A new tree-ring and fire scar study from SMU and the University of Arizona finds that today’s mega-wild fires in the Southwest are unusual.

The 1,400-year record encompassed the Little Ice Age (1600 to mid 1800s A.D.) and the Medieval Warm Period (800-1300 A.D.) and found that fire incidence was nearly the same under both cool and warm, wet and dry conditions.

Forest policy of fire suppression prevented forests being naturally thinned by relatively small ground fires. The result was a build up of brush which exacerbated fires to produce even larger, more destructive wild fires. The researchers say, “The U.S. would not be experiencing massive large-canopy-killing crown fires today if human activities had not begun to suppress the low-severity surface fires that were so common more than a century ago.”

“This new study is based on a first-of-its-kind analysis that combined fire-scar records and tree-ring data for Ponderosa Pine forests in the southwest United States.”

“Fire scientists know that in ancient forests, frequent fires swept the forest floor, often sparked by lightning. Many of the fires were small, less than a few dozen acres. Other fires may have been quite large, covering tens of thousands of acres before being extinguished naturally. Fuel for the fires included grass, small trees, brush, bark, pine needles and fallen limbs on the ground.”

“The fires cleaned up the understory, kept it very open, and made it resilient to climate changes because even if there was a really severe drought, there weren’t the big explosive fires that burn through the canopy because there were no fuels to take it up there.”  ”The trees had adapted to frequent surface fires, and adult trees didn’t die from massive fire events because the fires burned on the surface and not in the canopy.”

Read the entire press release from SMU here.

This study implies that attempts at “sustainable” forest management and endangered species issues have in fact made our forests more unsustainable.

See also:

Drought in the West
Droughts in the Southwest put in perspective

A new study has produced “statistically significant” results supporting the (false) hypothesis that adults who listen to children’s songs become younger, not just feel younger, but become actually chronologically younger.

In statistical studies, there is always the possibility that results that seem to support a hypothesis could have occurred purely by chance and actually have nothing to do with the hypothesis.  In statistical jargon, the measure of this possibility is called the “P-value” which is an estimate of the probability that results occurred purely by chance.  Researchers strive for a low P-value.  The current arbitrary gold standard in some sciences is a P-value less than or equal to 0.05, meaning that there is only a one in twenty chance that the results are accidental.  The results with a P-value less than 0.05 are called “statistically significant” and are deemed to support the hypothesis.

To publish a statistics-based paper in a prestigious scientific journal, the results must be “statistically significant.” There are many ways, however, in which researchers can manipulate their data to meet the mathematical requirements for “statistical significance” and still be very wrong in their conclusions.  This seems especially common in health studies.

The study mentioned above was part of a larger study on how researchers can game the system to produce statistically significant results supporting their hypotheses.

Statistician William M. Briggs has an interesting post “How To Present Anything As Significant” in which he reviews a new paper: “False-Positive Psychology : Undisclosed Flexibility in Data Collection and Analysis Allows Presenting Anything as Significant“

In the paper, the authors “show that despite empirical psychologists’ nominal endorsement of a low rate of false-positive findings (P≤0.05), flexibility in data collection, analysis, and reporting dramatically increases actual false-positive rates. In many cases, a researcher is more likely to falsely find evidence that an effect exists than to correctly find evidence that it does not. We present computer simulations and a pair of actual experiments that demonstrate how unacceptably easy it is to accumulate (and report) statistically significant evidence for a false hypothesis.”

In the paper introduction the authors say:

“Our job as scientists is to discover truths about the world. We generate hypotheses, collect data, and examine whether or not the data are consistent with those hypotheses. Although we aspire to always be accurate, errors are inevitable.  Perhaps the most costly error is a false positive, the incorrect rejection of a null hypothesis. First, once they appear in the literature, false positives are particularly persistent. Because null results have many possible causes, failures to replicate previous findings are never conclusive. Furthermore, because it is uncommon for prestigious journals to publish null findings or exact replications, researchers have little incentive to even attempt them. Second, false positives waste resources: They inspire investment in fruitless research programs and can lead to ineffective policy changes. Finally, a field known for publishing false positives risks losing its credibility.

In this article, we show that despite the nominal endorsement of a maximum false-positive rate of 5% (i.e., p≤.05), current standards for disclosing details of data collection and analyses make false positives vastly more likely. In fact, it is unacceptably easy to publish “statistically significant” evidence consistent with any hypothesis.

