SAMPLE SIZE DETERMINATION IN HEALTH STUDIES A Practical Manual s. K. Lwanga Epidemiological and Statistical Methodology World Health Organization Geneva, Switzerland and S. Lemeshow Division of Public Health University of Massachusetts at Amherst MA USA World Health Organization Geneva 1991 WHO Library Cataloguing in Publication Data Lwanga, S K. Sample size determination in health studies' a practical manual 1 Sampling studies 2.Health surveys I. Lemeshow, S II.Title ISBN 92 4 154405 8 (NLM Classification WA 950) © World Health Organization 1991 Publications of the World Health Organization enjoy copyright protection in accordance with the provisions of Protocol 2 of the Universal Copyright Convention. For rights of reproduction or translation of WHO publications, in part or in toto, application should be made to the Office of Publications, World Health Organization, Geneva, Switzerland. The World Health Organization welcomes such applications. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the Secretariat of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities. or concerning the delimitation of its frontiers or boundaries. The mention of specific companies or of certain manufacturers' products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters. The authors alone are responsible for the views expressed in this publication. Typeset In India Printed In England 89/8087 -Macm ,llan/Clays-5000 Contents Introduction One-sample situations Estimating a population proportion with specified absolute preclslOn Estimating a population proportion with specified relative preClSlon Hypothesis tests for a population proportion Two-sample situations Estimating the difference between two population proportions with specified absolute precision Hypothesis tests for two population proportions Vll 1 2 3 6 6 7 Case-control studies Estimating an odds ratio with specified relative precision Hypothesis tests for an odds ratio 9 10 Cohort studies Estimating a relative risk with specified relative precision Hypothesis tests for a relative risk 12 12 13 Lot quality assurance sampling Accepting a population prevalence as not exceeding a specified value Decision rule for "rejecting a lot" 15 Incidence-rate studies Estimating an incidence rate with specified relative precision Hypothesis tests for an incidence rate Hypothesis tests for two incidence rates in follow-up (cohort) studies 17 17 17 Definitions of commonly used terms 21 Tables of minimum sample size 1. Estimating a population proportion with specified absolute preclslOn 2. Estimating a population proportion with specified relative preClSlon 3. Hypothesis tests for a population proportion 4. Estimating the difference between two population proportions with specified absolute precision 23 iii 9 15 15 18 25 27 29 33 Sample size determination 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. iv Hypothesis tests for two population proportions Estimating an odds ratio with specified relative precision Hypothesis tests for an odds ratio Estimating a relative risk with specified relative precision Hypothesis tests for a relative risk Acceptjng a population prevalence as not exceeding a specified value Decision rule for "rejecting a lot" Estimating an incidence rate with specified relative precision Hypothesis tests for an incidence rate Hypothesis tests for two incidence rates in follow-up (cohort) studies (study duration not fixed) 36 42 50 52 60 63 69 72 73 77 Preface In many of WHO's Member States, surveys are being undertaken to obtain information for planning, operating, monitoring and evaluating health services. Central to the planning of any such survey is the decision on how large a sample to select from the population under study, and it is to meet the needs of health workers and managers responsible for making that decision that this manual has been prepared. It is essentially a revised and expanded version of a popular unpublished document on sample size that has been widely used in WHO's field projects and training courses. The examples and tables presented, which have been selected to cover many of the approaches likely to be adopted in health studies, will not only be of immediate practical use to health workers but also provide insight into the statistical methodology of sample size determination. The authors would like to thank Dr B. Grab, formerly Statistician, WHO, Geneva, Dr R. J. Hayes, London School of Hygiene and Tropical Medicine, and colleagues in the Unit of Epidemiological and Statistical Methodology, the Diarrhoeal Diseases Control Programme and the Expanded Programme on Immunization of WHO for their comments. The financial support of the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases is gratefully acknowledged. v Introduction Among the questions that a health worker should ask when planning a surveyor study is "How large a sample do I need?" The answer will depend on the aims, nature and scope of the study and on the expected result, all of which should be carefully considered at the planning stage. For example, in a study of the curative effect of a drug on a fatal disease such as the acquired immunodeficiency syndrome (AIDS), where a single positive result could be important, sample size might be considered irrelevant. In contrast, if a new malaria vaccine is to be tested, the number of subjects studied will have to be sufficiently large to permit comparison of the vaccine's effects with those of existing preventive measures. The type of "outcome" under study should also be taken into account. There are three possible categories of outcome. The first is the simple case where two alternatives exist: yes/no, dead/alive, vaccinated/not vaccinated, existence of a health committee/lack of a health committee. The second category covers multiple, mutually exclusive alternatives such as religious beliefs or blood groups. For these two categories of outcome the data are generally expressed as percentages or rates. The third category covers continuous response variables such as weight, height, age and blood pressure, for which numerical measurements are usually made. In this case the data are summarized in the form of means and variances or their derivatives. The statistical methods appropriate for sample size determination will depend on which of these types of outcome the investigator is interested in. Only once a proposed study and its objectives have been clearly defined can a health worker decide how large a sample to select from the population in question. This manual is intended to be a practical guide to making such decisions. It presents a variety of situations in which sample size must be determined, including studies of population proportion, odds ratio, relative risk and incidence rate. 1 In each case the information needed is specified and at least one illustrative example is given. All but one example are accompanied by tables of minimum sample size for various study conditions so that the reader may obtain solutions to problems of sample size without recourse to calculations (more extensive tables are available in the publication by Lemeshow et al. mentioned below). Random sampling is assumed for all examples, so that if the sample is not to be selected in a statistically random manner the tables are not valid. 