Jones RH | Bayesian information criterion for longitudinal and clustered data | Stat in Med | Nov 10 2011 | 3050 | Useful effective sample sizes for different correlation patterns including AR1 and unequal measurement times |

Begg MD, Vaughan RD | Are biostatistics students prepared to succeed in the era of interdisciplinary science? (And how will we know?) | Am Statistician | May 2011 | 71 | Excellent definitions of multi-, inter-, and trans-disciplinary and how this should affect training; in one experiment a multi-disciplinary approach did not benefit students early in their programs |

Kass, R | Statistical inference: The big picture | Statistical Sci | Feb 2011 | 1 | Bridges frequentist and Bayesian inference with an alternative "big picture" pragmatic depiction; excellent discussions by others |

Wingspan Productions | The Joy of Stats | Broadcast by BBC | | | Hans Rosling says there’s nothing boring about stats, and then goes on to prove it. A one-hour long documentary produced by Wingspan Productions and broadcast by BBC, 2010. |

Lehrer, J | http://www.newyorker.com/reporting/2010/12/13/101213fa_fact_lehrer?currentPage=1 The truth wears off | The New Yorker | Dec 13 2010 | 52 | Beautifully written article on the shrinkage of scientific findings over time; related to reproducible research and publication bias; see also http://www.scientificamerican.com/blog/post.cfm?id=in-praise-of-scientific-error-2010-12-20 this |

Nolan D, Temple Lang D | Computing in the statistics curricula | Am Statistician | 64:2 | 97 | A definitive article on what the statistical computing component of a graduate stat program should contain, and what goes wrong when only cursory attention is paid to computing |

Brown EN, Kass RE | What is statistics? | Am Statistician | 63:2 2009 | 105 | Defines statistical thinking and what happens when graduate programs do not incorporate it |

Divine G, Kapke A et al | Exemplary data set sample size calculation for Wilcoxon-Mann-Whitney tests | Stat in Med | Jan 15 2010 | 108 | Use of pilot data and noncentrality parameter estimates to scale up to sample size needed for definitive study |

Wickham H | A layered grammar of graphics | J Comp Graph Stat | March 2010 | 3 | Excellent article about foundations of statistical graphics and how to use ggplot2 including a concise specification of Napoleon's march to Moscow |

London AJ, Kimmelman J, Emborg ME | Beyond access vs. protection in trials of innovative therapies | Science | May 14 2010 | 829 | Harm to science of badly designed or badly motivated studies |

Tu Y-K, Law GR, Ellison GTH, Gilthorpe MS | Ratio index variables or ANCOVA? Fisher's cat revisited | Pharm Stat | 9;2010 | 77 | Shows contradictory findings resulting from use of ratios when measurement error or imperfect correlations are present |

Pocock SJ, Ware JH | Translating statistical findings into plain English | Lancet | Jun 6 2009 | 1926 | Concise reference on how to word results from P-values and confidence limits in medical papers |

Ransohoff D | How to improve reliability and efficiency of research about molecular markers: roles of phases, guidelines, and study design | J Clin Epi | Dec 2007 | 1205 | Study design is all important; principles are more important than phases and guidelines; strong case for making early studies reliable; excellent bibliography and list of examples of omics studies gone bad |

Young SS | Gaming the system: Chaos from multiple testing | IMS Bulletin | Dec 2007 | 13 | Editorial about the scientific crisis caused by poor statistical practice in epidemiology |

Sonnenburg et al | The Need for Open Source Software in Machine Learning | J Machine Learning Res | 2007 | 2443 | Excellent summary of benefits of open source and of various licenses |

Palmer CR, Shahumyan H | Implementing a decision-theoretic design in clinical trials:Why and how? | Stat in Med | 30 Nov 2007 | 4939 | Benefits of a decision-theoretic approach to clinical trials;excellent summary of ethical issues;first known clinical trial; data-dependent design; software |

Campbell I | Chi-squared and Fisher-Irwin tests of two-by-two tables with small sample recommendations | Stat in Med | 30 Aug 2007 | 3661 | Why not to use Fisher's test or the Yates correction; why to use N-1 instead of N in formula for Pearson chi-square. See DataAnalysisDisc |

Louis T | Our future as history | Biometrics | March 2007 | 1 | A nice article about the future and challenges of biostatistics |

Royston P, Altman D, Sauerbrei W | Dichotomizing continuous predictors in multiple regression: a bad idea | Stat in Med | 15 Jan 2006 | 127 | Covers all the terrible problems caused by dichotomization |

Wang A, Gehan EA | Gene selection for microarray data analysis using principal component analysis | Stat in Med | 15 Jul 2005 | 2069 | Data reduction method of potentially great use in gene microarray analysis and many other areas |

Morton S | Statistical collaboration to impact policy decisions | Stat in Med | 28 Feb 2005 | 493 | Definitions of consulting and collaboration, excellent case studies where a statistician made a fundamental impact |

Senn S | Added values controversies concerning randomization and additivity in clinical trials | Stat in Med | 30 Dec 2005 | 3729 | Interesting discussion on randomization and covariate-balanced randomization, blocking, covariate adjustment, choice of metrics, history |

Efron B | Statistics and the rules of science | Amstat News | July 2004 | 2 | Excellent summary of the scientific process |

Dufouil C, Brayne C, Clayton D | Analysis of longitudinal studies with death and drop-out: a case study | Stat in Med | Jul 2004 | 2215 | Good overview, discusses problems in targeting immortal subjects |

Hogan J, Roy J, Korkontzelou C | Handling drop-out in longitudinal studies | Stat in Med | May 2004 | 43 | Excellent tutorial on nonrandom dropout |

Mason RL | http://proquest.umi.com/pqdlink?index=12&did=000000624575701&SrchMode=3&sid=1&Fmt=6&VInst=PROD&VType=PQD&RQT=309&VName=PQD&TS=1085151174&clientId=2335 Does the statistics profession have an identity crisis? | JASA | Mar 2004 | 1 | Strong case for certification of statisticians |

Brumback BA, Hernan MA, Haneuse SJPA, Robins JM | Sensitivity analyses for unmeasured confounding assuming a marginal structural model for repeated measures | Stat in Med | Mar 2004 | 747 | |

Lin DY, Ying Z | Semiparametric regression analysis of longitudinal data with informative drop-outs | Biostat | 2003 | 385 | Perhaps the most useful approach for comparing 2 treatments when nonrandom dropouts occur |

Greenland S | When should epidemiologic regressions use random coefficients? | Biometrics | 2000 | 915 | See this and Biometrics 62:760;2006 |

Nick TG, Hardin JM | Regression Modeling Strategies: An Illustrative Case Study From Medical Rehabilitation Outcomes Research | Am J Occ Therapy | Sep 1998 | 459 | This is like a mini-course in modeling. Helpful for Masters level (or anyone) working on the Puzzlers. |

| Special Section: Opportunities and Challenges for the Discipline | American Statistician | Feb 2010 | | This special section discusses some important issues facing our profession. |

| | | | | |

| | | | | |