O FAZ UM ESTATÍSTICO?

[DSFontes]

O FAZ UM ESTATÍSTICO?

Essa é uma pergunta que muitos fazem mesmo hoje, mesmo depois de mais de 40 anos que a profissão foi regulamentada.

Quando dizemos para os estudantes do segundo grau (ensino médio) que a profissão de ESTATÍSTICO está em alta, a primeira pergunta que eles fazem é justamente essa: O que faz um estatí­stico?

Pensando em ajudar o jovem aluno do ensino médio a entender um pouco do que fazemos, o CONRE-3 montou um pequeno flyer que costumamos distribuir aos que procuram uma carreira para cursar na universidade.

POR QUE ESCOLHER O BACHARELADO EM ESTATÍSTICA?

Que tal ser um profissional super versátil, que possa trabalhar em qualquer área e se dar sempre muito bem?

E que tal trabalhar com equipes diferentes, poder interagir com vários profissionais e ainda exercer uma função importante?

E, melhor ainda, não ter muito concorrente?

Gostou, não é? Pois este profissional existe e pode ser um Estatí­stico!

O Estatí­stico é aquele que se forma Bacharel em Estatí­stica. No Brasil há 29 universidades e/ou escolas que oferecem cursos de Bacharelado em Estatí­stica.

Mas, para que serve a ESTATÍSTICA, afinal de contas?

Imagine um médico e um farmacêutico querendo saber se um remédio em desenvolvimento é bom ou ruim. Para testar o remédio, é preciso PLANEJAR muito bem o experimento, COLETAR corretamente os dados, ANALISAR com muito cuidado e DIVULGAR seus resultados de forma honesta e com confiança no que está dizendo. Imagine o perigo de uma pesquisa mal feita num assunto tão importante! Bom, para não colocar a vida de ninguém em risco, é preciso tomar muitos cuidados. Antes de mais nada, é preciso planejar cada etapa abaixo:

O remédio será testado em quem?
-- Homens? Mulheres? Idosos? Crianças? Obesos? Jovens? Quem?

Quantas pessoas serão necessárias para testar?
-- Basta testar em uma ou duas pessoas? Ou será melhor testar em 10 pessoas? 30? 500? 2.000? Como saber?
-- Há dinheiro para testar em tanta gente?

E se houver dois grupos de pessoas?
-- Para um grupo de voluntários dá-se o remédio a ser testado; para o outro grupo, dá-se um remédio "de mentirinha", chamado placebo, mas não se conta a verdade para ninguém. Será que há diferença nos resultados de um grupo para outro?

-- Mas o remédio foi testado só com um grupo de pessoas, em geral voluntários, como é que depois pode-se afirmar que este remédio vai ser bom para todo mundo? É certeza absoluta?

O Estatí­stico é exatamente o profissional que auxiliará tanto o médico como o farmacêutico em cada uma destas etapas: desde o tipo de voluntário, quantidade e controle das pessoas que farão parte do experimento (amostragem), na coleta cuidadosa e minuciosa dos dados (campo), na organização destes dados no computador (banco de dados e tabulação), na hora de fazer todas as comparações interessantes, interpretar os resultados (testes estatí­sticos) e divulgá-los para todos os envolvidos (análises estatí­sticas). Como os testes são feitos somente num grupo de pessoas, existe uma pequena chance de haver um erro, não é mesmo? O Estatí­stico saberá dizer que tipo de erro poderá ocorrer e com que grau de certeza o resultado será divulgado.

A Estatí­stica é um conjunto de técnicas e métodos que vai ajudar o Estatí­stico em todas as etapas acima: na amostragem, na organização dos dados, na geração de tabelas e análises comparativas, na interpretação dos resultados, de forma que todas as afirmações possam ser feitas dentro de um limite de segurança estabelecido.

Mas a Estatí­stica não é usada só para ver se o remédio é bom ou não. Se você pensar bem, muita coisa do nosso dia-a-dia acontece em conseqüência de estudos que levam em conta análises estatí­sticas. Vejam alguns exemplos:

· Você abre o jornal e lê a manchete: "Cruzamentos: perigo à vista". A matéria mostra um gráfico sobre criminalidade na cidade e traz evidências de que num certo cruzamento houve muito mais assaltos do que noutros. Quantas pessoas evitarão este cruzamento ou passaraõ a ter atenção redobrada nestes locais?

