Biostatistics course Part 11 Comparison of two proportions Dr. Sc. Nicolas Padilla Raygoza Department of Nursing and Obstetrics Division of Health Sciences.

Slides:



Advertisements
Similar presentations
Chapter 12: Inference for Proportions BY: Lindsey Van Cleave.
Advertisements

Biostatistics course Part 13 Effect measures in 2 x 2 tables Dr. Sc. Nicolas Padilla Raygoza Department of Nursing and Obstetrics Division Health Sciences.
Biostatistics course Part 9 Comparison between two means Dr. Sc Nicolas Padilla Raygoza Department Nursing and Obstetrics Division Health Sciences and.
Biostatistics course Part 6 Normal distribution Dr. en C. Nicolas Padilla Raygoza Department of Nursing and Obstetrics Division of Health Sciences and.
Chapter 4 Inference About Process Quality
Populations & Samples Objectives:
Biostatistics course Part 2 Types of studies in epidemiology Dr. en C. Nicolas Padilla Raygoza Departrment of Nursing and Obstetrics Division of Health.
STA291 Statistical Methods Lecture 23. Difference of means, redux Default case: assume no natural connection between individual observations in the two.
“Students” t-test.
Confidence intervals for means and proportions FETP India
McGraw-Hill, Bluman, 7th ed., Chapter 9
Statistical Inferences Based on Two Samples
Biostatistics course Part 14 Analysis of binary paired data
Copyright (c) 2004 Brooks/Cole, a division of Thomson Learning, Inc. Chapter 9 Inferences Based on Two Samples.
Previous Lecture: Distributions. Introduction to Biostatistics and Bioinformatics Estimation I This Lecture By Judy Zhong Assistant Professor Division.
Research planning Dr. Nicolas Padilla Raygoza Department of Nursing and Obstetrics MCM María de Lourdes García Campos Department of Clinical Nursing Division.
Biostatistics course Part 4 Probability Dr. C. Nicolas Padilla Raygoza Department of Nursing and Obstetrics Division of Health Sciences and Engioneering.
1 Test a hypothesis about a mean Formulate hypothesis about mean, e.g., mean starting income for graduates from WSU is $25,000. Get random sample, say.
DATA ANALYSIS I MKT525. Plan of analysis What decision must be made? What are research objectives? What do you have to know to reach those objectives?
Sample size computations Petter Mostad
Basic Elements of Testing Hypothesis Dr. M. H. Rahbar Professor of Biostatistics Department of Epidemiology Director, Data Coordinating Center College.
T-Tests Lecture: Nov. 6, 2002.
5-3 Inference on the Means of Two Populations, Variances Unknown
Standard error of estimate & Confidence interval.
Two-Sample Proportions Inference. Sampling Distributions for the difference in proportions When tossing pennies, the probability of the coin landing on.
University of Guanajuato Campus Celaya Salvatierra Division of Health Sciences and Engineering Department of Nursing and Obstetrics Dr. Nicolas Padilla.
Biostatistics course Part 16 Lineal regression Dr. Sc. Nicolas Padilla Raygoza Department of Nursing and Obstetrics Division Health Sciences and Engineering.
Biostatistics course Part 15 Correlation Dr. Sc. Nicolas Padilla Raygoza Department of Nursing and Obstetrics Division Health Sciences and Engineering.
Comparing Two Proportions
Biostatistics course Part 8 Inferences of a mean Dr. Sc Nicolas Padilla Raygoza Department of Nursing and Obstetrics Division of Health Sciences and Engineering.
Go to index Two Sample Inference for Means Farrokh Alemi Ph.D Kashif Haqqi M.D.
Copyright (c) 2004 Brooks/Cole, a division of Thomson Learning, Inc. Chapter 9 Inferences Based on Two Samples.
Course on Biostatistics Part 1 What is statistics? Dr. Nicolas Padilla Raygoza Department of Nursing and Obstetrics Division of Health Sciences and Engineering.
Copyright © 2007 Pearson Education, Inc. Publishing as Pearson Addison-Wesley Slide
Biostatistics course Part 5 Binomial distribution
Inference for Regression Simple Linear Regression IPS Chapter 10.1 © 2009 W.H. Freeman and Company.
Biostatistics course Part 3 Data, summary and presentation Dr. en C. Nicolas Padilla Raygoza Department of Nursing and Obstetrics Division of Health Sciences.
Chapter 7 Sampling and Sampling Distributions ©. Simple Random Sample simple random sample Suppose that we want to select a sample of n objects from a.
Biostatistics course Part 12 Association between two categorical variables Dr. Sc. Nicolas Padilla Raygoza Department of Nursing and Obstetrics Division.
Copyright © 2010 Pearson Education, Inc. Chapter 22 Comparing Two Proportions.
Inference for 2 Proportions Mean and Standard Deviation.
Copyright © 2010, 2007, 2004 Pearson Education, Inc. Chapter 22 Comparing Two Proportions.
Testing Differences between Means, continued Statistics for Political Science Levin and Fox Chapter Seven.
- We have samples for each of two conditions. We provide an answer for “Are the two sample means significantly different from each other, or could both.
T Test for Two Independent Samples. t test for two independent samples Basic Assumptions Independent samples are not paired with other observations Null.
Biostatistics course Part 7 Introduction to inferential statistics Dr. Sc. Nicolas Padilla Raygoza Department of Nursing and Obstetrics, Division Health.
P-values and statistical inference Dr. Omar Aljadaan.
Biostatistic course Part 10 Inferences from a proportion Dr. Sc. Nicolas Padilla Raygoza Department dof Nursing and Obstetrics Division Health Sciences.
Sample Size Needed to Achieve High Confidence (Means)
Two-Sample Proportions Inference
Introduction For inference on the difference between the means of two populations, we need samples from both populations. The basic assumptions.
Two-Sample Proportions Inference
One-Sample Inference for Proportions
Two-Sample Proportions Inference
Chapter 4. Inference about Process Quality
STAT 312 Chapter 7 - Statistical Intervals Based on a Single Sample
Statistics in Applied Science and Technology
Biostatistics course Part 2 Types of studies in epidemiology
Comparing Two Proportions
Comparing Two Proportions
CHAPTER 6 Statistical Inference & Hypothesis Testing
Summary of Tests Confidence Limits
CHAPTER 12 Inference for Proportions
CHAPTER 12 Inference for Proportions
Two-Sample Proportions Inference
Chapter 24 Comparing Two Means.
Inference for Proportions
Statistical Inference for the Mean: t-test
Presentation transcript:

Biostatistics course Part 11 Comparison of two proportions Dr. Sc. Nicolas Padilla Raygoza Department of Nursing and Obstetrics Division of Health Sciences and Engineering Campus Celaya-Salvatierra Universidad de Guanajuato Mexico

Biosketch Medical Doctor by University Autonomous of Guadalajara. Pediatrician by the Mexican Council of Certification on Pediatrics. Postgraduate Diploma on Epidemiology, London School of Hygiene and Tropical Medicine, University of London. Master Sciences with aim in Epidemiology, Atlantic International University. Doctorate Sciences with aim in Epidemiology, Atlantic International University. Associated Professor B, School of Nursing and Obstetrics of Celaya, university of Guanajuato.

Competencies The reader will apply a Z test to obtain inferences from two independent proportions. He (she) will calculate confidence interval from two independent proportions.

Introduction Often, we make comparisons of two proportions from independent samples. In class we learned earlier to calculate confidence intervals and hypothesis test for a proportion; we can use the same methods to make inferences on proportions, if the sample size is large. For a large sample we can use a Normal approximation to the binomial distribution.

Examples In a study of urinary tract infection not complicated, patients were assigned to be treated with trimethoprim / sulfamethoxazole and fosfomycin / trometamol. 92 of 100 treated with fosfomycin / trometamol showed bacteriological cure while 61 of 100 treated with trimethoprim / sulfamethoxazole were cured infection.

Introduction When comparing proportions of independent samples, we must first calculate the difference in proportions. Analysis to compare two independent proportions is similar to that used for two independent means. We calculate a confidence interval and hypothesis test for difference in proportions.

Notation The notation we use for analysis of two proportions is the same as that for a proportion. The numbers below are for distinguishing the two groups. ParametersPopulation 1 2 Sample 1 2 Proportionπ1 π2p1 p2 Standard deviation √π1(1-π2) √π2(1-π2)√p1(1-p1) √p2(1-p2)

Inferences from two independent proportions The square of the standard error of a proportion is known as the variance of proportion. The variance of the difference between two independent proportions is equal to the sum of the variances of the proportions of each sample. The variances are summed because each sample contributes to sampling error in the distribution of differences.

