lie upward about their age if it entitles them to certain privileges or benefits but downward if they are concerned about perceptions of their youth. Taking age responses from these types of questions to then represent the demographics of a population are difficult to take into account given these potential biases.
5. Does it make sense?
Pretty straightforward idea- statistics have a way of forcing the suspension of common sense and it is always worth the extra minute to consider if what you are being told makes sense compared to what is already known to be true.
Welcome to Proper Admin for June 2016.
If you need a concept refresher, click here.
Thanks for reading this past month.
Tim
Blog Admin
Commentary...
Nothing specific for this month- just got a couple of nice comments after my Bluets post and wanted to mention my appreciation for those.
What was all the chatter about? May 2016 blog posts...
*5/3- Life Changing Books, 2012
As a follow up to this post, I am going to go back and pull out any reading from this year that I recommend. I might have that ready in time for next week's May book recap.
*5/6- Proper Admin, May 2016 (Part One- general)
*5/8- Proper Admin, May 2016 (Part Two- Bluets)
In the ongoing saga regarding the nationality of my phone- I recently got the auto-fill to return 'Okehampton'- a small town within Devon in the southwest of England. A town of just over seven thousand, Okehampton does not even have a properly professional soccer club.
*5/10- Proper Admin, May 2016 (Part Three- books)
*5/17- More Or Less- TB test and Simpson's Paradox
I do not plan on turning this space into a statistics blog. However, in light of this post, I thought putting my book notes up for How To Lie With Statistics might be of some value. Please see below the footnotes if interested.
I first read this book for my high school statistics class. If I recall correctly, the point of reading the book was to learn how statistics are inherently deceiving. We then proceeded to spend all of junior year learning just how to deceive people with statistics.
I merged my TB test experience with the topic of that podcast because I thought it highlighted an oft-cited statistical anomaly about disease incidence- as testing becomes better or more frequent, observed cases of a disease will increase even if the rate at which the disease infects people does not change.
Learning how to respond to increased observance of positive test results is one of the many significant challenges facing healthcare in the coming decades. For the news consuming public, learning how to respond to sensational news headlines about exploding disease incidences is just a new form of an old challenge.
*5/20- Pre-2011 life changing books pt 1- Eat This, Not That + The Seven Habits of Highly Effective People
*5/24- Leftovers- Eat This, Not That (Dieting) + The Seven Habits of Highly Effective People (Peter principle)
The Peter principle is an example of how systems are assumed to run well until something exposes that this is not the case. From my perspective, the biggest challenge that will face any system's success comes when an individual acting in his or her best interest causes outright harm at the system level (or, at best, has no ability to bring the system any closer to its intended goal). After all, if an individual's impact to the system is negative, what is the point of remaining in the system at all?
Once an individual reaches this line of reasoning, the only logical outcomes are the impending collapse of the failing system or the removal of the individual from the system by those tasked with maintaining its smooth operation. Does this mean all systems are 'doomed' to run smoothly until an unforeseen complication alienates the individuals within it?
This pattern might be inevitable in some cases but I think there is plenty of room overall for improved designs. Unfortunately, too often I see system designs that weigh the details over the principles and the result is an endless parade of rules, restrictions, and exceptions. Keeping the simplest idea in mind at all times- individuals acting in their best interest should make things better overall- and making sure changes maintain this concept would go a long way toward solving some of the chronic issues that plague many systems.
*5/30 (Monday) This Post Took Forever
This post did take a long time to sort out and finalize. I probably took out all kinds of stuff, half-written or otherwise, and found a few other things that I wanted to go on about but failed to figure out a way to include it in the post. Some of it is on the back burner for future posts. I guess this is bound to happen from time to time.
One angle I wanted to explore more in this post was the concept of getting 'back to the basics'. I wanted to write about the idea because I've always found that going 'back to the basics' has left me feeling re-energized in times where I've felt sluggish. As the post developed, it became clear to me that this idea was only loosely related to the topic at hand and focusing on the impact of the layoff was a truer thread along which to describe the events.
This blog, in a certain way, represents something close to the 'back to the basics' concept. I've always found myself writing over the years in some form, no matter the specific situation. One great part of my old job was the platform to write meaningfully. In reflecting on this most recent post, I do not think it is accidental that, despite the long list of things I described losing track of post-layoff, finding a way to continue writing in some fashion was the first of those things I properly addressed.
What else happened in May?
We'll cover my usual reading recap separately in my post next Tuesday. However, just as a preview, here are three books that only if you already know that you like the author:
*Jane by Maggie Nelson
*Episode in the Life of a Landscape Painter by Cesar Aira
*Downtown Owl by Chuck Klosterman
One note about Jane. One of my least favorite things about a book is one that starts with a long introduction or preface that is not 'numbered' as part of the main text. Usually, these sections are numbered separately, perhaps using Roman numerals. What happens in these sorry publications is that you will read page after page, perhaps for an hour, before reaching page...three.
