Public opinion polls

There will be elections in Brazil in 2014. This time we will elect the president, the state governors, part of the senate, the congress and the state assemblies. Due to Brazilian law, open campaign is still prohibited, but some movement already started. For example, everybody knows that president Dilma Roussef will apply for reelection and there are two declared "opposition" candidates. In this context the public opinion polls acquire a great importance. According to these polls, President Dilma is frankly favorite  to win the election in the first turn and her adversaries barely reach  16%  and 12% of the "vote intentions".

A recently published poll gained considerable attention in Brazil. Two different polls organized by the same institute (IBOPE, short for "Brazilian Institute for Public Opinion Research") gave different results concerning the same population, in one of the polls Pres. Dilma would have about 43% of vote intentions, while in the second  this number would be 38%.

There was a fuzz in the social networks about this result, first because the News Services (which are in the majority supporting the opposition) published headlines like "President Dilma fell in the IBOPE poll!". There was an outcry from Pres. Dilma's supporters because they  insinuated political use of the result (which is obvious), because the former poll range extended for a longer period and ended after the second poll (which gained full coverage from the press).

Apart from the political use of such results, one should look at the problem with a scientific point of view. What is a public opinion poll?

The answer to this question is: a statistical inference measure.

Let us consider what is in fact a poll. This process probes one population (the country voters) extracting a small sample and probes one question. Let us remain simple, let us suppose the question is binary, having only two possible outcomes. As every math, physics, engineering student learns, this problem is equivalent to probing a box containing a large number of pebbles colored black and white. Supposing the fraction of white peebles is p, that the sample size is N, the number n of white pebbles in the sample will be given by the Binomial probability distribution:

Let us look what this means for a typical sample size used in these polls (N=2600) and let us assume p=0.38 (that is, the population is composed of 38% of white pebbles). The result is given by the red line in the figure below.

 The fact that we measured p=0.38 means that we actually drew 988 white pebbles in the 2600 sized sample, but let us come back a little and ask before the measure what would be the probability to draw 988 pebbles provided p=0.38. This number is f=0.01612 (that is 1.6212%). Let us suppose now our measure of p is wrong and it is actually larger (it could be smaller too). I plotted in the figure (blue line) what would be the result if p=0.40. The probability to draw 988 white pebbles in this case would be f=0.00182 (or 0.182%). It looks small, but it is actually more than 1/10 of the ideal value. In fact, if we drew 100 N=2600 samples out of the box, any result between ~960 and ~1020 would be likely obtained in p=0.38.

Everybody who works with statistical inference knows this fact. In a statistical measure we will never be fully confident in the result. We will always run into the possibility to commit two types of errors, the first error, called "type I" error means that you accepts as true a proposition that is actually false, in the present case, that assuming p=0.38 is wrong. This is the origin of the "confidence interval" concept, which most laymen know as "error margin". There is, however, the type II error, which is rejecting a statement which is actually true, in the present case, rejecting p=0.40 after drawing 988 white pebbles. The type II error is more difficult to control and one could easily loose track of it if one tries to circumscribe the type I error to low probability values.

The determination of the confidence interval to this problem usually requires approximating the binomial distribution by the normal distribution (which is a good hypothesis in the present case) or maybe using more sophisticated methods, like using bayesian inference, but one should take care not to use the central limit theorem here, since the number of samples is actually 1.

Therefore the result of a public opinion poll is simply a (educated) guess. Its results must be analysed with care and never in the way described above (the "conclusion" that president Dilma is falling in the public preference). This use of statistics is simply political misuse. Naturally all this analysis is based on one premise, namely, that the sample is extracted from an homogeneous population. In my opinion this is one of the largest failures in the public opinion polls. I believe it is possible to fraud a public opinion poll by carefully choosing the place and the time in which the interviews are made. Naturally there are protocols which have to be followed, but even so, I believe one can direct the answers depending on the will of the institute. Naturally in a civilized country, a public opinion institute which resorts to this kind of strategy would end up loosing credibility, but the one sided nature of the Brazilian Press will surely ensure that any misuse of the public opinion polls will be unpunished.

