P. 1
Consulting. Real Problems, Real Interactions, Real Outcomes

Consulting. Real Problems, Real Interactions, Real Outcomes

|Views: 0|Likes:
Published by edgardoking

More info:

Published by: edgardoking on Apr 13, 2012
Copyright:Attribution Non-commercial


Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less





Real Problems, Real Interactions, Real Outcomes Author(s): Richard Tweedie and Sue Taylor Source: Statistical Science

, Vol. 13, No. 1 (Feb., 1998), pp. 1-3 Published by: Institute of Mathematical Statistics Stable URL: http://www.jstor.org/stable/2676708 Accessed: 08/11/2010 21:38
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=ims. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Institute of Mathematical Statistics is collaborating with JSTOR to digitize, preserve and extend access to Statistical Science.


feltthat this also indiabout what most cated a seriousmisapprehension was thosewhoserolewas consulting about. of and in sion beyondmere advisingand coauthorship. 13.Statistical Science 1998. in Key wordsand phrases: Statisticalconsulting. Johnson organizersat Pullman) a (the conference and Directorof the CenterforApplied Statistical session was organized:the first double-barrelled he in Expertise there. 1. student. FlindersUniversity South of face. giving 1 .a statistician. clientinteractions.evenfor conratherthan full-time to be "new researchers" be (Of sultingstatisticians. I organizedthis session. THE PULLMANSESSIONS The papersin the Pullmanpanel discussion were presented invitation the IMS-WNAR meeting by at in held at Washington State University June 1996. she is nowa Ph.) morerealisticthan one might is Even so. consultants well as illustrative the manyfacetsofourprofesof DenUniversity ColoradoHealthSciencesCenter. Colorado Colorado State University. course. New Researchers the IMS. consulting bibliography. analyzing and presenting information the contained the data. in the beliefthat consulting valuable RichardTweedieis Chair. of nonacato academicand perhapsmoreparticularly Fort Collins.as one ofitsthemes. CNR mayhave been the wish in theiradvice. termsoftime or developing new techniques.showing diverseroleofthe statistician collecting..at least notin a real consulting situationwhere it is criticalto understanda that will interactions varietyof client-consultant consultafor influence the requirements effective views on the tion: see the Appendixfornumerous of role.. iftheyadoptthatrole.D.Department Statistics. organizedand participatedin a numberof such sessions at conferences both in the United States and in Australia. was a consultant CSIRO and problems halfwould describesome real consulting the private sectorin Australia (1974-1987). shouldask to be you on a co-author" the paperthatis assumedto be the I of end-product consulting. perhaps verypessimisticdepending or I In on one'sviewpoint. some commentsI had made on the role of consultants some In thatthe yearspreviously. 80523-1877 (e-mail: tweedie@stat. if in you have put in substantialeffort.an attempt made to drawtogether is someofthe lessonsofthese the in papers.. Tweedie(1992) I wrote comments consulting the Committee New on of on Researchers Committee of (CNR. ologyand medicalstatistics." This seemed to me to be very optimistic. 1-29 Consulting: Real Problems. not usually seen by clientsas a will notgo far Statisticians guruon a mountaintop.colostate. real complexity the consulting Moreover.and an activeconsultant epidemithe secondpartwouldbe a panel discussion. This timeI was asked to organizethe session in orderto atone for. the CNR article advises that ".of interestin theirown rightas Australia(1990-1995). unless you need the data analysis fiablycynical about the reasons for these stateis experience yourrole[as a consultant] to dispense ments:giventhe prevailing criteriaforpromotion in academia in particular.Colorado. The client should be responsiblefor the actual analysis. 1991) were "glib": in particular. Real Interactions. Real Outcomes Richard Tweedie (Resources Appendix by Sue Taylor) Abstract. Vol. with Deb Nolan and LuAnn demic statisticians.only perhaps rivalled by sessionson the gap between"academic"and "practical" statisticians.I diswith the Committee stateagreed ratherstrongly advice. Sue that mightillustratethe range of activitiesthat Taylor was Consultant.. consulting resources 1." this overview the case studiespresented of there.one might justiment that ". almostanyinteraction have especiallya with. I have.or sessions on how to teach undergraduate service courses in our notoriously uncharismatic subject.edu). everbeen involved is new researcher. over the years. ver.The Pullmanmeeting IMS-WNARhad. No.or perhaps amplify. of In "Statisticalconsulting.and it is clearlya perennialtopic for statisticalmeetings.

