Vindication! August 3, 2010Posted by Matt in science, statistics.
Tags: Climate Change, global warming, mathematics, science, statistics
If you recall, we had a rather harsh winter several months ago, complete with snow and ice and days of frigid temperatures that we are usually spared in the Mid-South. This, in turn, led many to proclaim with certainty that Global Warming was a complete and total hoax. It was plastered across blogs and headlines and, of course, Fox News that the entire idea of warming was an Al Gore fabrication.
I got into a bit of a heated argument on a friend’s site at the time over this when I claimed that a cold winter in a small location meant nothing when measured against the mountains of statistical data collected over the years. I was summarily vilified for even entertaining the thought that mathematics could show that it existed. Around this same time the pseudo-controversy named “Climategate” became a phenomena in the reality-denying community, and I went into a bit more detail on my blog about the nature of statistics.
In the final analysis, we cannot use a small amount of data to statistically disprove or to prove the existence of a scientific theory like global warming. Just as the cold winter did not disprove it, the hot summer today does not prove it. It takes a lot of data, collected worldwide, to show that this theory is most likely true – and that is just what scientists have.
I only wish that the Global Warming deniers who were so very sure of themselves in the winter would use their same brand of logic today.
Life as a Random Variable March 8, 2010Posted by Matt in Christian Beliefs.
Tags: Christian thought, god, life, providence, random variable, Satan, statistics, theodicy, theology
Please note that the following blog entry is the product of some of my thoughts from the past day or so and I realize that the ideas presented here may have several holes in them that can’t be covered in just over 400 words.
Life is a series of variables. Some are random occurrences that take place like the roll of a dice and some are dependent upon our actions, which are in turn dependent upon a whole litany of datum from our lives, whether it is education, experience, temperament or others of untold number. Though it may be realistic for us to correctly predict the outcome of individual events or some series of events within a band of error, the quantity of variables becomes so great that it is impossible for us to map out the entirety our lives with any degree of certainty.
I say this because I find myself greatly bothered by the common idea in Christian circles that all things happen according to a plan set forward by an all-powerful God. Once we place the Zoroastrian-influenced concept of a devil in its appropriate mythological place, things become even hairier. The absence of cosmic warfare places the responsibility for evil at our feet and at the divine feet of God. Thus, we must sincerely ask the questions that bother us so without resorting to the pat answers pervading the Christian religion, particularly when it deals with events beyond our control. We look at the random misery around us – from small children dying of cancer to earthquakes decimating some of the poorest regions on earth – and we must ask the simple question, why? If God is truly good and in power of all things upon the earth, why do such events take place?
Maybe there is no plan. Maybe things just happen beyond our control, events which are beyond our ability to stop, and the important thing is how we respond to them. When the die is cast and the random variables of life turn up in a way beyond explanation within the classic paradigm of Christian thought, perhaps we would be better off acknowledging that some things are even outside the realm of God’s providence. This then absolves God of responsibility for evil and instead places the “blame” upon the random statistical fluctuations of an imperfect universe.
Stuff happens and that’s really all there is to it. All we can do is get up, dust ourselves off and move forward, keeping in mind that there is no force in this world more powerful than love. We embrace God and come to the realization that maybe some things are even outside the sphere of divine power. And that, in turn, is okay and perhaps even a bit comforting as we confront the mysteries of life.
On Climate Change and Statistics December 9, 2009Posted by Matt in statistics.
Tags: Climate Change, linear regression, math, science, statistics
add a comment
My job is to foretell the future.
No, I’m not like Miss Cleo. I don’t own a crystal ball or use chicken bones or practice astrology. I don’t receive prophetic visions of impending doom. No, my tools for seeing into the future are found in the realm of mathematics.
I provide much of the statistical background when it comes to forecasting for the multi-billion dollar medical device company for whom I am employed. I work heavily with regression analysis to do this form of numerical soothsaying and have developed a fairly good understanding of the behind-the-scenes processes needed to use these methods. My education background is firmly rooted in this area, with a bachelor’s degree in mathematics and graduate work in statistics, so my life and livelihood are based in the power of numbers.
Thus, I was a bit dismayed to read about the criticism that had been leveled at a few climate scientists over a series of emails that were hacked and stolen. Now, I do not understand all of the intricacies of the science, I don’t claim to know anything about their data collection techniques, nor have I sifted through the piles of data upon which they base their hypotheses, but my ears perked up when talk turned to their statistical methods, especially when one hears the vehemence with which critics based their accusations.
There are two points that immediately come to mind when I think of this case:
1. There is no perfect statistical method. Regardless of the model used or the amount of data collected, there is always room for error. Statisticians can account for this with a probable range of error, but that fact is often overlooked by critics in their zeal for proving an opponent wrong.