The first thing I noticed about Jane was that this book 'starts' on page 13, suggesting immediately that it was potentially, if not a literary masterpiece, at least a publishing milestone. As far as I am concerned, reading would be a better activity if all books correctly numbered their pages.
And it never hurts to give a reader credit for finishing twelve pages after the first syllable (*).
*This post just hit its seven thousandth word so I'll start to wrap up. Nice job so far, reader.
Anything else new?
My hair is now officially long enough to have a ponytail. This happened on May 28.
I see...
I went bowling again this month for the first time in ages. Bowling is fine. Sometimes I do OK, sometimes I don't. It is usually a good time for a group.
What I cannot figure out is how this game came into existence. Bowling, I learned after several minutes of focused research, is hundreds of years old. Its history is somewhat colorful- I enjoyed the story of how ten-pin bowling was invented after gambling on nine-pin bowling was outlawed by the state of Connecticut in the 1800s. The Simpsons episode where Homer buys Marge a bowling ball for her birthday is a minor classic.
But I could not find an answer to my question of why anyone would invent a game that required someone at the far end to stack the pins back up and return the ball to those playing at the other end. Perhaps doing this was considered a plush job among 12th century servants or maybe it was 'first to nine'. Who knows? I stopped googling.
There are some loosely related cousins of the game, such as bocce, whose objective is to throw a ball close to another thrown ball. This one makes sense to me- no extra setup required. I could see this being popular while everyone was waiting for the ink to dry on the Magna Carta.
Did you leave the apt at all?
The start of the month marked my annual appearance at the Walk For Hunger. Unlike famous or important people, I use the world 'appearance' literally here for I appeared at it, technically volunteering, and noticeable to anyone with functioning eyes. This marks the eleventh or twelfth year since my first time there (when I walked) and perhaps my seventh or eighth time actually there.
Although the event is basically the same thing every year, there does always seem to be something that happens which makes a given year memorable. This year, a chain of events led me to ride the #1 bus along its entire route- Dudley Square to Harvard Square. This highly pointless accomplishment was, I believe, notable enough for being the first time I have ever ridden an MBTA bus from start to finish.
A few days later, I attended a book reading for When Breath Becomes Air. The memoir was written by a young surgeon named Paul Kalanithi while he was dying from brain cancer. Although he passed before the manuscript was entirely finished, it was completed by his wife, Lucy Kalanithi. She was the one who did the actual reading and fielded questions from the crowd at the Cambridge Main Library.
I've only attended a handful of book readings and I always come away wondering why certain questions get asked. In this case, a lot of the questions asked were technical ones about death, dying, and cancer. Some of the others were about her own feelings or experiences after the publication of the book.
At the time, some questions seemed to me like they were missing the whole point of the book. The questions I thought mattered were those addressing the concerns of meaning, of purpose, of how to live courageous within the process of dying (1).
It did occur to me after the reading, though, that this reaction says a lot more about me than about anyone else in the audience. Like most books, When Breath Becomes Air means something different to each person who reads it. For some, the book really is about love, family, and relationships. For others, it is about facing mortality. A doctor might find this book useful as a tool to improve her job performance while a reader like me might find it very helpful in the grieving process.
This thought hit home when I was discussing the reading a couple of days later. One person remarked how they really liked the title of the book- the idea that breath becoming air beautifully captured the essence of what they found remarkable about the book. Others more or less seemed to agree with that idea.
I can understand where that thought is coming from but I'll admit here that I gave the title little thought until that remark. At the time of reading, it had no impact on me. My breath always becomes air, I suppose. But the comment did underline for me how the book will mean something different for different people. I think the points of universal agreement about this book are the great truth found in its writing and the courage of the author in openly describing his own experiences.
The other angle about book readings is that I always come up with good questions to ask five minutes after the event ends. In this case, I would have liked to hear Lucy Kalanithi elaborate more on her remark that end of life care is the biggest question facing health care in the 21st century.
In terms of all-time book reading rankings, I probably would still slot it just behind Chuck Klosterman's appearance a few years ago at Porter Square Books. This was memorable for his late arrival (fifteen minutes past seven) which he blamed on airlines. He then launched off on a five minute speech about how there should be a government option for airlines so that even the lousiest private carriers would seem luxurious by comparison.
On a final note, I made another trip to the Museum of Fine Art. What I realized on my most recent trip is how much I enjoy art that captures power. This might mean a painting of a massive storm, figuratively portraying the size and force of nature, or a fifty foot tall statue towering over you that must have required great physical exertion to complete.
I also find fascinating anything that makes me think differently of the way other people live. The MFA has a couple of replica rooms that you can walk through. These are set in a specific place and point in time- ancient China, Victorian England, etc- and the rooms show the furniture and art of the time in a way that comes across more directly to me than is managed by a standard exhibit layout.
These displays reminded me of two exhibits I enjoyed at other museums. The first, at Chicago's Art Institute, was forty or fifty dioramas of American rooms from the 1600s through today. The other was at Boston's Science Museum and showed different dinner plates of people from around the world along with stories about the diners and the nutritional information of each meal.