than the specifictechnicalinput." as somethe data. althoughthe bibliography readersto sourcesappendedwill enable interested on into the available literature this delve further topic and share at least one of the more tangible the bits ofadvicefrom panel. She illustrates way in whichcareful the of tigation the data sourcesis vital forany analysis of (and lack ofknowledge the data to be meaningful of or collection. of the writing a sole-authored Finally. where the Sue Taylor describes sortofrareproject the statistician actuallyinvolved is earlyenoughto data collection enhancethe and be able to influence As a consultant theAustralian to to ability analyze. thus saving a large amountof later workin analysis. studentfrom of the classic "on-campus" the best ley. and war storiesthatmanyofus tell or have manyother twothingsoverwhelmingly. rather it is oftenthe mere fact of such thinking.of university full working timeas consultants and ofstatisticians instruin academia.D.the statistician but not the problem. strong well.but is even here the consultant doingthe analysis. value to bothparties. However. and the paThis workedout surprisingly part ofthe sesthe are pers that follow from first and wit of experience the sion:regrettably. these case studies.all illustrate First. no matterwhat subject areas we enter. only after is of wholecontext the problem understood. interact the tionsbeingasked and consider impactofstatistics withinthe real contextof the problem.wearingan oldishhat as a privatesector I consultant. thatcontribution be truly its of mustenterintothe context best. StudyofAgeing. management searchproblems. This omissionreveals again the priorities even dutiespreclude his manyconsultants: ongoing paper.This is verymuchin themold in of the projectsenvisioned the CNR article. on problems a local and a tices and environmental none of global scale. with(we hoped) anecdotesand adviceand insights. theyshow that themwould be solvable.a Ph.he described majorenvironmental role lay in and one wheremuchofthe consultant's to what the real goals and contrying determine straintswere. in-housewith a government and mentality in a privatesectorcapacity. of and an outcome academic neededand developed.C. just as an "advisor. These examples coverbench realmost everyone. BatBob O'Brienfrom Missingis the paper from telle. BerkeKarl Broman. the secondpart are lost to all but thosewho were and WWW rethere. speakers. 3.illustrates (even a collaboration a interactions: good scientific some new techniques potential co-authorship!).in the end. that was not something statisticscan contribute to and we have much to offer there previously. audienceparticipation.that proves inhowpowerfully valuable.not just assumingthe clientwill carryit out. 2. sis to be inappropriate). CONCLUSIONS in will findsomething familiar Most statisticians these case studies. of at to for But second. our We know that understanding data and the questions of the client are of paramountimporto tance:it is reassuring see an examplewherethe statisticiancan controlthat process (Taylor).Moreover. U.We are knowthat problems rarelystandard.she was able to enLongitudinal sure(withmuchwork)thatanalysiswouldactually be carried out on reasonablyclean and reliable data.2 TWEEDIE ET AL. analyze the one preparedto understand withthosewho reallyown the quesdata.and that can come up with at its best statisticalthinking (Broman).we looked fora spectrumof statistical the the represent experibackgrounds: participants faculty ences of graduate students.or new ways to cope withnew problems we perhaps moretypically can see the role of our and decide how well the old ways fit assumptions (Tweedie). heard. It is hard to overestimate trainsus to thinkabout complicated our discipline issues in ways that allow us to quicklydiagnose to in difficulties esoteric disciplines whichwe have had only several minutesof introduction-a fact by reinforced these examplesand even further by the referees this collection. THE PARTICIPANTS In what followsyou will see fourarticlesfrom When inviting fourrather different perspectives.and an appropriateanalysis yields a solution-but yet and effective counterintuitive the data are revisitedand the again. From the-perspective an in-houseconsulof a problem. insight. Withoutstatisticalthinking.We can reveal far more know that close examination than the client originallytold us (Hoeting).a new researcher orado State. pracsocial surveys.The . a limitedclientunderstanding the can cause a well-planned analydata understudy. since therewere verymany stake- holdersin the problem:statisticalanalysis would provide the answers if only the questions were of clear. now at ColJennifer Hoeting. describea situationwheremodelling does work. tant. describes the role of the academic on as advisor-consultant a projstatistician funded invesect.