· Um estudo científico mostra que mulheres fumantes têm probabilidade maior de desenvolver câncer do pulmão do que homens fumantes. Quantas mulheres não pararam de fumar diante desta notí­cia?

· A CET faz um estudo sobre o trânsito na cidade de São Paulo e decide se o rodí­zio será necessário ou não com base nas estatí­sticas sobre a quantidade e tipos de veí­culos diariamente nas ruas, locais mais congestionados, horários de pico, etc.

· A prefeitura da cidade reformula o sistema de transporte público com base nos relatórios estatí­sticos contendo informações detalhadas sobre fluxo de passageiros, linhas mais requisitadas, tempo de ociosidade, demanda vs. oferta, etc.

· O governo divulga dados estatí­sticos que influenciam todos os í­ndices financeiros que usamos no dia-a-dia: comércio, indústria, transporte, clientes, etc... que vão influenciar nas contas que vamos pagar (prestações, crediário, luz, água, telefone, gás, etc.)

Veja onde o estatí­stico pode atuar no tópico MERCADO DE TRABALHO PARA O ESTATÍSTICO.

[DSFontes]

Achei o post abaixo muito interessante, pois traz alguns pontos que deveriam sempre estar na mente do estatí­stico, antes de pensar no aspecto técnico do trabalho.

Abaixo, reproduzi também alguns comentários feitos por leitores do blog.

Think like a statistician " without the math
Posted by Nathan on Mar 4, 2010 to Data Design Tips, Statistics / 54 comments

*

I call myself a statistician, because, well, I'm a statistics graduate student. However, ask me specific questions about hypothesis tests or required sampling size, and my answer probably won't be very good.

The other day I was trying to think of the last time I did an actual hypothesis test or formal analysis. I couldn't remember. I actually had to dig up old course listings to figure out when it was. It was four years ago during my first year of graduate school. I did well in those courses, and I'm confident I could do that stuff with a quick refresher, but it's a no go off the cuff. It's just not something I do regularly.

Instead, the most important things I've learned are less formal, but have proven extremely useful when working/playing with data.
Here they are in no particular order.

Attention to Detail

Oftentimes it's the little things that end up being the most important. There was this one time in class when my professor put up a graph on the projector. It was a bunch of data points with a smooth fitted line. He asked what we saw. Well, there was an increase in the beginning, a leveling off in the middle, and then another increase. However, what I missed was the little blip in the curve in the first increase. That was what we were after.
The point is that trends and patterns are important, but so are outliers, missing data points, and inconsistencies.

See the Big Picture

With that said, it's important not to get too caught up with individual data points or a tiny section in a really big dataset. We saw this in the recent recovery graph. Like some pointed out, if we took a step back and looked at a larger time frame, the Obama/Bush contrast doesn't look so shocking.

No Agendas

This should go without saying, but approach data as objectively as possible. I'm not saying you shouldn't have a hunch about what you're looking for, but don't let your preconceived ideas influence the results. Because if you go to length looking for some specific pattern, you're probably going to find it. It'll just be at the sacrifice of accurate results.

Look Outside the Data

Context, context, context. Sometimes this will come in the form of metadata. Other times it'll come from more data.
The more you know about how the data was collected, where it came from, when it happened, and what was going on at the time, the more informative your results and the more confident you can be about your findings.

Ask Why

Finally, and this is the most important thing I've learned, always ask why. When you see a blip in a graph, you should wonder why it's there. If you find some correlation, you should think about whether or not it makes any sense. If it does make sense, then cool, but if not, dig deeper. Numbers are great, but you have to remember that when humans are involved, errors are always a possibility.

*Photo by misterbisson - Prof. Jiri Cisek (JiÅ™í ÄŒížek). Photographer Milan Kollinger took it at the Faculty of Applied Sciences University of West Bohemia in Pilsen in 1998.