1 Continuous response variables are not considered in this manual because of the wide range of possible parameter values. vii Sample size determination The manual is designed to be used in "cookbook" fashion; it neither helps the reader to decide what type of study, confidence level or degree of precision is most appropriate, nor discusses the theoretical basis of sample size determination. Before using the manual, therefore, the investigator should have decided on the study design, made a reasonable guess at the likely result, determined what levels of significance, power and precision (where relevant) are required and considered operational constraints such as restrictions on time or resources. The reader who wishes to learn more about the statistical methodology of sample size determination is referred to Lemeshow, S. et aI., Adequacy of sample size in health studies (Chichester, John Wiley, 1990; published on behalf of the World Health Organization) or to any standard textbook on statistics. viii One-sample situations Estimating a population proportion with specified absolute precision Required information and notation (a) Anticipated population proportion (b) Confidence level (c) Absolute precision required on either side of the proportion (in percentage points) P 100(1-1X)% d A rough estimate of P will usually suffice. If it is not possible to estimate P, a figure of 0.5 should be used (as in Example 2); this is the "safest" choice for the population proportion since the sample size required is largest when P = 0.5. If the anticipated proportion is given as a range, the value closest to 0.5 should be used. Tables 1a and 1b (pages 25-26) present mInImUm sample SIzes for confidence levels of 95% and 90%, respectively. Simple random sampling is unlikely to be the sampling method of choice in an actual field survey. If another sampling method is used, a larger sample size is likely to be needed because of the "design effect". For example, for a cluster sampling strategy the design effect might be estimated as 2. This would mean that, to obtain the same precision, twice as many individuals would have to be studied as with the simple random sampling strategy. In Example 2, for instance, a sample size of 192 would be needed. Example 1 Solution A local health department wishes to estimate the prevalence of tuberculosis among children under five years of age in its locality. How many children should be included in the sample so that the prevalence may be estimated to within 5 percentage points of the true value with 95% confidence, if it is known that the true rate is unlikely to exceed 20%? (a) Anticipated population proportion (b) Confidence level (c) Absolute precision (15%-25%) 20% 95% 5 percentage points Table 1a (page 25)shows that for P = 0.20 and d = 0.05 a sample size of 246 would be needed. If it is impractical, with respect to time and money, to study 246 children, the investigators should lower their requirements of confidence to, per- Sample size determination haps, 90%. Table 1b (page 26) shows that, in this case, the required sample size would be reduced to 173. Example 2 Solution An investigator working for a national programme of immunization seeks to estimate the proportion of children in the country who are receiving appropriate childhood vaccinations. How many children must be studied if the resulting estimate is to fall within 10 percentage points of the true proportion with 95% confidence? (It is not possible to make any assumption regarding the vaccination coverage.) (a) Anticipated population proportion ("safest" choice, since P is unknown) (b) Confidence level (c) Absolute precision (40%-60%) 50% 95% 10 percentage points Table 1a (page 25) shows that for P=0.50 and d=O.1O a sample size of96 would be required. Estimating a population proportion with specified relative precision Required information and notation (a) Anticipated population proportion (b) Confidence level (c) Relative precision P 100(1-IX)% E The choice of P for the sample size computation should be as "conservative" (small) as possible, since the smaller P is the greater is the minimum sample size. Tables 2a and 2b (pages 27-28) present mmImum sample SIzes for confidence levels of 95% and 90%, respectively. Example 3 Solution An investigator working for a national programme of immunization seeks to estimate the proportion of children in the country who are receiving appropriate childhood vaccinations. How many children must be studied if the resulting estimate is to fall within 10% (not 10 percentage points) of the true proportion with 95% confidence? (The vaccination coverage is not expected to be below 50%.) (a) Anticipated population proportion (conservative choice) (b) Confidence level (c) Relative precision (from 45% to 55%) 50% 95% 10% (of 50%) Table 2a (page 27) shows that for P = 0.50 and E = 0.10 a sample size of 384 would be needed. If it is impractical, with respect to time and money, to study 384 children, the investigators should lower their requirements of confidence to, per2 One-sample situations haps, 90%. Table 2b (page 28) shows that, in this case, the required sample size would be reduced to 271. Simple random sampling is unlikely to be the sampling method of choice in an actual field survey. If another sampling method is used, a larger sample size is likely to be needed because of the "design effect". For example, for a cluster sampling strategy the design effect might be estimated as 2. This would mean that, to obtain the same precision, twice as many individuals would have to be studied as with the simple random sampling strategy. In this example, therefore, for a confidence level of 95%, a sample size of 768 would be needed. Example 4 Solution How large a sample would be required to estimate the proportion of pregnant women in a population who seek prenatal care within the first trimester of pregnancy, to within 5% of the true value with 95% confidence? It is estimated that the proportion of women seeking such care will be between 25% and 40%. (a) Anticipated population proportion (b) Confidence level (c) Relative precision 25%-40% 95% 5% (of 25%-40%) Table 2a (page 27) presents the following sample sizes for population proportions in the range 25%-40%. P Sample size 0.25 0.30 0.35 0.40 4610 3585 2854 2305 E= 0.05 and for Therefore a study of roughly 4610 women might be planned to satisfy the stated objectives. If necessary a smaller sample size could be used, but this would result in a loss of precision or confidence or both if the true value of P was close to 25%. Hypothesis tests for a population proportion This section applies to studies designed to test the hypothesis that the proportion of individuals in a population possessing a given characteristic is equal to a particular value. Required information and notation (a) Test value of the population proportion under the null hypothesis (b) Anticipated value of the population proportion (c) Level of significance (d) Power of the test (e) Alternative hypothesis: either or Po Pa 1000(% 100(1- fJ)% Pa>P o or Pa<P o (for one-sided test) Pai=P o (for two-sided test) 3 Sample size determination Tables 3a-d (pages 29-32) present minimum sample sizes for a level of significance of 5%, powers of90% and 80%, and both one-sided and twosided tests. For Tables 3c and 3d the complement of Po should be used as the column value whenever Po> 0.5. Example 5 Solution The five-year cure rate for a particular cancer (the proportion of patients free of cancer five years after treatment) is reported in the literature to be 50%. An investigator wishes to test the hypothesis that this cure rate applies