Applied
Environmental
Statistics

Statistics, down to Earth

Course Outline
Our flagship course, taught over 4.5 days.

DAY 1
Describing Data in a Group
blankWhat's your best estimate?
blankHow noisy is it?
blankDealing with outliers
blankWhen to transform the scale

How Hypothesis Tests Work
blankTheir common denominators
blankTheir jargon explained
blankTips and tricks for computer and hand calculation
blank1-sided and 2-sided tests
blankExercise: testing means


Statistical intervals
blankCoping with uncertainty
blankCoping with skewed data
blankConfidence, prediction, tolerance intervals
blankExercise: the UCL95 and other intervals

Contingency Tables
blankDoes the frequency change between groups?
blankUse with censored data
blankExercise: uranium in drinking water


DAY 2
Comparing Two Groups of Data
blankAre means, medians different?
blankParametric and nonparametric tests
blankPaired data
blankHave standards been met?
blankThe quantile test
blankExercise: Precipitation chemistry

How many observations do I need?
blankWeaknesses of standard formulae
blankInteractions between variation, power, and dollars
blankSoftware available

Comparing Three or More Groups
blankone- and two-factor ANOVA
blanknon-parametric alternatives
blankmultiple comparison tests: who’s different?
blankExercise: Metals at six locations

Testing differences in Variability/Precision
blankCharacterizing differences in variability
blankLevene’s & Squared Ranks tests
blankPROBLEM: variability of concentrations

Correlation
blankLinear and monotonic correlation
blankr, rho and tau
blankKendall’s linear model
blankExercise: Three correlation coefficients


DAY 3
Linear Regression
blankBuilding a good regression model
blankdetermining improvements over background noise
blankhypothesis tests, confidence and prediction intervals
blankExercise: estimating total load

Multiple Regression
blankmeasures of a good model
blankplot the data !
blankmulti-collinearity
blankmodel selection methods better than stepwise
blankExercise: estimating urban non-point loads


DAY 4
Analysis of Covariance
blankDo two lines differ?
blankSeasonal changes
blankExercise: how many regression lines are needed?

Trend Analysis
blankselecting a trend test
blankregression vs. Mann-Kendall approaches
blankmonotonic vs. step trends
blankdealing with seasonality
blankdetecting regional trends
blankExample: Seasonal Kendall test for trend

FINAL EXAM



DAY 5
Logistic Regression
blankregression for categorical responses
blankapplication to nondetects, ratings, qualitative field methods
blankExercise: regression for atrazine concentrations with nondetects

Equivalence Tests
blankdifference from standard statistical tests
blanktesting for differences that are "big enough"
blankequivalence between groups, trends

Class Discussion and Applications