The culprit is a construct we refer to as researcher degrees of freedom. In the course of collecting and analyzing data, researchers have many decisions to make: Should more data be collected? Should some observations be excluded? Which conditions should be combined and which ones compared?

Which control variables should be considered? Should specific measures be combined or transformed or both?

It is rare, and sometimes impractical, for researchers to make all these decisions beforehand. Rather, it is common (and accepted practice) for researchers to explore various analytic alternatives, to search for a combination that yields “statistical significance,” and to then report only what “worked.”

The problem, of course, is that the likelihood of at least one (of many) analyses producing a falsely positive finding at the 5% level is necessarily greater than 5%.”

The authors provide guidance for authors and reviewers to remedy the situation.  Briggs summarizes them as follows:

The authors list six major mistakes that users of statistics make. They themselves used many of these mistakes in “proving” the results in the experiment above.

“1.Authors must decide the rule for terminating data collection before data collection begins and report this rule in the article.” If not, it is possible to use a stopping rule which guarantees a publishable p-value: just stop when the p-value is small!

 ”2. Authors must collect at least 20 observations per cell or else provide a compelling cost-of-data-collection justification.” Small samples are always suspicious. Were they the result of just one experiment? Or the fifth, discarding the first four as merely warm ups?

“3. Authors must list all variables collected in a study.” A lovely way to cheat is to cycle through dozens and dozens of variables, only reporting the one(s) that are “significant.” If you don’t report all the variables you tried, you make it appear that you were looking for the significant effect all along.

“4. Authors must report all experimental conditions, including failed manipulations.” Self explanatory.

 ”5. If observations are eliminated, authors must also report what the statistical results are if those observations are included.” Or: there are no such things as outliers. Tossing data that does not fit preconceptions always skews the results toward a false positive.

 ”6. If an analysis includes a covariate, authors must report the statistical results of the analysis without the covariate.” This is a natural mate for rule 3.

The authors also ask that peer reviewers hold researchers’ toes to the fire: “Reviewers should require authors to demonstrate that their results do not hinge on arbitrary analytic decisions.”

The bottom line here is to always be somewhat suspicious of papers whose results depend upon statistical manipulation or modeling versus papers that present actual observations.

“There are three kinds of lies: lies, damned lies and statistics.” – Mark Twain

See also:

Statistical Games #1

Statistical Games #2 Stroke for Stroke

To help celebrate Arizona’s centennial, there is a new web portal that “offers a tour of the people, places, and events that defined our past and are shaping our future. The Arizona Experience is your passport to Arizona’s hidden treasures. Interactive features allow you to customize your tour. Visit Arizona’s iconic landscapes, listen to the oral histories of descendants of early explorers, settlers, and miners, or discover how our leading innovations in biotechnology, alternative energy, and high-tech products are creating a promising tomorrow. Each month during the 2012 Centennial year will launch a new theme to showcase the 48th state.”

The theme for March is mining and minerals.  The features include:

Mining Arizona’s Metals – interactive map of active mines in Arizona, Morenci mine flyover, and surface and underground mining techniques slide show.

Rock Products – Building Arizona – interactive cement plant tour, map with  locations and mineral commodities of more than 300 quarries or mines, videos and photo gallery.

Featured Artist – World renowned mineral photographer Jeff Scovil presenting a photo gallery of some of his best images of Arizona minerals, as well as a short video on “how to photograph minerals”.

Miners Story – Video gallery of the men and women of San Manuel recounting their experiences living and working in one of Arizona’s historic mining communities.

H. Mason Coggin Photo Collection – Arizona historic mines and miners photo gallery.

The Arizona Experience is a dynamic, multimedia, 4D web environment with interactive maps, hundreds – soon to be thousands – of images, historical time-lines, flyovers of iconic landscapes, interviews with Arizona leaders, featured artists, hours of videos – onsite and at the Arizona Experience YouTube channel, and oral histories that capture the experiences of the men and women that shaped the state.

According to Dr. Michael Conway of the Arizona Geological Survey, “We used Microsoft Research’s new Layerscape visualization software to produce the 3D flyovers, and we worked closely with ESRI to broadcast interactive maps that incorporate spatial data, content, interactive timelines, and photo galleries.  These dynamic tools and extraordinary content are tailor made for teachers challenging their students to explore Arizona’s past, examine its present, and imagine its future.”

Take a few minutes to look over the home page, and sample the various features.  There is more to it than initially meets the eye.  There are lots of nooks and crannies that bring up very interesting material.

Click on http://arizonaexperience.org/ to start your tour.