2. When forecasting the future, “massaging” the data to fit an obvious trend is often needed. Many times you will find unforecasted anomalies (whether too high or too low) in the data which could skew your statistical model and give an unrealistic forecast. In order to combat this, a statistician may either remove the data point completely or replace it with a more realistic value.
Climate science is a complicated field and one that I certainly do not know enough about to comment. The multiple variables involved in building a statistical model for this sort of event is far more complicated than the regression models I employ. I can imagine that it is an exceedingly difficult field to predict and one fraught with uncertainty, but certainly not one that is impossible.
And I think that if everyone had at least a basic understanding of statistics this debate would be far different and much more productive.
The Battle Against the Brain – part 2 September 15, 2009Posted by Matt in The Battle Against the Brain.
Tags: Christianity, college degrees, education, Evangelicals, statistics
Let us begin our discussion on education and religion with a quick statistic.
A second study, one conducted by the University of Michigan and published earlier this year, breaks down the data into college majors in order to find which areas of study effect one’s “religiosity” to the greatest degree. The research found that the largest negative effect on one’s religiosity (measured by religious attendance and the importance of religion in the students’ lives) came from humanities or social science majors, students who focused on business or education showed an increase in their religiosity, and those majoring in biological or physical science showed no significant changes in church attendance, but their regard for the importance of religion showed a large decrease.
It has become increasingly obvious over the years that churches have lost many of the educated along the way, particularly those who focus on the arts and sciences, two fields that take different paths as they focus on a search for meaning. The phenomena may be attributed to any number of reasons, but for our purposes I have indentified two points in particular that illustrate this large discrepancy:
1) Many beliefs espoused by Evangelical Christians are logically incoherent to the educated.
2) Many Evangelicals use certain scriptures to dismiss education on “secular” topics as unimportant or even evil.
To be continued…
Church Statistics – Pt.3 March 10, 2009Posted by Matt in Christianity, statistics.
Tags: Christianity, church, diversity, education, geography, race, statistics
Yesterday I posted two entries looking at Trinity College’s 2008 American Religious Identification Survey in which we analyzed a few of the statistics regarding denominational preferences, as well as the demographic measures of gender, age, and marital status. Today I would like to sort out some of the findings on the issues of race and on some overall societal changes.
Decades ago, the great Martin Luther King called 11:00 Sunday morning “the most segregated hour in America,” and today, some 4o+ years later, the statistics show that to still be the case. Being a member of a racially diverse church, I find these numbers to be unsurprising, yet still saddening in today’s world. On first look at the numbers on race, I was surprised to see that the highest concentration of black membership is in the Baptist church, but that was before I read what each religious body listed in the survey contained. My initial thought, when looking at the word Baptist, was of the Southern Baptists, a decidedly homogenous group (92% white) if there ever was one. This survey, though, listed not only the SBC, but also the other Baptist conventions, as well as the somewhat cryptic “African-American” denominations. Other than that, it can easily be seen that nearly every religious group has shown a decline for all of the racial groups listed – White, Black, Hispanic, and Asian. The only Christian group that showed a discernible increase was the one marked Christian Generic (non-denominational). The proportion of the “nones” again showed the largest increase across all groups, nearly doubling for each one.
The second group of numbers under the banner of societal changes deals with education, namely the percentage of college graduates age 25 and over. In the overall U.S. population, 27% of the citizens fit into this category, yet, when analyzing American religious groups, we find that only two of them, Mainline Christians (35%) and Mormons (31%), have a greater concentration of the educated. At the bottom of the spectrum of the learned we find those of the Pentecostal/Charismatic tradition (13%) and the Baptists (16%). On the positive side, every grouping showed a percent increase from 1990 to 2008, yet, for the most part, the faithful remain less educated than the general populace. As an aside, the “Nones” also increased to 31%.
The final measure found in this report deals with geography and religion, a calculation that is again enlightening and maybe even a bit surprising. The decline of faith in America can be seen across nearly every state and region, as Catholics and Protestants both show decreases and those known as the “Nones” post, in some cases, very large increases. In no other place is the waning of Christian beliefs more evident than in the Northeast, where we find former bedrocks of Catholicism falling into unbelief – Massachusetts posted a 15 point loss of Catholics, while the “Nones” gained 14 points, Connecticut displayed Catholics losing 12 points and the “Nones” gaining 8, Maine showed Catholics down 9 and Protestants down 7, while the “Nones” gained a whopping 14 points, increasing to a full 25% of the population. It is in Vermont, though, that the “Nones” have really made their case, for it is here that they (34%) are now greater than either the Catholics (26%) or the Protestants (29%). But it is not just in the Northeast that we see this transition taking place, a full 15% of Midwestern count themselves among the “Nones,” while Protestant Christians falter – most noticably in Iowa (decrease from 69% in 1990 to 54% in 2008) and Kansas (from 74% to 61%). Even in the South, where church attendance and right living have long been ways of life, Protestant Christians are hitting the proverbial wall, displaying double digit decreases across a multitude of states – Florida (from 62% to 49%), Georgia (86 to 72), North Carolina (86 to 73), South Carolina (88 to 73), Virginia (76 to 65), and Texas (68 to 48). Today in the South even the percentage of “Nones” has doubled since 1990, from 6% to 12%.
