JOEL BEST

JOEL BEST ~ BECOMING FAD-PROOF ~ FLAVOR OF THE MONTH ~ 2006

Posted by

0

~ DFWHLT-VEIAUMTLP ~ Don’t Forget What Happened Last Time – Viewing Each Innovation As Unique Misses The Larger Pattern

~ BSAAC ~ Be Skeptical About Astonishing Claims

~ WEDWHTTIWAW?-ITAETIMNW? ~ What Evidence Do We Have That This Innovation Will Actually Work? – Is There Any Evidence Suggesting That It Might Not Work?

~ DWHAEFWTCIETC-AFTVOTEB-A-WITBFTE? ~ Do We Have Any Estimates For What The Change Is Expected To Cost – And For The Value Of The Expected Benefits – And – What Is The Basis For Those Estimates?

~ TBOPOTBOTMTC ~ The Burden Of Proof Ought To Be On Those Making The Claims

~ CTIOPE-TLAIHBA-TMEOTBARIE ~ Continue To Insist On Persuasive Evidence – The Longer An Innovation Has Been Around – The More Evidence Ought To Be Available Regarding Its Effectiveness

~ AFAOCAEIOHDNCPSE ~ A Favorable Anecdote Or Comments About Early Indications Or Hopes Does Not Constitute Particularly Strong Evidence

~ DFOTFOBLB ~ Don’t Focus On The Fear Of Being Left Behind

~ EII-ATGTR-TMIHGEOTBIDWTTTR ~ Evidence Is Important – And The Greater The Risk – The More Important Having Good Evidence Ought To Be In Deciding Whether To Take That Risk

~ WTFTMOTPGTAMB-BWUDTC ~ When They Focus Too Much On The Potential Gains That Adoption Might Bring – People Wind Up Downplaying The Potential Costs

~ R:PRPTD-DATAONTAWWTTBAMQP-NWTPTM ~ Remember: People Rarely Proclaim Their Disappointments – Disillusionment And The Abandonment Of Novelties That Aren’t Working Well Tend To Be A Much Quieter Process – Nobody Wants To Publicize Their Mistakes

~ ISEWTTEIN-YNTBL ~ If Somebody Eventually Whispers That The Emperor Is Naked – You Need To Be Listening

JOEL BEST ~ COMMON SIGNS OF DUBIOUS DATA ~ STAT-SPOTTING ~ 2008

Posted by

0

~ B ~ Background

~ NSIWBF(BAFF) ~ Numbers Seem Inconsistent With Benchmark Figures (Basic And Familiar Facts)

~ SEAUTIASCP ~ Severe Examples Are Used To Illustrate A Supposedly Common Problem

~ B ~ Blunders

~ NTSTHOTLMBCBAMDP ~ Numbers That Seem Too High Or Too Low May Be Caused By A Misplaced Decimal Point

~ BTCSISBIL ~ Botched Translations Convert Statistics Into Simpler But Incorrect Language

~ MGDTRVIOD ~ Misleading Graphs Distort The Reader’s Visual Impression Of Data

~ EISOCATFF ~ Errors In Strings Of Calculations Affect The Final Figures

~ S ~ Sources

~ BRNMBASOG ~ Big Round Numbers May Be A Sign Of Guessing

~ H-‘TB’-O-‘TW’-MRE ~ Hyperbole – “The Biggest” – Or – “The Worst” – May Reveal Exaggeration

~ CTSUSMIBU ~ Claims That Seem Unbelievably Shocking May Indeed By Unbelievable

~ AP-IGADN-CTAC ~ A Problem – Is Given A Disturbing Name – Calculated To Arouse Concern

~ D ~ Definitions

~ BDLTBN ~ Broad Definitions Lead To Big Numbers

~ EAPDMTPSL ~ Expanding A Problem’s Definition Makes The Problem Seem Larger

~ CAPDDMOC ~ Changing A Problem’s Definition Distorts Measures Of Change

~ APDMELDC ~ A Problem’s Definition May Exclude Less Disturbing Cases

~ M ~ Measurements

~ NSITQ-HWTMC? ~ New Statistics Invite The Question – How Was This Measure Created?

~ UUOACLTQC ~ Unusual Units Of Analysis Can Lead To Questionable Conclusions

~ SMULQTEPR ~ Surveys May Use Loaded Questions That Encourage Particular Responses

~ CIMMATRS ~ Changes In Measurements May Affect The Resulting Statistics

~ CMOMMPDR ~ Competing Methods Of Measurement May Produce Different Results

~ P ~ Packaging

~ GMBBOABOMS ~ Generalizations May Be Based On A Biased Or Misleading Sample

~ TFACTEAPT ~ Time Frames Are Chosen To Emphasize A Particular Trend

~ AOBIUTCP ~ An Odd Base Is Used To Calculate Percentages

~ TNIASC(LATCMLTBA) ~ The Number Involves A Selective Comparison (Looking Only At Those Cases Most Likely To Be Affected)

~ ACRTSSMHBP ~ A Claim Reports That Some Statistical Milestone Has Been Passed

~ TW-‘A’-MRTET-‘M’-OT-‘M’ ~ The Word – “Average” – May Refer To Either The – “Mean” – Or The – “Median”

~ AEMBCBPRCATB ~ Apparent Epidemics May Be Caused By Problems Receiving Closer Attention Than Before

~ CIIAPOC ~ Correlation Is Implied As Proof Of Causation

~ DDMPI ~ Dramatic Discoveries May Prove Incorrect

~ D ~ Debates

~ REIDCOTP ~ Rival Explanations Identify Different Causes Of The Problem

~ OSDATNOE ~ Opposing Sides Disagree About The Nature Of Equality

~ ADPC ~ Advocates Debate Policy Choices