in a certain local health district. What minimum sample size would be needed if the investigator was interested in rejecting the null hypothesis only if the true rate was less than 50%, and wanted to be 90% sure of detecting a true rate of 40% at the 5% level of significance? (a) Test cure rate (b) Anticipated cure rate ( c ) Level of significance ( d) Power of the test (e) Alternative hypothesis (one-sided test) 50% 40% 5% 90% cure rate < 50% Table 3a (page 29) shows that for Po = 0.50 and P a = 0.40 a sample size of 211 would be needed. Example 6 Solution Previous surveys have demonstrated that the usual prevalence of dental caries among schoolchildren in a particular community is about 25%. How many children should be included in a new survey designed to test for a decrease in the prevalence of dental caries, if it is desired to be 90% sure of detecting a rate of 20% at the 5% level of significance? (a) Test caries rate (b) Anticipated caries rate ( c ) Level of significance ( d) Power of the test (e) Alternative hypothesis (one-sided test) 25% 20% 5% 90% caries rate < 25% Table 3a (page 29) shows that for Po = 0.25 and P a = 0.20 a sample size of 601 would be needed. If the investigators use this sample size, and if the actual caries rate is less than 20%, then the power of the test will be larger than 90%, i.e. they will be more than 90% likely to detect that rate. Example 7 Solution The success rate for a surgical treatment of a particular heart condition is widely reported in the literature to be 70%. A new medical treatment has been proposed that is alleged to offer equivalent treatment success. A hospital without the necessary facilities or staff to provide the surgical treatment has decided to use the new medical treatment for all new patients presenting with this condition. How many patients must be studied to test the hypothesis that the success rate of the new method of treatment is 70% against an alternative hypothesis that it is not 70% at the 5% level of significance? The investigators wish to have a 90% power of detecting a difference between the success rates of 10 percentage points or more in either direction. (a) Test success rate (b) Anticipated success rate 4 70% 80% or 60% One-sample situations ( c ) Level of significance ( d) Power of the test ( e ) Alternative hypothesis (two-sided test) 5% 90% success rate #- 70% Table 3c (page 31) shows that for (1- Po) = 0.30 and iFa - Pol = 0.1 0 a sample size of 233 would be needed. Example 8 Solution In a particular province the proportion of pregnant women provided with prenatal care in the first trimester of pregnancy is estimated to be 40% by the provincial department of health. Health officials in another province are interested in comparing their success at providing prenatal care with these figures. How many women should be sampled to test the hypothesis that the coverage rate in the second province is 40% against the alternative that it is not 40%? The investigators wish to be 90% confident of detecting a difference of 5 percentage points or more in either direction at the 5% level of significance. (a) Test coverage rate (b) Anticipated coverage rate ( c ) Level of significance ( d) Power of the test (e) Alternative hypothesis (two-sided test) 40% 35% or 45% 5% 90% coverage rate #- 40% Table 3c (page 31) shows that for Po =0.40 and iFa-Pol=0.05 a sample size of 1022 would be needed. 5 Two-sample situations Estimating the difference between two population proportions with specified absolute precision Required information and notation (a) Anticipated population proportions ( b) Confidence level (c) Absolute precision required on either side PI and P z 100(1-a)% of the true value of the difference between the proportions (in percentage points) d (d) Intermediate value V=P I (1-Pd+P 2 (1-P z ) For any value of d, the sample size required will be largest when both PI and P z are equal to 50%; therefore if it is not possible to estimate either population proportion, the "safest" choice of 0.5 should be used in both cases. The value of V may be obtained directly from Table 4a (page 33) from the column corresponding to P z (or its complement) and the row corresponding to PI (or its complement). Tables 4b and 4c (pages 34-35) present mInImUm sample SIzes for confidence levels of 95% and 90%, respectively. Example 9 Solution What size sample should be selected from each of two groups of people to estimate a risk difference to within 5 percentage points of the true difference with 95 % confidence, when no reasonable estimate of PI and P z can be made? (a) Anticipated population proportions ("safest " choice) ( b ) Confidence level ( c ) Absolute precision (d) Intermediate value 50%,50% 95% 5 percentage points 0.50 Table 4b (page 34) shows that for d = 0.05 and V = 0.50 a sample size of 769 would be needed in each group. Example 10 In a pilot study of 50 agricultural workers in an irrigation project, it was observed that 40% had active schistosomiasis. A similar pilot study of 50 agricultural workers not employed on the irrigation project demonstrated that 32% had active schistosomiasis. If an epidemiologist would like to carry out a larger study to estimate the schistosomiasis risk difference to 6 Two-sample situations within 5 percentage points of the true value with 95% confidence, how many people must be studied in each of the two groups? Solution (a) ( b) ( c) ( d) Anticipated population proportions Confidence level Absolute precision In termedia te value 40%,32% 95% 5 percentage points 0.46 Table 4b (page 34) shows that for d = 0.05 and V = 0.46 a sample size of 707 would be needed in each group. Hypothesis tests for two population proportions This section applies to studies designed to test the hypothesis that two population proportions are equal. For studies concerned with very small proportions, see Example 13. Required information and notation (a) Test value of the difference between the population proportions under the null hypothesis (b) Anticipated values of the population proportions PI and P 2 ( c) Level of significance 1000:% ( d) Power of the test 100(1- /3)% ( e) Alternative hypothesis: either P 1 - P 2 > 0 or P 1 - P 2 < 0 (for one-sided test) or P 1 -P 2 ¥0 (for two-sided test) Tables 5a-h (pages 36-41) present minimum sample sizes for a level of significance of 5 %, powers of 90% and 80%, both one-sided and two-sided tests, and the special case of very small proportions. Tables 5e-h should be used whenever the proportion under consideration is less than 5%.1 Example 11 It is believed that the proportion of patients who develop complications after undergoing one type of surgery is 5% while the proportion of patients who develop complications after a second type of surgery is 15%. How large should the sample size be in each of the two groups of patients if an investigator wishes to detect, with a power of 90%, whether the second procedure has a complication rate significantly higher than the first at the 5% level of significance? Solution 0% (a) Test value of difference in complication rates 5%,15% (b) Anticipated complication rates ( c ) Level of significance 5% ( d ) Power of the test 90% (e) Alternative hypothesis risk difference (P 1 - P 2) < 0% (one-sided test) 1 For further discussion of small proportions, see Lemeshow, S. et ai., Adequacy of sample size in health studies (Chichester, John Wiley, 1990; published on behalf of the World Health Organization). 