Original post: http://flowingdata.com/2010/03/04/think ... -the-math/


ALGUNS COMENTÁRIOS

Sobre a necessidade do estatí­stico conseguir se explicar sem estatistiquês:

Pete
Good post, but these are really basic things that anyone doing any kind of quantitative analysis should regard as fundamental. These guidelines should be so familiar that they should not require any conscious thought. They are essential to a quantitative perspective of the world. I read them and thought, well, yeah, of course. I suppose it is worth reading as a reminder, or for beginners.

I agree with @jasprice about including "Explain Why". I would go one step further and add "Translate conclusions into easily understandable results." I do a lot of quantitative analysis. It is extremely important to be able to effectively communicate the importance of a key result to a person (or group) that does not have a quantitative background…especially if it is someone in a leadership position, i.e. your boss! March 4, 2010 at 9:40 am

Nathan
Absolutely. You have to remember though that are a lot of people who are doing this ad hoc. Even for me this was interesting, because I don’t typically think about, well, how I think :) March 4, 2010 at 11:18 am

Pete
I own a book that goes into some of this, ‘Turning Numbers Into Knowledge: Mastering the Art of Problem Solving’, by Jonathon Koomey. It addresses the issues you mention on a qualitative basis. It is useful and a quick read. March 5, 2010 at 11:34 am

Sobre a necessidade de se investir no conhecimento intrí­nseco ao assunto do projeto:

Cedar
I would add that these basic rules are actually not nearly as much help as you would expect without a working background knowledge of the topic being studied. Applying statistical techniques, without knowing background facts of the research design and the particular variables being measured only gets you so far. In fact, a few of the things you cite above, such as looking outside the data, and asking why (or digging deeper) are almost entirely dependent on background knowledge of that topic.
Sometimes some of this background knowledge is near universal, but that obscures the fact that we are using it (and can also lead to overconfidence in other situations which demand specialized background knowledge). I think education research is a classic example of this, where people think they have enough background knowledge to interpret the results, but in fact few do. March 4, 2010 at 10:25 am

Nathan
@Cedar " i agree with you completely, and that’s what i was trying to get at with looking outside the data. the context, or background, plays a huge role in the analysis and the results. March 4, 2010 at 11:21 am

Sobre a impressão que se tem por aí­ que basta fazer uma disciplina de Introdução à Estatí­stica para se considerar um Estatí­stico:

Joe The User
Sorry, I don’t this advice as helpful. I’m not statistician but I was least somewhat exposed to it in grad school.

You underestimate the difficulty someone with little maths understanding would have doing the operation "think about whether or not it makes any sense." If someone doesn’t know things like what is or isn’t a normal distribution and why it matters, they really can’t distinguish wishful thinking from evidence based reasoning. Look at the number of wrong statical arguments, by statisticians and mathematicians. These indeed come by not "digging deep" but this deep digging only happens when you know what the different correlations mean. Black-Scholes and the Guassian Copula are good examples of statistic gone wildly wrong with serious consequences. A moderate of understanding of distributions lets someone somewhat understand why people are questioning these now. But without the maths, a person "digging deeper" is going to be lost. March 4, 2010 at 2:35 pm

Nathan
@Joe the User " I’m not saying this stuff is easy to learn though. All I’m saying is that this is the most important stuff that I’ve learned over 4+ years in graduate school. So many people think that statistics is all hypothesis tests and similar formal tests " many who took intro stat in college and think that’s everything. All that stuff is secondary to the above. March 5, 2010 at 1:22 am

Sobre como enriquecer sua análise a fim de permitir outras leituras (releituras), e sobre a disponibilização de dados:

TrendWatcher
A few more principles to complement those in your post and in the comments.
1. Always ask "what’s missing?"
2. Create multiple views of the data. A single view is never enough for real situations
3. Craft a story, tying together the different views
4. Make the data available to others in a readily reusable format so others can work/play with it and possibly discover views you may have missed, thereby opening up a dialogue
5. When creating views, experiment with different granularity. One size does not fit all.
6. Write down your assumptions, biases, hypotheses BEFORE gathering, assembling, and analyzing the data. Keep a record of how often you are proved wrong. It better not be zero% or you are either fooling yourself or not tackling very serious issues. March 5, 2010 at 1:04 pm

[DSFontes]