Inferences from two independent proportions SE = √p(1-p)/n Variance = p(1-p)/n p1(1- p1) p2(1- p2) Variance (p1-p2) = variance of p1 + variance of p2 = n1 n2 The standard error of the difference between two proportions is given by the square root of the variances. SE (p1-p2) = √[p1(1-p1)/n1 + p2(1-p2)/n2]

Confidence intervals for two independent proportions To calculate the confidence interval we need to know the standard error of the difference between two proportions. The standard error of the difference between two proportions is the combination of the standard error of two independent distributions, ES (p1) and (p2). We estimated the magnitude of the difference of two proportions from the samples; now, calculate the confidence interval for this estimate.

The general formulae for confidence interval 95% is: Estimate ±1.96 x SE The formulae for IC 95% of two proportions should be: (p1-p2) ± 1.96 SE (p1-p2) Confidence intervals for two independent proportions

In the study of urinary tract infection, the proportion in the group of fosfomycin / trometamol was 0.92 and trimethoprim / sulfamethoxazole was 0.61 Difference in proportions = = 0.31 ES = √ [(0.92 (1-0.92) / (1-0.61) / 100] = 0056 The confidence interval at 95% would be: 0.31 ± 1.96 (0,056) = 0.31 ± 0.11 = 0.2 to 0.42 Confidence intervals for two independent proportions

The confidence interval at 95% would be: 0.31 ± 1.96 (0,056) = 0.31 ± 0.11 = 0.2 to 0.42 I have 95% confidence that the difference in the proportions in the population would be between 0.2 and As the difference does not include 0, we are confident that the proportion of the population treated with fosfomycin / trometamol is different than with trimethoprim sulfamethoxazole. Confidence intervals for two independent proportions

Hypothesis test for two independent proportions A hypothesis test uses the difference and standard error of difference. However, we use a slightly different standard error to calculate the hypothesis test. This is because we are assessing the probability that the observed data assume that the null hypothesis is true. The null hypothesis is that there is no difference in the proportions of both samples and both groups have a common π.

The best estimate we can get from π is the common proportion, p of the two proportions of the sample. P = r1 + n2 + r2/n1+n2 Where: r1 and r2 are numbers of positive responses in each sample n1 and n2 are the sample sizes in each sample. Common proportion will be between two individual proportions. Hypothesis test for two independent proportions

The standard error can be calculated by replacing p by p1 and p2. SE (p1-p2) =√p(1-p)(1/n1 +1/n2) This is known as a pooled standard error. Hypothesis test for two independent proportions

Example In the study of urinary tract infection, the proportion in the group of fosfomycin / trometamol was 0.92 and trimethoprim / sulfamethoxazole was integrants were in each group. Common p = / = 153/200 = SE (p1-p2) = √ 0.77 (1-0.77) (1 / / 100) = √ x = 0.019

Example Assuming a normal approximation to the binomial distribution, we calculate the Z test, as before. To calculate the hypothesis test, we must: 1.- Identify the null hypothesis Ho 2.- Identify the alternative hypothesis H1 3.- Calculate the hypothesis test Z.

Example Null hypothesis: when comparing two independent proportions of populations is usually the two proportions are equal. Ho: π1 = π2 It is as if the difference in the proportions of the two populations is 0. Ho: π1 - π2 = 0 Alternative hypothesis: is usually that the two proportions are not equal. H1: π1 ≠ π2 This is the same as the difference in proportions is not equal to zero. H1: π1 - π2 ≠ 0

Z statistic test The general formula for the Z test is the same as for the difference in two means. (p1-p2) – 0 z= SE (p1-p2) When the null hypothesis is that the difference in two proportions is zero estimate: (p1-p2) – 0 p1-p2 z= = SE (p1-p2) SE (p1-p2)

Example 0.92 success for fosfomycin / trometamol and 0.61 for trimethoprim / sulfamethoxazole SE = (p1-p2) – z= = = SE (p1-p2) P<0.05

Bibliografía 1.- Last JM. A dictionary of epidemiology. New York, 4ª ed. Oxford University Press, 2001: Kirkwood BR. Essentials of medical ststistics. Oxford, Blackwell Science, 1988: Altman DG. Practical statistics for medical research. Boca Ratón, Chapman & Hall/ CRC; 1991: 1-9.