7 Sample size determination Table 5a (page 36) shows that for PI = 0.05 and P 2 = 0.15 a sample size of 153 would be needed in each group. Example 12 Solution In a pilot survey in a developing country, an epidemiologist compared a sample of 50 adults suffering from a certain neurological disease with a sample of 50 comparable control subjects who were free of the disease. Thirty of the subjects with the disease (60%) and 25 of the controls (50%) were involved in fishing-related occupations. If the proportion of people involved in fishing-related occupations in the entire population is similar to that observed in the pilot survey, how many subjects should be included in a larger study in each of the two groups if the epidemiologist wishes to be 90% confident of detecting a true difference between the groups at the 5% level of significance? (a) Test value of difference between proportions involved in fishing-related occupations 0% (b) Anticipated proportions involved in fishing-related occupations ( c ) Level of significance ( d ) Power of the test (e) Alternative hypothesis (two-sided test) 60%,50% 5% 90% risk difference # 0% The required sample size is obtained from Table 5c (page 38) from the column corresponding to the smallest of PI' P 2 and their complements and the row corresponding to IP 2 - P 11. In this case, for (1- PI) = 0.40 and I P 2 - P 11 = 0.10, the required sample size would be 519 in each group. Example 13 Solution Two communities are to participate in a study to evaluate a new screening programme for early identification of a particular type of cancer. In one community the screening programme will include all adults over the age of 35, whereas in the second community the procedure will not be used at all. The annual incidence of the type of cancer under study is 50 per 100000 ( = 0.0005) in an unscreened popUlation. A drop in the rate to 20 per 100000 (= 0.0002) would justify using the procedure on a widespread basis. How many adults should be included in the study in each of the two communities if the investigators wish to have an 80% probability of detecting a drop in the incidence of this magnitude at the 5% level of significance? (a) Test difference in cancer rates ( b ) Anticipated cancer rates ( c ) Level of significance ( d) Power of the test (e) Alternative hypothesis (one-sided test) 0% 0.05 %, 0.02 % 5% 80% risk difference (P 1 - P 2) > 0% Table 5f (page 40) shows that for PI = 0.0005 and P 2 = 0.0002 a sample size of 45 770 would be needed in each group. 8 Case-control studies Examples 14 and 15 concern the odds ratio, which is the ratio ofthe odds of occurrence of an event in one set of circumstances to the odds of its occurrence in another. For example, if the "event" is a disease, people with and without the disease may be classified with respect to exposure to a given variable: Disease No disease Exposed Unexposed a c b d The odds ratio is then ad/be. Estimating an odds ratio with specified relative precision Required information and notation (a) Two of the following should be known: • Anticipated probability of "exposure" for people with the disease [a/(a + b)] • Anticipated probability of "exposure" for people without the disease [c/(c + d)] • Anticipated odds ratio (b) Confidence level (c) Relative precision Pi Pi OR 100(1 - Ct)% [; When the number of people in the population who are affected by the disease is small relative to the number of people unaffected: e~(a+c) and d ~(b + d). In this case, therefore, the probability of "exposure" given "no disease" (P i) is approximated by the overall exposure rate. Tables 6a-h (pages 42-49) present minimum sample sizes for confidence levels of 95% and 90% and relative precisions of 10%, 20%, 25% and 50%. F or determining sample size from Table 6 when 0 R ~ 1, the values of both Pi and OR are needed. Either of these may be calculated, if necessary, 9 Sample size determination provided that Pi is known: OR = [Pf/(l- Pi)]/[Pi/(1 - P!)] P! = Pf/[OR(l- Pi) + PiJ. and If OR < 1, the values of Pi and 1/0R should be used instead. Example 14 Solution In a defined area where cholera is posing a serious public health problem, about 30% of the population are believed to be using water from contaminated sources. A case-control study of the association between cholera and exposure to contaminated water is to be undertaken in the area to estimate the odds ratio to within 25% of the true value, which is believed to be approximately 2, with 95% confidence. What sample sizes would be needed in the cholera and control groups? Anticipated probability of "exposure" given "disease" Anticipated probability of "exposure" given "no disease" (approximated by overall exposure rate) Anticipated odds ratio ( b) Confidence level (c) Relative precision (a) Table 6c (page 44) shows that for OR would be needed in each group. = 2 and P! = ? 30% 2 95% 25% 0.3 a sample size of 408 Hypothesis tests for an odds ratio This section outlines how to determine the minimum sample size for testing the hypothesis that the population odds ratio is equal to one. Required information and notation (a) Test value of the odds ratio under the null hypothesis (b) Two of the following should be known: • Anticipated probability of "exposure" for people with the disease [al(a + b)] • Anticipated probability of "exposure" for people without the disease [cl(c + d)] • Anticipated odds ratio ( c) Level of significance ( d) Power of the test (e) Alternative hypothesis (for two-sided test) ORo=1 P*1 P*2 ORa 100a% 100(1- fJ)% Tables 7a and 7b (pages 50-51) present minimum sample sizes for a level of significance of 5% and powers of 90% and 80% in two-sided tests. For determining sample size from Table 7 when ORa> 1, the values of both P! and OR are needed. Either of these may be calculated, if necessary, 10 Case--control studies provided that Pi is known: ORa = [Pi/(I - Pi)]/[P!I(I- Pi)] Pi = Pi/[ORa(1- Pi) and + PiJ. If ORa < I, the values of Pi and IIOR a should be used instead. Example 15 The efficacy of BeG vaccine in preventing childhood tuberculosis is in doubt and a study is designed to compare the vaccination coverage rates in a group of people with tuberculosis and a group of controls. Available information indicates that roughly 30% of the controls are not vaccinated. The investigators wish to have an 80% chance of detecting an odds ratio significantly different from 1 at the 5 % level. If an odds ratio of 2 would be considered an important difference between the two groups, how large a sample should be included in each study group? Solution 1 (a) Test value of the odds ratio ? (b) Anticipated probability of "exposure" given "disease" Anticipated probability of "exposure" given "no disease" 30% 2 Anticipated odds ratio (c) Level of significance 5% 80% (d) Power of the test odds ratio # 1 (e) Alternative hypothesis Table 7b (page 51) shows that for OR 130 would be needed in each group. 