Um dia na vida do Estatí­stico segundo a revista eletrônica PRINCETON REVIEW:

Statistician

A Day in the life of a Statistician

Statisticians collect data and analyze it, looking for patterns that explain behavior or describe the world as it is. A good statistician is involved in survey development and data collection from the beginning, ensuring the validity and usefulness of the data. Statisticians are employed by private and public concerns and apply their skills to specific industry issues, such as economic analysis, inventory control problems, health problems, and even television demographics. Statisticians must be familiar with valid scientific protocol and be able to quickly familiarize themselves with baselines and historical industry figures in order to structure an uncompromised analysis. Statisticians spend over half their day in front of a computer, setting up models, manipulating data, analyzing data, or writing reports. "You don’t just crunch numbers. You explain them," wrote one veteran statistician, who said that writing skills are important for those hoping to advance in the field. They spend the rest of their day in meetings, in planning sessions, or on the telephone exchanging ideas with colleagues. Respondents said, "Statistics is a visual science. You have to be able to picture the data and how it fits in with other known data shapes. It is not a Ônumbers only’ job." Most statisticians are applied statisticians, tailoring studies to real-life problems. It takes mathematical, visual, and practical skills to excel in this occupation as well as flexibility, curiosity, and a rigorous mind. Statisticians said the most difficult part of their job is explaining the implications of their studies to non-statisticians. Many said that statistics are best used as a starting point for investigation, not as a conclusion, and upper-level managers find this concept difficult to grasp. One wrote, "Statistics is a science of trends and probabilities, not certainties. We can tell what things happened, and suggest why they might have happened, but they’re only suggestions." Many statisticians cited the support of the statistician community as important to their satisfaction in the field. Statisticians feel challenged, involved and invigorated by their work, and this is evidenced by the small number (under 11 percent) of statisticians who leave the field each year.

Paying Your Dues

There are strict academic requirements for becoming a statistician. Entry-level positions require a Bachelor’s degree in Mathematics or Statistics. Those who wish to rise in the profession should consider obtaining a Master’s degree or a Ph.D. Just under 100 universities offer graduate degrees in statistics. Suggested coursework includes mathematics (calculus and linear algebra), probability, logic, psychology, and computer science. Candidates who combine statistical skills with another major that reflects their professional direction-such as economics and econometrics, computer and material science, or biology-have a distinct competitive advantage when seeking employment. Membership in professional organizations is not required, but many choose to join the ones affiliated with their occupation, such as organizations for economists or manufacturers.

Associated Careers

With their strong mathematical aptitude, many careers are open to statisticians. Wall Street is the most common employer of mathematical statisticians who leave their field, followed closely by the government, which regularly hires statisticians as "area analysts". A number become accountants, insurance analysts, and actuarial analysts when they want to improve their quality of life. A significant few enter computer science and become programmers and systems analysts.

LINK TO ORIGINAL: http://www.princetonreview.com/Careers.aspx?cid=149

[DSFontes]

Há uma definição bastante coerente das funções do ESTATÍSTICO no site da Organização Internacional do Trabalho (International Labour Organization - http://www.ilo.org/global/lang--en/index.htm):

http://www.ilo.org/public/english/burea ... 8/2122.htm

2122 STATISTICIANS



Statisticians conduct research, improve or develop mathematical and other aspects of statistical concepts, theories and operational methods and techniques, and advise on or engage in their practical application,in such fields as business or medicine as well as. in other areas of natural, social or life sciences.

Tasks include:

(a) studying, improving and developing statistical theories and methodologies-.

(b) planning and organising surveys and other statistical collections, and designing questionnaires;

(c) evaluating, processing, analysing, and interpreting statistical data and preparing them for publication;

(d) advising on or applying various data collection methods and statistical methods and techniques, and determining reliability of findings, especially in such fields as business or medicine as well as in other areas of natural, social or life sciences:.

(e) preparing scientific papers and reports-.

(f) performing related tasks;

(g) supervising other workers.

Examples of the occupations classified here:

Demographer

Statistician

Statistician, applied statistics

Statistician, mathematical


Some related occupations classified elsewhere:

Assistant, statistical: 3434

Clerk, statistical: 4122