Multicausal failures

We all are aware of failures that are caused by a single event. From the point o view of engineering these are lucky cases, since controlling these events, the failure can be prevented. Engineering project usually assumes this hypothesis for this purpose. For example, the maximum stress level of a structure is calculated based on the yield or fracture stress of the individual components, some parts in airplanes are designed such that the cyclic stress intensity factor does not exceed the fatigue threshold level measured in a Paris plot.
There are, however, cases in which the failure is caused by multiple critical events happening in series or in parallel simmutaneously or not. The most famous example was the fire in the Kiss club in Santa Maria-RS, Brazil, last year. The causes ranged from corrupt city officials and firemen, who alloed the place to open without minimal safety conditions, greed by the owners, which led to bad material's selection of the foam used for acoustic insulation,which produced HCN when burnt, and the stupidity of the band members, who lit inappropriate fireworks in a closed space. Any of these events, if they were avoided, would prevent the tragedy too.This went surely through the mind of all people involved, but they surely decided that the probability of everything going wrong in the right sequence in the right time was too low to consider. The tragedy is there to prove they were wrong.
In fracture as Prof. Bažant teaches, this leads to two different failure propability distributions. In case of a single critical event (as in cleavage), the probability is described by the Weibull distribution. In the case the failure is a consequence of infinite individual critical events, like ductile fracture by microvoid coalescence, this leads to the gaussian distribution. There are also intermediate cases. The point here is to remind that multicausal failures do exist. They require nonlinear thinking by the engineer, who is forced to consider not only what could go wrong, but also in which sequence and in which time.
Worse, as I repeat to exhaustion to my students, when you decrese the probability of the unicausal failure, the multicausal failure becomes increasingly more probable.

Science or mysticism?

We are always criticising opinions and interpretations which lack the rigour of the scientific method, but how do we defend science to theordinary public? I remember reading in a book (don't remember which) the following critiscism: everyone of us believes in the first law of thermodynamics, because we were told it holds the most careful tests made until today, but only a handful scientists in the whole world are able to understand and interpret these tests. The majority of the population feels confortable with believing in science just because someone with a lab coat said it is science.

The h index

I was explaining to a.colleague how to calculate.the h index of our department, since he got surprized that an emmeritus professor of our department (who does not publish too much) appears three times in the list of the 34 most quoted papers. I explained that this is a problem of the h index (if II remember well, Hirch pointed to this in his paper): it gives too much value to old published papers, which, due to their long lives, received lots of citations. Then I got the idea of using a time-dependent h index, let us say, based on the list of papers published in a given time span (for example, the last three years). This will give a best estimate of the citation strength os a set of researchers. Of course, this time-dependent h index may be artificially inflated by the own researcher citing his own papers, thefore I suggest introducing a secondary index, which I call the citation strengrh, which is the number of citations (full) divided by the citations excluding self-citations. In my opinipn this index computes the probability that an article published by that author is cited by someone else. I would like to hear your opinion about this idea.

Brilliant minds

We are still mourning the passing away from our friend Larry Kaufman. I was thinking about him today, thinking about how lucky I was to have the opportunity to know him in person and thought about other friends, some who already left us, and some who are still with us. I decided to choose three of them, to share with you some of my recollections of the interaction with these extraordinary persons. This is no biography, I just want to share few memories which are special to me.

The first one, of course, must be Larry (Lawrence) Kaufman. I first heard his name as co-autor (with Berstein) of a book called CALPHAD, which we (still) have in our library. I first met him in Erice, at the 1996 Calphad conference in Italy, He knew I worked in the MPI Eisenforschung so he called me aside, to show a calculation of Al-Ti-Fe phase diagram using his database (which uses only crude models, regular solutions, ideal solutions, line compounds), showing he could reproduce a particularly tricky phase equilibriium involving the t2 phase, my colleagues had experimentally determined. This was characteristic from him. It does not matter that he coined the term CALPHAD, that in some sense, he was a kind o "father" to us all, he was always available and interacting, even with the humblest of the graduate students (myself). It in not everyday that you have the opportunity to "chat" with the person who created an entire scientific discipline (computational thermodynamics) and a whole industry (e.g. Thermo-calc, Pandat and so on).