11 = 2 and Pi = 0.30 a sample size of Cohort studies Estimating a relative risk with specified relative precision Required information and notation (a) Two of the following should be known: • Anticipated probability of disease in people exposed to the factor of interest • Anticipated probability of disease in people not exposed to the factor of interest • Anticipated relative risk (h) Confidence level (c) Relative precision P2 RR 100(I-IX)% B Tables 8a-h (pages 52-59) present minimum sample sizes for confidence levels of95% and 90%, and levels of precision of 10%, 20%, 25% and 50%. For determining sample size from Table 8 when RR ~ 1, the values of both P 2 and RR are needed. Either of these may be calculated, if necessary, provided that P I is known: and If RR< 1, the values of PI and I/RR should be used instead. Example 16 Solution An epidemiologist is planning a study to investigate the possibility that a certain lung disease is linked with exposure to a recently identified air pollutant. What sample size would be needed in each of two groups, exposed and not exposed, if the epidemiologist wishes to estimate the relative risk to within 50% of the true value (which is believed to be approximately 2) with 95% confidence? The disease is present in 20% of people who are not exposed to the air pollutant. (a) Anticipated probability of disease given "exposure" Anticipated probability of disease given "no exposure" Anticipated relative risk (b) Confidence level (c) Relative precision ? 20% 2 95% 50% Table 8d (page 55) shows that for RR = 2 and P 2 = 0.20 a sample size of 44 would be needed in each group. 12 Cohort studies Hypothesis tests for a relative risk This section outlines how to determine the minimum sample size for testing the hypothesis that the population relative risk is equal to one. Required information and notation (p) Test value of the relative risk under the null hypothesis (h) Two of the following should be known: RRo = 1 • Anticipated probability of disease in people exposed PI to the variable • Anticipated probability of disease in people not exposed to the variable Pz • Anticipated relative risk RRa ( c) Level of significance 100()(% (d) Power of the test 100(1-rn% ( e) Alternative hypothesis (for two-sided test) Tables 9a~c (pages 60~62) present minimum sample sizes for a level of significance of 5% and powers of 90%, 80% and 50% in two-sided tests. For determining sample size from Table 9 when RRa > 1, the values of both P z and RRa are needed. Either of these may be calculated, if necessary, provided that P I is known: RRa= PI/P z and If RRa < 1, the values of PI and l/RRa should be used instead. Example 17 Two competing therapies for a particular cancer are to be evaluated by a cohort study in a multicentre clinical trial. Patients will be randomized to either treatment A or treatment B and will be followed for 5 years after treatment for recurrence of the disease. Treatment A is a new therapy that will be widely used if it can be demonstrated that it halves the risk of recurrence in the first 5 years after treatment (i.e. RRa = 0.5); 35 % recurrence is currently observed in patients who have received treatment B. How many patients should be studied in each of the two treatment groups if the investigators wish to be 90% confident of correctly rejecting the null hypothesis (RRo = 1), if it is false, and the test is to be performed at the 5 % level of significance? Solution 1 ? 35% Anticipated probability of recurrence given treatment B 0.5 Anticipated relative risk 5% ( c ) Level of Significance 90% ( d) Power of the test rela ti ve risk =I- 1 (e) Alternative hypothesis (a) Test value of the relative risk (b) Anticipated probability of recurrence given treatment A 13 Sample size determination Table 9a (page 60) shows that for RRa = 0.5 (1/ RRa = 2) and P 2 = 0.35 (P 1 = 0.175) a sample size of 135 would be needed in each group (figure obtained by interpolation; the exact sample size is 131 by computation). 14 Lot quality assurance sampling Accepting a population prevalence as not exceeding a specified value This section outlines how to determine the minimum sample size that should be selected from a given population so that, if a particular characteristic is found in no more than a specified number of sampled individuals, the prevalence of the characteristic in the population can be accepted as not exceeding a certain value. Required information and notation (a) Anticipated population prevalence P (b) Population size (c) Maximum number of sampled individuals showing N characteristic ( d) Confidence level d* 100(1-C()% Tables lOa-j (pages 63-68) present minimum sample sizes for confidence levels of 95% and 90% and values of d* of 0-4. Example 18 Solution In a school of 2500 children, how many children should be examined so that if no more than two are found to have malaria parasitaemia it can be concluded, with 95% confidence, that the malaria prevalence in the school is no more than 10%? (a) Anticipated population prevalence ( h) Population size (c) Maximum number of malaria cases in the sample ( d) Confidence level 10% 2500 2 95% Table 10c (page 64) shows that for P = 0.1 0 and N = 2500 a sample size of 61 children would be needed. Decision rule for "rejecting a lot" This section applies to studies designed to test whether a "lot" ( a sampled population) meets a specified standard. The null hypothesis is that the proportion of individuals in the population with a particular characteristic is equal to a given value, and a one-sided test is set up such that the lot is accepted as meeting the specified standard only if the null hypothesis can 15 Sample size determination be rejected. For this purpose a "threshold value" of individuals with the characteristic (d*) is computed as a basis for a decision rule; if the number of sampled individuals found to possess the characteristic does not exceed the threshold, the null hypothesis is rejected (and the lot is accepted), whereas if the threshold is exceeded, the lot is rejected. Required information and notation (a) Test value of the population proportion under the null hypothesis (b) Anticipated value of the population proportion ( c) Level of significance ( d) Power of the test Po Pa 1001):% 100(1-(1)% Tables lla-c (pages 69-71) present minimum sample sizes for a level of significance of 5% and powers of 90%, 80% and 50% in one-sided tests. Example 19 Solution In a large city, the local health authority aims at achieving a vaccination coverage of 90% of all eligible children. In response to concern about outbreaks of certain childhood diseases in particular parts of the city, a team of investigators from the health authority is planning a survey to identify areas where vaccination coverage is 50% or less so that appropriate action may be taken. How many children should be studied, as a minimum, in each area and what threshold value should be used if the study is to test the hypothesis that the proportion of children not vaccinated is 50% or more, at the 5% level of significance? The investigators wish to be 90% sure of recognizing areas where the target vaccination coverage has been achieved (i.e. where only 10% of children have not been fully vaccinated). (a) Test value of the population proportion (b) Anticipated value of the population proportion ( c ) Level of significance ( d) Power of the test 50% 10% 5% 90% Because the mistake of accepting groups of children as adequately vaccinated, when in fact the coverage is 50% or less, is the more important, Po=0.50 and Pa=O.lO. Table lla (page 69) shows that in this case a sample size of 10 and a threshold value of 2 should be used. Therefore, a sample of 10 children should be taken from each of the areas under study. If more than 2 children in a sample are found not to have been adequately vaccinated, the lot (the sampled population) should be "rejected", and the health authority may take steps to improve vaccination coverage in that particular area. If, however, only 2 (or fewer) children are found to be inadequately vaccinated, the null hypothesis should be rejected and the group of children may be accepted as not being of immediate priority for an intensified vaccination campaign. 