The Mega Cities exhibit at MFA features a project showing the 'pigeon's eye' view of one of Asia's most crowded cities (New Delhi?). This was set in a 8x4x8 cube and designed so you could walk in and see, on three walls plus the ceiling, the densely packed slums of the city from the vantage point of a bird flying overhead.
Stunning is the right word to describe this one, I think, at least from the way it impacts a viewer. In terms of the nuts and bolts, I would imagine a determined eight year old could make something similar with a box of Lego blocks- provided those blocks were of two and three story aluminum-roofed buildings which housed several families, each block pressed right against (and occasionally, on top of) its neighbor.
Back on Tuesday with a reading recap for May, a little more about art, and a podcast I recently listened to.
Tim
Footnotes...
1. The other thing I should say about the book reading...
To be fair, since Lucy Kalinithi did not write most of the book, I recognize that it is probably challenging to come up with questions about the writing of the book itself since she did not actually write it. Still, I did enjoy some of the process nuggets that came through, such as how some of the writing in the book was not part of the original manuscript but rather was compiled from Paul Kalinithi's emails or papers.
OK- the post is over but if you want to read a bit more about statistics...
A few weeks ago, I suggested in a post that I would start sharing completed 'book notes' in some fashion on this space. Here are the notes I pulled together back in January for the statistics book I referenced above. There is no commentary or analysis here beyond what I did back at the time of reading- so feel free to skip if you are here just for the pure blog element of what I am putting up.
*How To Lie With Statistics*
January 16, 2016
*Core Purpose*
This book is a simple walkthrough of how statistics cannot simply be accepted for face value immediately upon their presentation. It describes numerous themes or types of statistical problems and provides some instruction on what to look out for.
Ultimately, knowing basic statistical techniques is vital given how much of our time is spent being asked to assess statistics. Outright rejection of such methods is not an option- people do not give up reading because some authors have been known to deceive with words- so improving one’s ability to interpret and best use statistics is vital.
*Using a sampling technique*
*The danger in sampling lies with how representative it is of the population
-> The less representative, the more likely that the results will be biased
*A good test to determine the quality of your sample- does every member of the population have an equal chance of being in the sample?
*Any product of a sampling technique comes with a statistical error- this represents how all the possible combinations of samples interact with the one that is being used
*Ways to imply causation…*
*Correlation is not causation- just because the changes in one measurement lead to changes in another does not mean that the change in the first measurement causes the changes in the second- it could just as easily have to do with a third factor to which the first two are linked
*When comparing groups, make sure they are comparable!
-> Rates of death in two groups might differ wildly if people more likely to die are also more likely to join one group over another
*Try to find out the denominators!*
*Using a small denominator can help sell a result because small changes can be reported as big percentage adjustments
*One technique to prove something you are unable to prove is to use a different figure and then try to relate it back to what you are trying to prove
-> A common way to accomplish this is to state changes in numerators without noting a proportional change in denominators- the observations are different but the percentage of occurrence is unchanged
*Dangerous ways to present the facts*
*Stating averages or medians when being less than those measures causes no relevant harm is dangerous because it frames the natural but perhaps slower than average progress of a perfectly normal person as being slow or a sign of defect
*Percentages lend an air of precision to which none exists- and percentages carried to multiple decimal places can often be more silly than informative
*A fifty percent pay cut is offset by a one hundred percent pay raise- this illustrates a type of problem that comes up when the base changes from two statistics that on the surface appear linked
*Five simple questions to ask of any statistic*
1. Who says so?
Essentially, before accepting the conclusions of any statistic, consider if the party reporting the statistic is invested in how the result is interpreted. Also consider how much the statistic has changed since the raw figures were calculated or recorded- the more analysis and reporting that takes place, the more you need to look for both conscious and unconscious bias within those analysis and reporting steps.
2. How do they know?
Focus on the compilation technique and determine if it influenced the result. Was the sample used biased or self-selected? Are the conclusions based on a large enough sample size?
3. What’s missing?
Is there a comparison statistic that needs to be presented in order to provide context? Oftentimes, a seemingly shocking observation is put into context when the same result is measured within entirely different groups or populations. It could also be that a seemingly significant percentage is rendered meaningless when the raw figures are presented (or vice-versa).
4. Did somebody change the subject?
This is perhaps the most important question if the target statistic is being used to represent something different- for example, reported cases of a disease can be used to represent incidence of a disease. A variation of the same issue is seen when the same question is asked in different contexts- people may lie upward about their age if it entitles them to certain privileges or benefits but downward if they are concerned about perceptions of their youth. Taking age responses from these types of questions to then represent the demographics of a population are difficult to take into account given these potential biases.
5. Does it make sense?
Pretty straightforward idea- statistics have a way of forcing the suspension of common sense and it is always worth the extra minute to consider if what you are being told makes sense compared to what is already known to be true.