So, now we must take some time for reflection and ask ourselves, “What do we do now?” Obviously, the way that the Christian faith has worked in the past decades is no longer applicable.
Church Statistics – pt.2 March 9, 2009Posted by Matt in Christianity, statistics.
Tags: age, America, Baptists, Catholics, Christianity, denominations, gender, Mainline Christians, marital status, non-denominational, statistics
add a comment
Earlier today I posted part 1 of my look at the latest American Religious Identification Survey as done by Trinity College. This entry will take a look at the demographics of religion in America as seen by the report.
When looking at gender, it is interesting to note that females outnumber males in every Christian religious group looked at by the report. The difference is especially pronounced in the Baptist Church, where women make up 57% of the professing members, in the Pentecostal/Charismatic churches, where they lead the way at 58%, and in the Mainline churches, at 56%. For the purpose of this survey, the Church of Christ has been lumped into the category of Protestant Denominations (has lightning struck these people yet?) along with the Jehovah’s Witnesses and Sevent Day Adventists, where women lead the way (well, sort of) at 55%. It should be noted, however, that men outnumber women in non-Christian religions, including a variety of Eastern faiths, Islam, and New Religious Movements. Also, men hold a huge advantage in the “Nones” category (indicating atheists, agnostics, or those not professing a religion) at 60%.
Age compositon makes up another important component of this demographic study and we again can glean some important and enlightening lessons when carefully looking at these statistics. In the United States, 60% of the adult population is between the ages of 18 and 49, and 40% is age 50 years or older. The largest Christian group, the Roman Catholic Church, follows the age breakdown of the general populace almost exactly, but there are others whose results come in a bit more skewed to either end of the age spectrum. A glaring example of this can be seen in the Baptist church, which is decidedly older than any other group. Only 42% of their membership is below the age of 50 and only 11% is under 30, while a full 21% of their roll is over 70, making them the Methuselahs of the American Christian landscape. On the other hand, the group labeled Generic Christians, which includes non-denominational churches, skews much farther to the youthful side, with 66% of their members under the age of 50, 25% under the age of 30, and only 9% over the age of 70.
The last piece of data to look at under this type of religious demographics deals with marital status, where we find that the majority of those identifying as Christians are currently married, though the percentages for all but two of the surveyed groups are actually a bit lower than the national average of 56%. The two denominations that have a greater percent of married members than the general populace are the Mormons (68%) and the Baptists (60%). The largest proportions of single, never married, members are found in the Generic Christians and Protestant Denominations (both at 23%), and the smallest percentage of singles can be found with the Baptists (13%). Also interesting in this group of numbers is the divorce rate for Christian groups. In the overall American population, 13% of adults are either divorced or separated from their spouse. With the lone exception of the Mormons, every Christian group come in at greater than 10%, with the Pentecostal/Charismatics (16%) and Mainline churches (14%) outpacing the others.
So, if you are male and looking for that special someone and hope to find her in church, you should be careful with what you choose. The idea of a large percent of women in the Baptist church might be enticing, but one should remember that they are most likely married or over 50. My suggestion to you would be to try one of the so-called Generic Christian churches. Though the ratio of women to men may not be as high (52 to 48), they will be more likely to be single and young.
Thoughts? Part 3 will deal with race and social distinctions.
Church Statistics – Pt.1 March 9, 2009Posted by Matt in church, statistics.
Tags: Christianity, god, members, religion, statistics, Trinity College
Trinity College has just released the 2008 American Religious Identification Survey and the results are interesting to say the least. For this first part of our look at the survey, let us investigate the religious self-identification findings and what this may mean for us in the Christian community.
According to the survey, the U.S. adult population (defined as age 18 and over) has increased significantly over the past two decades, from 175,440,000 in 1990 to 207,983,000 in 2001 to 228,182,000 in 2008. This represents an increase of 30.1% since 1990. Of the 1990 total (175M), 86.2% of adults self-identified as Christians, a proportion that has decreased over the years by 10 perentage points, to 76.0% in 2008. At the same time, the number of “Nones” (atheists, agnostics, and those with no stated religion) increased from 8.2% (14.3M) in 1990 to 14.2% (29.5M) in 2001 to 15.0% (34.2M) in 2008, almost doubling their share of the total.
Further breaking down the numbers, we find that the largest denomination remains Catholicism, with 25.1% (57.2M) of the total population, a level that has remained fairly steady for them over the years (26.2% in 1990 and 24.5% in 2001).