16 Incidence-rate studies Estimating an incidence rate with specified relative precision Required information and notation (a) Relative precision (b) Confidence level f: 100(1- et)% Table 12 (page 72) presents minimum sample sizes for confidence levels of 99%,95% and·90%. Example 20 Solution How large a sample of patients should be followed up if an investigator wishes to estimate the incidence rate of a disease to within 10% of its true value with 95% confidence? (a) Relative precision (b) Confidence level 10% 95% Table 12 shows that for [; = 0.10 and a confidence level of 95% a sample size of 385 would be needed. Hypothesis tests for an incidence rate This section applies to studies designed to test the hypothesis that the incidence rate of a characteristic is equal to a particular value. Required information and notation (a) Test value of the popUlation incidence rate under the null hypothesis (b) Anticipated value of the population incidence rate (c) Level of significance (d) Power of the test Ao Aa 100et% 100(1- fJ)% (e) Alternative hypothesis: either or Aa > Ao or Aa < Ao (for one-sided test) Aa #- Ao (for two-sided test) Tables 13a-d (pages 73-76) present minimum sample sizes for a level of significance of 5 %, powers of 90% and 80% and both one-sided and twosided tests. 17 Sample size determination Example 21 Solution On the basis of a 5-year follow-up study of a small number of people, the annual incidence rate of a particular disease is reported to be 40%. What minimum sample size would be needed to test the hypothesis that the population incidence rate is 40% at the 5% level of significance? It is desired that the test should have a power of90% of detecting a true annual incidence rate of 50% and the investigators are interested in rejecting the null hypothesis only if the true rate is greater than 40%. (a) Test value of the incidence rate (b) Anticipated incidence rate ( c ) Level of significance ( d) Power of the test (e) Alternative hypothesis (one-sided test) Table 13a (page 73) shows that for ..1.0 sample size of 169 would be needed. = 40% 50% 5% 90% incidence rate> 40% 0.40 and Aa = 0.50 a minimum Hypothesis tests for two incidence rates in followup (cohort) studies This section applies to studies designed to test the hypothesis that the true incidence rates of a disorder or characteristic in two groups of individuals are equal. Subjects either have a common date of entry into the study and are followed up until they develop the characteristic in question or cannot be followed up any more (Example 22), or are inducted into the study as they become available but are followed up only until a specified date (Example 23). Required information and notation (a) Test value of the difference between the popu- lation incidence rates under the null hypothesis . 1. 1 -..1. 2=0 Anticipated values of the incidence rates Al and ..1.2 Level of significance 100a% Power of the test 100(1-[3)% Alternative hypothesis: either )'1 - )'2 > 0 or Al - ..1.2 < 0 (for one-sided test) or ..1.1-..1.2 #0 (for two-sided test) (f) Duration of study (if fixed) T (b) ( c) ( d) ( e) If the study is terminated at a fixed point in time, before all the subjects have necessarily experienced the end-point of interest, the observations are said to be censored. The values of A then have to be modified according to the formula as in Example 23. Tables 14a-d (pages 77-80) present minimum sample sizes for a level of significance of 5%, powers of 90% and 80% and both one-sided and two18 Incidence-rate studies sided tests, when the duration of the study is not fixed and the two groups studied are of equal size. No tables are given for studies of fixed duration because too many parameters are involved to permit easy tabulation. Example 22 Solution As part of a study of the long-term effect of noise on workers in a particularly noisy industry, it is planned to follow up a cohort of people who were recruited into the industry during a given period of time and to compare them with a similar cohort of individuals working in a much quieter industry. Subjects will be followed up for the rest of their lives or until their hearing is impaired. The results of a previous small-scale survey suggest that the annual incidence rate of hearing impairment in the noisy industry may be as much as 25%. How many people should be followed up in each of the groups (which are to be of equal size) to test the hypothesis that the incidence rates for hearing impairment in the two groups are the same, at the 5% level of significance and with a power of 80%? The alternative hypothesis is that the annual incidence rate for hearing impairment in the quieter industry is not more than the national average of about 10% (for people in the same age range), whereas in the noisy industry it differs from this. (a) Test value of the difference in incidence rates (b) Anticipated incidence rates ( c ) Level of significance ( d) Power of the test ( e ) Alternative hypothesis (two-sided test) (f) Duration of study o 25% and 10% 5% 80% Al =P )'2 not applicable Table 14d (page 80) shows that for Al = 0.25 and A2 = 0.10 a sample size of 23 would be required in each group. Example 23 Solution A study similar to that outlined in Example 22 is to be undertaken, but the duration of the study will be limited to 5 years. How many subjects should be followed up in each group? o (a) Test value of the difference in incidence rates (b) Anticipated incidence rates 25% and 10% 5% 80% Al =P A2 5 years ( c) Level of significance ( d) Power of the test (e) Alternative hypothesis (two-sided test) (f) Duration of study The values of ), must be modified according to the formula for f(A) given on page 18: f(I =0.175) = 0.0918 where X = (AI f(AI =0.25) = 0.1456 f(A2 =0.10) = 0.0469. + A2 )/2 The appropriate sample size formula is where k is the ratio of the sample size for the second group of subjects (n 2 ) to that for the first group (nd (in this example k = 1). 19 Sample size determination Thus n1 = {1.96)[2(0.0918)J + 0.842)(0.1456 + 0.0469)}2/(0.25 - 0.10)2 = 1.462/0.023 = 65.0. A sample size of 65 would therefore be needed for each group. For a one-sided test the corresponding sample size formula is 20 Definitions of commonly used terms The brief definitions listed here are intended to serve only as reminders for the reader. Fuller explanations of statistical terms and a discussion of the statistical theory relevant to sample size determination are to be found in Lemeshow, S. et al., Adequacy of sample size in health studies (Chichester, John Wiley, 1990; published on behalf of the World Health Organization). a The significance level of a test the probability of rejecting the null hypothesis when it is true (or the probability of making a Type I error). p The probability of failing to reject the null hypothesis when it is false (or the probability of making a Type II error). Case--control studies Studies in which subjects are selected on the basis of their status with respect to a given characteristic (such as the presence of a disease); the "cases" show the characteristic and the "controls" do not. Both groups are studied with respect to their prior and current exposure to suspected risk factors. Cluster sampling A sampling process in which sampling units are made up of clusters or groups of study units. Cohort