SEPP Science Editorials

Index of Editorials Misinformation Statistics Misuse	All Editorials for 2020 2013 2012 2011 2010 2009 2008 Categories Subcategories Antarctic Warming Skepticism [2] Book Review [3] Climate Change CO2 Emissions [1] Climate Models Uncertainty [2] Climate Science Climate Cycles [1] Climate Sensitivity [1] Holes [1] Thermal History [1] Unsolved Problems [1] Energy Issues American Power Act [1] Clean and Sustainable [1] Nuclear Waste Storage [1] Renewable Electricity Standard (RES) [1] Environmentalism Surrogate Religion [1] Foreword Energy Primer for Kids [1] Geo-Engineering Applications [2] Global Climate - International French Academy [1] Global Warming Anthropogenic Global Warming (AGW) [6] Confusion [1] Economics [1] General [2] Greenhouse Gases [1] Hockeystick [4] Ice Cores [1] Junkscience [9] Oceans' Role [2] Skepticism [1] Sun's Role [2] Health Issues Second Hand Smoke [1] Measurements Arctic Sea Ice [1] Atmospheric Temperature Data [2] Sea Surface Temperature [1] Surface Data [2] Misinformation Statistics Misuse [1] Modern Empirical Science v. Medieval Science [1] NIPCC China [1] Nuclear Fuel Supplies [1] Organizations Climate Research Unit (CRU) [1] International Panel on Climate Change (IPCC) [2] Nongovernmental International Panel on Climate Change (NIPCC) [1] UK Met Office [1] World Meteorological Organization (WMO) [1] Political Issues Climate Realism [1] Climategate [3] Independent Cross Check of Temperature Data [1] Report IPCC Assessment Report [2] NOAA State of the Climate 2009 [1] NRC-NAS Advancing the Science of Climate Change [1] Sea-Level Rise West Antarctic Ice Sheet (WAIS) [1] Alarmism [1] Types of Energy Nuclear Energy [1]
20-Jun-09 How to Cheat with Statistics
SEPP Science Editorial #18-2009 (in TWTW Jun 20, 2009) *S. Fred Singer, Chairman and President , Science and Environmental Policy Project (SEPP)* How to Cheat with Statistics Jun 20, 2009 The standard way is to simply ignore contrary data: for example, the IPCC-AR4 [2007] does not mention or reference climate forcing from changes in solar activity in spite of much published evidence. A more sophisticated method is selectivity: for example, choosing a time interval that will lead to a desired temperature trend [see SEPP Science Editorial #7-08 of Oct 4, 2008]. More difficult to spot is 'selective smoothing' of data that can produce a trend where none exists [see SEPP Science Editorial #8-09 2/28/09]. We now come to the misuse of averaging, as used in the WH report released this week. Recall that the last National Assessment report (NACC 2000, under Al Gore) used TWO climate models to predict dire futures. Trouble was, their results disagreed violently: in half of the 18 regions they even gave opposite predictions [see NIPCC Summary, figure 16]: For example, the Rio Grande region (New Mexico and West Texas), Upper and Lower Colorado would turn into a desert, acc to one model -while the other model turned them into swamps. So how to fix this strategic error? The new WH Assessment uses an AVERAGE of models instead of showing the results of individual models. It's the old story about the statistician who had one foot in a bucket of ice water and the other in a bucket of boiling water: on the average, he was quite comfortable. View The Week That Was in which this editorial appeared. Return to Top of Page