Baptists come in next, though they have shown a decline in share since 1990, dropping from 19.3% in 1990 to 16.3% in 2001 to 15.8% in 2008.
Mainline Christians (including Methodists, Lutherans, Prebyterians, and Episcopalians) have shown the largest decline from an 18.7% share in 1990 (32.8M) to 12.9% in 2008 (29.4M).
Nondenominational Christians have proven to be one of the few bright spots, increasing their share from 0.1% in 1990 (194K) to 1.2% in 2001 (2.5M) to 3.5% in 2008 (8.0M).
Not surprisingly, my denominational heritage (Church of Christ) has decreased its share from 1.2% in 2001 (2.6M) to 0.8% in 2008 (1.9M).
Also in this section of the survey, those being questioned were asked about their belief in God:
Regarding the existence of God, do you think…?
There is no such thing – 2.3%
There is no way to know – 4.3%
I’m not sure – 5.7%
There is a higher power but no personal God – 12.1%
There is definitely a personal God – 69.5%
Refused – 6.1%.
What do you think of these findings? Part two will deal with demographics, so stay tuned.
Polling For Truth September 24, 2008Posted by Matt in President 2008.
Tags: Barack Obama, election, John McCain, polls, president, statistics
add a comment
This being an election year, the television news cycles and nearly all publications have become inundated with polling data and statistics, most of which are fairly innocuous, that carry the intent to inform (or perhaps coerce) the average consumer. As a statistician, I am fascinated with polls and numbers and how they are used in public discourse. There are a few pointers to always remember when looking at the reported data:
1. Pay attention to the error. If Obama is leading McCain by 47%-45%, with +/- 3% error, it is a dead heat. If, in the next poll, McCain is now leading 48%-46%, it probably does not show any actual change in the attitude of the general populace. Rather, it is just a random statistical fluctuation.
2. Don’t ever make assumptions based on only one poll. Though statisticians try and poll a solid sample of America, the potential always exists that they could choose more supporters of one candidate over the other, skewing the data.
3. Pay attention to the question asked by the organization giving the poll. Results can be easily manipulated depending on the wording of the question and the order of the choices given.
So, given the way that statistics can be used and misused, the best method for gleaning good information is to look at a cross-section of polling data and pay attention to all aspects of the polls. I’ve done a bit of research, using the data on pollingreport.com, in which I’ve averaged the results from several reports by week, ranging from mid-July to now. This is how it looks so far:
wk1 wk2 wk3 wk4 wk5 wk6 wk7 wk8 wk9
McCain 41 43 42 43 43 43 45 46 45
Obama 46 45 47 45 46 49 46 45 47
You can see it in graph format here:
Note: Week 6 was the Democratic National Convention and Week 7 was the Republican. You can see a definite bounce for each of them.
Each week has anywhere from 5 to 10 polls used and, I placed the data for each of these on a normal distribution curve for their respective weeks in order to identify data that may not be good. For instance,
A Fox News poll for September 8-9 showed Obama at only 42%. Given that his mean score for that week was 45% with a standard deviation of 1.5%, we can deduce that this score falls outside of 2 standard deviations, the 95% confidence level.
Does that make sense? I could talk for hours about this, so you may not want me to get started…
Me and the Office December 5, 2007Posted by Matt in statistics.
Tags: forecasting, math, office, statistics, work
add a comment
While at times it can be challenging and stressful, demand planning is a good field for a math nerd like me to be in. For those of you unfamiliar with what I do, here is a quick description.
I work for Smith & Nephew, a company that manufactures medical devices – the side of the business I work on is that of Trauma, where we produce plates and screws and nails and other equipment for helping along the healing process when a bone is broken. My position deals little with handling the actual products, though. In a work environment such as this, the people in manufacturing have to know what and how much of each product to produce. If they make too much, we’ll have a glut of inventory. If they don’t make enough, customers will be short on product and angry. So, there needs to be a balance between keeping inventory levels manageable and good customer service.
And that is my job.
Through a combination of statistical modeling, customer input, and gut instinct, I (and the other demand planners) create a forecast detailing how much of each product we think will be sold in the coming months.
It is not a very exact science, though and it doesn’t take long before you learn that old business maxim – the one thing that’s right about the forecast is that its always wrong – is true. It is just not possible to get a forecast exactly correct all, or even any, of the time, but we can figure it within a certain percent of error. That being said, many times you need a thick skin, because people are quick to blame the forecast (many times in error) when something goes wrong. So, you always need to be ready to give a defense for your numbers and many times I have cause to break out tables and graphs (sales and marketing people usually need pictures) and try to simplify complicated processes of linear regression to show where numbers come from.
I guess it takes a different (i.e. not normal) type of person to enjoy looking a numbers all day, but that’s fine with me. Who wants to be normal anyway?