studies Studies in which subjects are selected with respect to the presence and absence of a characteristic (such as exposure to a given factor) suspected of being associated with the particular outcome of interest (for example a disease). Both groups of subjects are followed up for development of the outcome. Confidence level The probability that an estimate of a population parameter is within certain specified limits of the true value; commonly denoted by "I-a". Design effect In cluster sampling, the design effect is an indication of the variation due to clustering. It is estimated by the ratio of the variance when cluster sampling is used to the variance when simple random sampling is used. Incidence rate The number of specific events (for example new cases of a disease) occurring in a specified population per unit time. Lot quality assurance sampling Sampling techniques, with industrial origins, designed to ascertain whether batches of items meet specified standards. Null hypothesis A statement concerning the value of a population parameter. It is the hypothesis under test in a test of significance, for example the hypothesis that an observed difference is entirely due to sampling errOr. 21 Sample size determination Odds ratio The ratio of the odds of occurrence of an event in one set of circumstances to the odds of its occurrence in another (see also page 9). One-sided test In hypothesis testing, when the difference being tested is directionally specified beforehand (for example when Xl < X 2, but not Xl> X 2, is being tested against the null hypothesis Xl = X 2). Population proportion The proportion of individuals in a population possessing a given characteristic. Power of a test The probability of correctly rejecting the null hypothesis when it is false; commonly denoted by "1 - {J". Precision A measure of how close an estimate is to the true value of a population parameter. It may be expressed in absolute terms or relative to the estimate. Prevalence The number of cases of a disease (or people with a particular characteristic) existing in a specified population at a given point in time. Relative risk The ratio of the risk (probability) of an outcome (for example disease or death) among people exposed to a given factor to the risk among people not exposed. Significance level See definition of iY.. Simple random sampling Sampling procedure in which every study unit has the same chance of being selected and every sample of the same size has the same chance of being chosen. Study units The individual members of a population whose characteristics are to be measured. Two-sided test In hypothesis testing, when the difference being tested for significance is not directionally specified beforehand (for example when the test takes no account of whether Xl> X 2 or Xl < X 2). Represent the number of standard errors from the mean; Zl-" and Zl-,,/2 are functions of the confidence level and Z 1 _ p is a function of the power of the test. 22 Tables of minimum sample size Table 1. Estimating a population proportion with specified absolute precision n = zi -a/2P(1- P)/d 2 (a) Confidence level 95% p 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.11 0.12 0.13 0.14 0.15 0.20 0.25 1825 456 203 114 73 51 37 29 23 18 15 13 11 9 8 5 3457 864 384 216 138 96 71 54 43 35 29 24 20 18 15 9 6 4898 1225 544 306 196 136 100 77 60 49 40 34 29 25 22 12 8 6147 1537 683 384 246 171 125 96 76 61 51 43 36 31 27 15 10 7203 1801 800 450 288 200 147 113 89 72 60 50 43 37 32 18 12 8067 2017 896 504 323 224 165 126 100 81 67 56 48 41 36 20 13 8740 2185 971 546 350 243 178 137 108 87 72 61 52 45 39 22 14 9220 2305 1024 576 369 256 188 144 114 92 76 64 55 47 41 23 15 9508 2377 1056 594 380 264 194 149 117 95 79 66 56 49 42 24 15 9604 2401 1067 600 384 267 196 150 119 96 79 67 57 49 43 24 15 9508 2377 1056 594 380 264 194 149 117 95 79 66 56 49 42 24 15 9220 2305 1024 576 369 256 188 144 114 92 76 64 55 47 41 23 15 8740 2185 971 546 350 243 178 137 108 87 72 61 52 45 39 22 14 8067 2017 896 504 323 224 165 126 100 81 67 56 48 41 36 20 7203 1801 800 450 288 200 147 113 89 72 60 50 43 37 32 18 12 6147 1537 683 384 246 171 125 96 76 61 51 43 36 31 27 15 10 4898 1225 544 306 196 136 100 77 60 49 40 34 29 25 22 12 8 3457 864 384 216 138 96 71 54 43 35 29 24 20 18 15 9 6 1825 456 203 114 73 51 37 29 23 18 15 13 11 9 8 5 d ~I 'Sample size less than 5. 13 I~ Table 1 ( continued) (b) Confidence level 90% p 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.11 0.12 0.13 0.14 0.15 0.20 0.25 1285 321 143 80 51 36 26 20 16 13 11 9 8 7 6 2435 609 271 152 97 68 50 38 30 24 20 17 14 12 11 6 3450 863 383 216 138 96 70 4330 1082 481 271 173 120 88 68 53 43 36 30 26 22 19 11 7 5074 1268 564 317 203 141 104 79 63 51 42 35 30 26 23 13 8 5683 1421 631 355 227 158 116 89 70 57 47 39 34 29 25 14 9 6156 1539 684 385 246 171 126 96 76 62 51 43 36 31 27 15 10 6494 1624 722 406 260 180 133 101 80 65 54 45 38 33 29 16 10 6697 1674 744 419 268 186 137 105 83 67 55 47 40 34 30 17 11 6765 1691 752 423 271 188 138 106 84 68 56 47 40 35 30 17 11 6697 1674 744 419 268 186 137 105 83 67 55 47 40 34 30 17 11 6494 1624 722 406 260 180 133 101 80 65 54 45 38 33 29 16 10 6156 1539 684 385 246 171 126 96 76 62 51 43 36 31 27 15 10 5683 1421 631 355 227 158 116 89 70 57 47 39 34 29 25 14 9 5074 1268 564 317 203 141 104 79 63 51 42 35 30 26 23 13 8 4330 1082 481 271 173 120 88 68 53 43 36 30 26 22 19 11 7 3450 863 383 216 138 96 70 2435 609 271 152 97 68 50 54 38 43 35 29 24 20 18 15 9 6 30 1285 321 143 80 51 36 26 20 16 13 11 9 8 7 6 d ~I 'Sample size less than 5. 54 43 35 29 24 20 18 15 9 6 24 20 17 14 12 11 6 * * en III 3 "CI iD 1/1 jij' CD Q. .... ... CD CD 3 S' .... III 0' :s Table 2. Estimating a population proportion with specified relative precision n=zi _>/z(l-P)/sz P (a) Confidence level 95% p 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 31431 7858 3492 1964 1257 873 641 491 388 314 140 79 50 35 26 20 13 25611 6403 2846 1601 1024 711 523 400 316 256 114 64 41 28 21 16 10 20686 5171 2298 1293 827 575 422 323 255 207 92 52 33 23 17 13 8 16464 4116 1829 1029 659 457 336 257 203 165 73 41 26 18 13 10 7 12805 3201 1423 800 512 356 261 200 158 128 57 32 20 14 10 8 5 9604 2401 1067 600 384 267 196 150 119 96 43 24 15 11 8 6 6779 1695 753 424 271 188 138 106 84 68 30 17 11 8 6 4268 1067 474 267 171 119 87 67 53 43 19 11 7 5 2022 505 225 126 81 56 41 32 25 20 9 5 G tjl 0.01 729904 345744 0.02 182476 86436 0.03 81100 38416 0.04 45619 21609 0.05 29196 13830 0.06 20275 9604 0.07 14896 7056 0.08 11405 5402 0.09 9011 4268 0.10 3457 7299 0.15 3244 1537 864 0.20 1825 0.25 1168 553 0.30 811 384 282 0.35 596 216 0.40 456 138 0.50 292 • Sample size less than 5. 