W.26 30-31.Sci. (1971). (1991). and ISENBERG. Statistical consultingin a university:dealing with people and other challenges. JOWELL. TheHuman Side ofStatistiCA. The J. Consulting. The papersbelowdo notpretend be to exhaustive: theyrepresent thosethat I have found ofmostuse in myprofessional practice. (1983).D. R.Statist. Six-state D. statistical.html . http://asa. J. (1987). Statist."and usefulif a year or twoolderthan desirable. 6 CHATFIELD.(1980). Biometrics 201-211. on mationor conduct discussions a timely basis. In Encyclopaedia of Statistical Sciences147-155. R. althoughtheyare by no means intendedto cover resourceavailable on the the growing information WorldWideWeb: Web VirtualLibrary:Statistics. Statistics stats. A search of the WorldWide Web as of October 1996 revealed a large numberof sites relevantto The following just a fewof these. and SCHUCANY. TheStatistical in Action.Statist. 240-268.html . Teaching Statistics Academic Press. A systematic study theconsulting as in literature an integral part of appliedtraining statistics. New York.37 5-20. *JOINER.execpc.Statist. Amer. International Review54 227-242.Statist.45 28-34. An "Essential Book List. analysis BOEN.html the A veryusefuland up-to-date document. (1982).uk/lists-k-o/minitab/ files/list-of-lists statisdetailsofall current contains This document intics-related e-mail lists. cal Consulting.35 72-76. International Statistical Institute Statistical declaration on professional ethics.Washington. (1981). The teaching interpersonal of in statistical Amer. C. D.html Resourceson the Web. statistical consulting. statistical Statist. Wiley. Bibliography 33 statistical consulting. B. A. Statisticson the Web.edu/chdocs/statistics/ stat_guide_home. and EVERITT. of *BASKERVILLE. 27 B. R. of conHYAMS. Those references markedwith an asteriskcontainan extensivelist of statisticalconsulting which are not references here.mailbase.Assoc.com/ statistics. A.umich.stat. psychology biostatistical sultation. Theyalso containpointers othermaterialcovto ering a wider range if needed. and WOLFE. "List StatisticsLists" is also available by sendingthe of one-line message sendminitab list-of-lists to mailbase@mailbase. (1981).ac. is of the HUNTER.Amer.wisc.37 1-4.stats.Amer. (1991).ac. How to consultwith a statistician.W. Lifetime Learning Publications. of RUSTAGI. Amer. aspectsof DC.Statist. Non-statistical D.lib. A Guide to StatisticalComputing the Internet. Cambridge Univ. Avoiding pitfalls. (1978). helberg/ http://www.CONSULTING 3 Pullman case studies show many of these attributes. . (1972). The World-Wide http://www. we have found them to be useful entrypoints. 9 Statistical Consultant 3-4. A. Proceedings theStatistical In of Education Section 67-72. J. London. J. R. . J. and ZAHN. G. BOEN. (1988). consulting.edu/vlib/statistics. The practice statistics: real world R. . Ethiof cal guidelines statistical for practice: report the Ad Hoc Committee Professional on Ethics. KIRK. RESOURCES APPENDIX ZAHN. also illustrate but the we vividly problems have in achieving such an idealisticstate.Statist. Press.edu/statistics/consult/ statbook. L. Amer.ufl. D.uk/cti/links http://www Resourceson . E. and Statistical Consulting. and FRYD. Biometrics 564-565. H. includingsubscription to These enable consultants share inforformation. (1977).ac. (1992). A. (1961). Electronic Resources Bibliography for Consultants is One ofthe mostusefultoolsforconsultants a of information together other by bibliography put consultants.35 121-123. A. transactional in statistical consulting. for *WOODWARD. S.stat. J. S. relationships BOEN.uk at or by pointing yourWeb browser http://www. C.ugl.Statist. duplicated ASA AD Hoc COMMITTEE PROFESSIONAL ON ETHICS (1983). SLOAN. (1982).32 58-60.gla. J.W. D. The practical L. Problem Solving: A Statistician's Chapmanand Hall. Guide. Amer. an idea whose time has come. C. R. Consultant HAND. J. http://www.html . for pracELLENBERG.. New York. CHATFIELD. Amer. (Chairman) (1986). J. Ethicalguidelines statistical tice: a historical prospective. are consultants.

You're Reading a Free Preview

/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->