Copyright 2002 David M. Hassenzahl Using r and  2 Statistics for Risk Analysis.

Slides:



Advertisements
Similar presentations
Statistics Review – Part II Topics: – Hypothesis Testing – Paired Tests – Tests of variability 1.
Advertisements

Chapter 6 Sampling and Sampling Distributions
Session 8b Decision Models -- Prof. Juran.
Regression Analysis Once a linear relationship is defined, the independent variable can be used to forecast the dependent variable. Y ^ = bo + bX bo is.
Quantitative Skills 4: The Chi-Square Test
Simple Linear Regression and Correlation
Simple Linear Regression
Data Analysis Statistics. Inferential statistics.
1-1 Copyright © 2015, 2010, 2007 Pearson Education, Inc. Chapter 25, Slide 1 Chapter 25 Comparing Counts.
1 Difference Between the Means of Two Populations.
PSY 307 – Statistics for the Behavioral Sciences
Statistics for Business and Economics
Data Freshman Clinic II. Overview n Populations and Samples n Presentation n Tables and Figures n Central Tendency n Variability n Confidence Intervals.
Lecture Inference for a population mean when the stdev is unknown; one more example 12.3 Testing a population variance 12.4 Testing a population.
Statistics: Data Presentation & Analysis Fr Clinic I.
Intro to Statistics for the Behavioral Sciences PSYC 1900 Lecture 6: Correlation.
Chapter Topics Types of Regression Models
Hypothesis Tests for Means The context “Statistical significance” Hypothesis tests and confidence intervals The steps Hypothesis Test statistic Distribution.
Simple Linear Regression Analysis
Sample Size Determination In the Context of Hypothesis Testing
Data Analysis Statistics. Inferential statistics.
Simple Linear Regression Analysis
PSY 307 – Statistics for the Behavioral Sciences
Statistics for the Social Sciences Psychology 340 Fall 2013 Tuesday, November 19 Chi-Squared Test of Independence.
Probability Tables. Normal distribution table Standard normal table Unit normal table It gives values.
Copyright © Cengage Learning. All rights reserved. 13 Linear Correlation and Regression Analysis.
1 Tests with two+ groups We have examined tests of means for a single group, and for a difference if we have a matched sample (as in husbands and wives)
Fundamentals of Data Analysis Lecture 4 Testing of statistical hypotheses.
Statistical Analysis A Quick Overview. The Scientific Method Establishing a hypothesis (idea) Collecting evidence (often in the form of numerical data)
Statistics for Business and Economics Chapter 10 Simple Linear Regression.
Regression. Idea behind Regression Y X We have a scatter of points, and we want to find the line that best fits that scatter.
Statistical Power 1. First: Effect Size The size of the distance between two means in standardized units (not inferential). A measure of the impact of.
Data Analysis (continued). Analyzing the Results of Research Investigations Two basic ways of describing the results Two basic ways of describing the.
© 2001 Prentice-Hall, Inc. Statistics for Business and Economics Simple Linear Regression Chapter 10.
6.5 One and Two sample Inference for Proportions np>5; n(1-p)>5 n independent trials; X=# of successes p=probability of a success Estimate:
Introduction to Linear Regression
Copyright © Cengage Learning. All rights reserved. 10 Inferences Involving Two Populations.
Multiple regression - Inference for multiple regression - A case study IPS chapters 11.1 and 11.2 © 2006 W.H. Freeman and Company.
Production Planning and Control. A correlation is a relationship between two variables. The data can be represented by the ordered pairs (x, y) where.
Class 4 Simple Linear Regression. Regression Analysis Reality is thought to behave in a manner which may be simulated (predicted) to an acceptable degree.
From Theory to Practice: Inference about a Population Mean, Two Sample T Tests, Inference about a Population Proportion Chapters etc.
Correlation and Prediction Error The amount of prediction error is associated with the strength of the correlation between X and Y.
Research Process Parts of the research study Parts of the research study Aim: purpose of the study Aim: purpose of the study Target population: group whose.
Inference for Regression Simple Linear Regression IPS Chapter 10.1 © 2009 W.H. Freeman and Company.
Chapter 9 Fundamentals of Hypothesis Testing: One-Sample Tests.
You will be given a data set (on a computer) and a hypothesis. You will be asked the following questions (word for word): 1. How many degrees of freedom.
Statistical Hypotheses & Hypothesis Testing. Statistical Hypotheses There are two types of statistical hypotheses. Null Hypothesis The null hypothesis,
The Statistical Analysis of Data. Outline I. Types of Data A. Qualitative B. Quantitative C. Independent vs Dependent variables II. Descriptive Statistics.
Chi square analysis Just when you thought statistics was over!!
Physics 270 – Experimental Physics. Let say we are given a functional relationship between several measured variables Q(x, y, …) x ±  x and x ±  y What.
Chapter 8 Parameter Estimates and Hypothesis Testing.
Chapter 13 Inference for Counts: Chi-Square Tests © 2011 Pearson Education, Inc. 1 Business Statistics: A First Course.
Multiple Regression I 1 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. Chapter 4 Multiple Regression Analysis (Part 1) Terry Dielman.
Copyright © 2010 Pearson Education, Inc. Warm Up- Good Morning! If all the values of a data set are the same, all of the following must equal zero except.
Copyright © 2011 by The McGraw-Hill Companies, Inc. All rights reserved. McGraw-Hill/Irwin Simple Linear Regression Analysis Chapter 13.
Regression Analysis Deterministic model No chance of an error in calculating y for a given x Probabilistic model chance of an error First order linear.
Medical Statistics Medical Statistics Tao Yuchun Tao Yuchun 9
Chi Square Test for Goodness of Fit Determining if our sample fits the way it should be.
Chapter ?? 7 Statistical Issues in Research Planning and Evaluation C H A P T E R.
The Chi Square Equation Statistics in Biology. Background The chi square (χ 2 ) test is a statistical test to compare observed results with theoretical.
Fundamentals of Data Analysis Lecture 4 Testing of statistical hypotheses pt.1.
Chapter 10: The t Test For Two Independent Samples.
Tests of hypothesis Contents: Tests of significance for small samples
Chapter 14 Introduction to Multiple Regression
Hypothesis Testing Hypothesis testing is an inferential process
INF397C Introduction to Research in Information Studies Spring, Day 12
Statistical Quality Control, 7th Edition by Douglas C. Montgomery.
Central Limit Theorem, z-tests, & t-tests
Chi square.
Chi square.
Presentation transcript:

Copyright 2002 David M. Hassenzahl Using r and  2 Statistics for Risk Analysis

Copyright 2002 David M. Hassenzahl Objectives Purpose: to compare model to data –“validate model” (or not) Two techniques –r (correlation coefficient) –  2 (Chi-squared) Apply to a familiar problem (barium decay)

Copyright 2002 David M. Hassenzahl Statistics Descriptive Comparison –Z-scores, hypotheses –Confidence levels –Evaluating models –Correlation and Chi-squared

Copyright 2002 David M. Hassenzahl Confidence Levels Given 100 flips of a coin. Would you bet $1000 that the next flip will yield heads if –50 heads? –90 heads? –99 heads? –999 heads out of the last 1000 flips? How about for $5? For 50% of your current net worth?

Copyright 2002 David M. Hassenzahl Statistical Significance Z = –(sample occurrence – number in sample times expected probability –Divide by square root of (np(1-p) Student’s t One-sided versus two sided tests! “p values” Confidence intervals

Copyright 2002 David M. Hassenzahl Type I and II errors Type I: reject the truth! (accuracy) Type II: accept an untruth! (precision) This is important… there’s often a tradeoff here!

Copyright 2002 David M. Hassenzahl Z-scores Intuition Z score will be big if –Numerator: if xbar >>  OR  >> xbar –s is very small –n is very big Bigger Z-score: confidence that   xbar Small Z-score: confidence that   xbar

Copyright 2002 David M. Hassenzahl From Z’s to r’s and  2 r and  2 compare more than one estimate Compare –Set of model predictions to –Set of data or observations If r is SMALL (little correlation) the model doesn’t fit If  2 is SMALL then the model does fit

Copyright 2002 David M. Hassenzahl “Goodness of fit” We say that r and  2 evaluate “goodness of fit” Note that a good fit does not mean that the model is right!

Copyright 2002 David M. Hassenzahl Barium Decay Theory: barium is removed as a constant function of concentration “Exponential decay” C(T) = C(0)e kT –k = /min –C(0) = 0.16 mgBa / liter blood (From SWRI page 56 – 63; hypothetical)

Copyright 2002 David M. Hassenzahl Exponential Decay Model Figure 2-9 from Should We Risk It?

Copyright 2002 David M. Hassenzahl Sample blood at 1 hour intervals Time (hours)Measured Concentration

Copyright 2002 David M. Hassenzahl Measured and Expected Time (hours)Measured Concentration Predicted Concentration

Copyright 2002 David M. Hassenzahl Graphical Comparison After Figure 2-9 from Should We Risk It?

Copyright 2002 David M. Hassenzahl How well does the model fit? Why do we care? –Future predictions –Is there a better model? Looks OK. Is that good enough? Try our two tools: r and  2

Copyright 2002 David M. Hassenzahl r Conceptual Compares model predictions to the data Asks –“What if there is no relationship (or correlation) between model and data?” –Is the model as close to the average value of the x’s as it is to the actual x’s?

Copyright 2002 David M. Hassenzahl r terms or components Predicted mean and standard deviation Observed mean and standard deviation “Covariance” –Do they go up and down together? –If independent, covariance = 0 r = Covariance (predicted, observed) (STDEV O)  (STDEV P)

Copyright 2002 David M. Hassenzahl Means Observed x o bar = (  x oi ) /n x o bar = ( )/9 = Observed x p bar = (  x pi ) /n = 0.051

Copyright 2002 David M. Hassenzahl Standard Deviations

Copyright 2002 David M. Hassenzahl Covariance

Copyright 2002 David M. Hassenzahl Calculated r r = Covariance (predicted, observed) (STDEV O)  (STDEV P)

Copyright 2002 David M. Hassenzahl Intuition Behind r If there is no relationship between observed and predicted, r = 0 If r  0, positive correlation If r  0, negative correlation

Copyright 2002 David M. Hassenzahl r Discussed 0.99 seems reasonably good Is there a better fit What about theory? Limitations: even low correlations may be okay…just a screening tool

Copyright 2002 David M. Hassenzahl t test for r n-2 = 7 degrees of freedom Look it up in the Student’s-t table Accept model validity at 99% confidence level if Student’s t is greater than 2.998

Copyright 2002 David M. Hassenzahl Chi-squared This formula “normalizes” to the size of the individual x oi If all x oi  x ip,  2 = 0 Look up value in table (page 398)

Copyright 2002 David M. Hassenzahl Chi-squared 9 data points Suppose we are concerned with 99% confidence level We would need a chi-squared of greater than 21.7 to reject this line Calculating, we find that  2 = 0.06! Note that it still might be possible to find a better line, even with the exponential

Copyright 2002 David M. Hassenzahl Conclusion Both r and Chi-squared appear to validate this model Suggests that our theoretical idea about the model may be valid Doesn’t tell us we are right, just that we may be acceptably wrong!