Uniquely Gifted

Search Now:
Amazon Logo

 

Warshaw's Field Guide to Atypical Statistics
 

by Meredith G. Warshaw

 

Analysis of variants (ANOVA) - a test to examine which variants of your experiment give the best results.

Bar chart - a map of the locations of the nearest taverns.

Box plot - a cabal planning to hide surreptitiously all data contrary to the investigator's hypotheses in a box for burial.

Case-control study - a study where subjects with the condition of interest are paired with matched persons who are expected to control the subjects and make them behave as the investigator wishes.

Clinical trial - malpractice suit.

Correlation - parent, sibling, or child of a subject.

Data cleaning - eliminating dirty data that will make scatter plots scatter and p-values nonsignificant.

Data mining - what investigators do in search of statistical significance.

Discriminant analysis - a test to see whether the investigator has been discriminated against by the funding agency.

Exploratory data analysis - running 100 tests on 20 variables to find the one statistically significant result (see data mining).

F-distribution - the proportion of students flunking Intro to Stats.

Fisher's exact test - a statistical test showing the distribution of trout in the local pond.

Histograms - plots that cause sneezes and other allergic reactions in researchers.

Kolmagorov-Smirnoff test - a test developed by the famous Russian statistician and party animal, Igor Kolmagorov, to compare two distributions to see if they are the same. The test involves measuring the amount of vodka the investigator needs to drink until the two distributions are indistinguishable. Standardized tables by weight and sex are available in most statistics textbooks.

Logistical regression - a way of figuring out the logistics of getting your study accepted into JAMA.

Loglinear analysis - a method of analyzing categorical data developed by a statistician who was confused by log-log graph paper.

Martingale residuals - the droppings of a rare bird found primarily in biostatistics departments, used by oracles to divine whether a survival analysis is likely to find statistically significant results.

Nonlinear regression - what happens to the behavior of an investigator in the midst of preparing a grant submission.

Path analysis - study of the safest ways to get to and from the lab late at night.

P-value - results of urinalysis.

Principal proponent's analysis - figuring out what the principal investigator for the study wants to find. Alternatively, figuring out what the main supporter of the grant on the funding agency's review committee wants to hear.

SAS - a company whose major business is selling multitudes of large manuals to hapless statisticians and collecting exorbitant licensing fees.

Statistical significance - the holy grail of research; essential to publication, funding, and tenure.

Survival analysis - strategizing how to get the most money from the funding agency so that your research center doesn't fold.

T-test - blinded taste comparison of Orange Pekoe, Oolong, and Earl Grey.

Zzzz-distribution - the proportion of students staying awake in Intro to Stats.

Copyright 2001 


"Children require guidance and sympathy far more than instruction."
       ~ Anne Sullivan (Helen Keller's Teacher)

– Table of Contents –

Site copyright 2000-2005, Meredith G. Warshaw
Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page.  Copyrights for components of this work owned by others than Meredith Warshaw must be honored. Abstracting with credit is permitted.  To copy otherwise, to republish, to post on servers, or to redistribute to lists, requires prior specific permission.  Request permissions from Meredith Warshaw.
This policy is adapted with permission from ACM.