The second extraordinary person is John Cahn (to my best knowledge, still alive and well). I met him at many occasions, but I remind in particular, my visit to NIST in 2001 on the way to the CALPHAD in Boston. I was showing some results on modelling crystal defects in intermetallics and said that I was doing that for my "Livre-docência" thesis (some Brazilian copy of the German habilitation). He told me he, in his entire life, had written only one thesis, his Ph. D. . The implication was clear. If he, himself, needed only a title, why should anyone else have other titles? I realized this and quickly answered that by doing this, I would get a 30% raise in my salary and he replied then, therefore, that I should do it, laughing. John Cahn is well known, even to physicist, due to the Hilliard-Cahn equation and some consider that he is the closest materials science will ever get to a Nobel prize winner.

The third special person was Ryioichi Kikuchi. One of the most privileged minds I ever met, I consider him to be of the level of Einstein, Bohr, Heisenberg and Feynman. I was working at the MPI Eisenforschung in Düsseldorf and met this tiny japanese guy walking in the corridor. I knew already he was coming for a three months for the fourth round of the Alexander von Humbolt prize grant, so I guessed right. At that time I remember him telling me his first recollection of Brazil was about watching one of his relatives departing on a ship to a distant country which was painted in green at the Mapa Mundi he had in his fundamental school. I made that time the decision to invite him to visit Brazil, so he could have opportunity to visit this lost branch of his family, what we did in year 2000. Sadly he passed away 10 years ago, but his memory will stay with me forever.

Results of the triennial evaluation by CAPES

Our graduate program on Metallurgical and Materials Engineering was evaluated by CAPES in the last triennial evaluation with grade 7, the highest possible and the first in the whole history of Escola Politécnica da USP. Congratulations to Professors +Jorge Tenório , +Denise Espinosa , Sérgio Duarte Brandi and Francisco Rolando Valenzuela Diaz, who led us to this unprecedented good result.

The world of Martian rocks: a cutting-edge system to materials characterization.

Launched in 2012 by NASA the Curiosity robot has revealed a lot of secrets from the red planet only studying Mars rocks and soil geology. What is the mechanism that allows human robots perform the Mars remote sensing of small rocks in a research to seek water, life and the own Mars's history?

The human robot is a real high-technology interplanetary laboratory. It is the state-of-the-art of a magnificent engineering achievement. It is equipped with a high-power Laser in an ingenious multipurpose device named by project’s scientists as ChemCam (Chemical Camera). The high-power Laser emits a short pulse with a big power which is focused on the surface of some rocks samples. The high energy density on the surface rock or in a sample of sand may induce supersonic plasma emission: so then, this light is collected by a spectrometer and with this information, scientists can identify the spectroscopic lines of the chemical elements in the rocks! Scientific literature refers to this technique as Laser-Induced Breakdown Spectroscopy (LIBS).

Fig. 1: ChemCam Artist's Conception by NASA.

The Curiosity robot has many devices developed in several countries including Russia, Netherlands, Germany and United States. Since 2012 a lot of discovers were reported with ChemCam and another devices. For example the first ionizing radiation measures in the red planet, a dry riverbed in a Martian crater and the first drilling in a Mars's rock in seek of water.

The amazing discover was, no doubt, performed by the ChemCam and its high-power remote sensing instrumentation. The robot reports that Mars sands are enriched with nutrients essential to life occurrence like Oxigen, Sulfur, Nitrogen, Phosphorus and Carbon. It also reports clay in Mars that indicates there were long periods of aqueous soil in the red planet. The great question is: Would be Mars an environment able to hold life? Better than this: Was Mars a planet with an ancient civilization?

Matheus Tunes, 23, is Graduate Student Candidate at Polytechnic School of the University of São Paulo  
Scientific Website

Fig. 2: A typical LIBS spectrum. In this case, we generate plasma in a deionized water sample. Tunes, MA; Schön, CG et al. INAC 2013 Proceedings, ENAN X121 E06, p. 1-6 (2013) ABEN-RJ.