217691 153664 115248 89637 71344 57624 46953 38416 54423 38416 28812 22409 17836 14406 11738 9604 7927 5217 4268 24188 17074 12805 9960 6403 2935 2401 13606 9604 7203 5602 4459 3602 2854 6147 4610 3585 2305 1878 1537 8708 1304 1067 6047 4268 3201 2490 1982 1601 784 2352 1829 1456 1176 958 4443 3136 1115 734 600 3401 2401 1801 1401 900 474 881 711 580 2688 1897 1423 1107 713 2177 1537 1152 576 470 384 896 317 171 968 683 512 398 256 209 224 178 144 117 96 544 384 288 61 246 143 114 92 75 348 184 79 242 171 128 100 64 52 43 31 178 125 94 73 58 47 38 45 136 96 72 56 36 29 24 29 19 15 87 61 46 36 23 . -I III cr CD N Table 2 ( continued) (b) Confidence level 90% p 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 33074 8268 3675 2067 1323 919 675 517 408 331 147 83 53 37 27 21 13 27060 6765 3007 1691 1082 752 552 423 334 271 120 68 43 30 22 17 11 22140 5535 2460 1384 886 615 452 346 273 221 98 55 35 25 18 14 9 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 11597 2899 1289 725 464 322 237 181 143 1'6 52 29 19 13 9 7 5 9020 2255 1002 564 361 251 184 141 111 90 40 23 14 10 7 6 6765 1691 752 423 271 188 138 106 4775 1194 531 298 191 133 97 75 59 48 21 12 8 5 3007 752 334 188 120 1424 356 158 89 57 40 29 22 18 14 6 e ~I 0.01 514145 243542 153341 108241 81181 0.02 128536 60886 38335 27060 20295 0.03 57127 27060 17038 12027 9020 0.04 32134 15221 9584 6765 5074 0.05 20566 9742 6134 4330 3247 0.06 14282 6765 4259 3007 2255 0.07 10493 4970 3129 2209 1657 0.08 8034 3805 2396 1691 1268 0.09 6347 3007 1893 1336 1002 0.10 5141 2435 1533 1082 812 0.15 682 2285 1082 481 361 0.20 271 1285 609 383 203 0.25 823 245 173 390 130 0.30 571 271 170 120 90 0.35 125 199 66 420 88 0.40 321 152 96 68 51 0.50 61 97 43 206 32 'Sample size less than 5. 63141 50255 40590 15785 12564 10148 5584 4510 7016 3141 3946 2537 2526 2010 1624 1754 1396 1128 1289 1026 828 987 785 634 620 780 501 631 503 406 223 281 180 126 158 101 101 80 65 70 56 45 41 52 33 31 39 25 25 20 16 18040 14571 4510 3643 1619 2004 1128 911 722 583 501 405 368 297 228 282 180 223 180 146 80 65 45 36 29 23 16 20 15 12 11 9 7 6 84 68 30 17 11 8 6 84 61 47 37 30 13 8 5 en CI 3 'C i' III j;j' CD .... ..o·= Q. CD CD 3 CI = Table 3. Hypothesis tests for a population proportion For a one-sided test 11= [Zl-",/[Po(l-PO)]+Zl-p,/[Pa(l-Pa)J}2/(Po-Pa)2. For a two-sided test 11 = {z 1 -,.'2 J [P 0(1 - P o)J + z 1 -pJ[P a(1 - P a)J}2 I(P 0 - Pa)2. (a) Level of significance 5%, power 90%, one-sided test pol 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 239 76 378 40 109 498 25 54 137 601 18 33 66 161 686 13 22 39 75 180 754 10 16 26 44 83 195 804 8 12 19 29 48 88 205 837 6 10 14 20 31 50 92 211 853 5 8 11 15 21 32 52 93 213 852 0.60 0.65 6 8 11 16 22 33 52 93 210 834 5 7 9 12 16 22 32 51 91 203 799 0.70 0.75 0.80 0.85 0.90 0.95 Pa ~I 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 221 67 34 21 15 11 8 7 5 362 102 49 30 20 14 11 9 7 6 5 • Sample size less than 5. 485 131 62 36 24 17 13 10 8 6 5 589 156 72 42 27 19 14 11 8 7 5 676 176 80 46 30 21 15 11 9 7 5 746 191 87 49 31 22 16 12 9 7 5 799 203 91 51 32 22 16 12 9 7 5 834 210 93 52 33 22 16 11 8 6 852 213 93 52 32 21 15 11 8 5 853 211 92 50 31 20 14 10 6 837 205 88 48 29 19 12 8 804 195 83 44 26 16 10 754 180 75 39 22 13 5 7 9 12 16 22 31 49 87 191 746 686 161 66 33 18 5 7 9 11 15 21 30 46 80 176 676 601 137 54 25 5 7 8 11 14 19 27 42 72 156 589 498 109 40 5 6 8 10 13 17 24 36 62 131 485 378 76 Ii * 5 6 7 9 11 14 20 30 49 102 362 239 5 7 8 11 15 21 34 67 221 Table 3 ( continued) (b) Level of significance 5%, power 80%, one-sided test , Po I 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 184 60 283 32 83 368 21 42 103 441 15 26 50 119 501 11 18 30 56 133 548 8 13 20 33 61 143 584 7 10 14 22 35 65 149 607 5 8 11 15 23 37 67 153 617 5 6 8 11 16 24 38 68 154 615 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 Pa W 0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 * 150 44 22 14 9 7 5 253 69 33 20 13 10 7 6 5 * Sample size less than 5. 342 91 43 25 16 12 9 7 5 419 109 50 29 19 13 10 8 6 5 483 125 57 32 21 15 11 8 6 5 535 137 62 35 23 16 11 9 7 5 574 145 65 37 23 16 12 9 7 5 601 151 67 38 24 16 12 9 7 5 615 154 68 38 24 16 11 8 6 5 617 153 67 37 23 15 11 8 5 607 149 65 35 22 14 10 7 5 7 9 12 16 24 38 67 151 601 584 143 61 33 20 13 8 5 7 9 12 16 23 37 65 145 574 548 133 56 30 18 11 en 3 'C ;I» 5 7 9 11 16 23 35 62 137 535 501 119 50 26 15 5 6 8 11 15 21 32 57 125 483 441 103 42 21 5 6 8 10 13 19 29 50 109 419 368 83 32 III N' III * 7 5 9 12 16 25 43 91 342 6 283 60 7 10 13 20 33 69 253' 184 ... ..o·= Q. 5 III III * 5 7 9 14 22 44 150 ~. I» = Table 3 ( continued) (c) Level of significance 5%, power 90%, two-sided test 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.01 0.02 0.03 0.04 0.05 5353 1423 668 395 264 9784 2524 1155 667 438 13686 3490 1580 905 589 17061 4324 1947 1109 718 189 19911 5026 2255 1279 826 22234 5597 2504 1417 912 24032 6036 2695 1522 978 25305 6344 2827 1594 1022 26052 6521 2901 1633 1045 26273 6565 2916 1639 1047 87 214 97 233 105 248 111 62 257 114 259 113 62 38 25 X ~ I Pa - Pol ~I 0.10 79 122 0.15 40 69 0.20 0.25 0.30 25 17 12 35 24 17 1. 74 43 29 20 0.35 0.40 0.45 0.50 10 8 6 5 13 10 8 6 15 12 9 7 X is the smaller of Po and (1 - Pol· • Sample size less than 5. 50 56 60 64 261 115 63 33 23 36 25 38 26 40 40 40 27 27 27 17 13 10 8 18 14 11 8 19 14 11 8 19 14 11 8 19 14 10 8 19 13 10 7 17 12 8 I! Table 3 ( continued) (d) Level of significance 5%, power 80%, two-sided test ~x 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.01 0.02 0.03 0.04 0.05 3933 1031 478 280 185 7250 1856 844 485 316 10172 2582 1164 664 430 12701 3209 1440 818 528 14837 3737 1673 947 610 16580 4167 1861 1052 676 19626 4905 2179 1225 783 53 26 16 86 114 131 53 53 172 18 82 85 31 36 41 44 46 20 14 24 17 9 7 11 8 7 5 12 9 7 6 18 13 10 8 28 20 14 11 8 7 30 20 15 11 8 7 48 30 21 15 11 8 181 86 48 114 41 24 16 12 161 11 18888 4732 2107 1188 761 111 19453 4868 2165 1219 780 0.10 17930 4499 2006 1132 727 184 lPa- Pol w N 0.11 0 •• 0.21 0.30 0.35 0.40 0.45 0.50 en 11 8 6 5 X is the smaller of Po and (1 - Po)' • Sample size less than 5. 6 5 a: 6 6 30 20 15 11 8 6 II 41 » 20 14 10 7 I» 3 ;- "0 1/1 i:r CD Q. CD .... 2. ..o· CD ::::I I» ::::I Table 4. Estimating the difference between two population proportions with specified absolute precision n =zI ~,dP dl- PI) + P z (1- Pz)]/d z or n=zi~"lz V/d z where V=P 1 (l-Pd+P 2 (l-P Z ) (a) Values of V X 0.01 0.02 0.03 0.04 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.08 0.09 0.10 0.12 0.13 0.14 0.16 0.17 0.18 0.19 0.20 0.21 0.22 0.23 0.23 0.24 0.25 0.25 0.25 0.26 0.26 0.26 0.26 0.03 0.04 0.05 0.06 0.07 0.08 0.08 0.09 0.10 0.11 0.13 0.14 0.15 0.17 0.18 0.19 0.20 0.21 0.22 0.23 0.24 0.24 0.25 0.26 0.26 0.26 0.27 0.27 0.27 0.27 0.04 0.05 0.06 0.07 0.08 0.09 0.09 0.10 0.11 0.12 0.13 0.15 0.16 0.18 0.19 0.20 0.21 0.22 0.23 0.24 0.25 0.25 0.26 0.26 0.27 0.27 0.28 0.28 0.28 0.28 0.05 0.06 0.07 0.08 0.09 0.09 0.10 0.11 0.12 0.13 0.14 0.16 0.17 0.19 0.20 0.21 0.22 0.23 0.24 0.25 0.26 0.26 0.27 0.27 0.28 0.28 0.28 0.29 0.29 0.29 0.06 0.07 0.08 0.09 0.10 0.10 0.11 0.12 0.13 0.14 0.15 0.17 0.18 0.20 0.21 0.22 0.23 0.24 0.25 0.26 0.27 0.27 0.28 0.28 0.29 0.29 0.29 0.30 0.30 0.30 0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.16 0.17 0.18 0.20 0.21 0.22 0.24 0.25 0.26 0.27 0.28 0.29 0.30 0.31 0.31 0.32 0.33 0.33 0.33 0.34 0.34 0.34 0.34 0.14 0.15 0.16 0.17 0.18 0.18 0.19 0.20 0.21 0.22 0.23 0.25 0.26 0.28 0.29 0.30 0.31 0.32 0.33 0.34 0.35 0.35 0.36 0.36 0.37 0.37 0.37 0.38 0.38 0.38 0.17 0.18 0.19 0.20 0.21 0.22 0.23 0.23· 0.24 0.25 0.27 0.28 0.29 0.31 0.32 0.33 0.34 0.35 0.36 0.37 0.38 0.38 0.39 0.40 0.40 0.40 0.41 0.41 0.41 0.41 0.20 0.21 0.22 0.23 0.24 0.24 0.25 0.26 0.27 0.28 0.29 0.31 0.32 0.34 0.35 0.36 0.37 0.38 0.39 0.40 0.41 0.41 0.42 0.42 0.43 0.43 0.43 0.44 0.44 0.44 0.22 0.23 0.24 0.25 0.26 0.27 0.28 0.28 0.29 0.30 0.32 0.33 0.34 0.36 0.37 0.38 0.39 0.40 0.41 0.42 0.43 0.43 0.44 0.45 0.45 0.45 0.46 0.46 0.46 0.46 0.24 0.25 0.26 0.27 0.28 0.28 0.29 0.30 0.31 0.32 0.33 0.35 0.36 0.38 0.39 0.40 0.41 0.42 0.43 0.44 0.45 0.45 0.46 0.46 0.47 0.47 0.47 0.48 0.48 0.48 0.25 0.26 0.27 0.28 0.29 0.30 0.31 0.31 0.32 0.33 0.35 0.36 0.37 0.39 0.40 0.41 0.42 0.43 0.44 0.45 0.46 0.46 0.47 0.48 0.48 0.48 0.49 0.49 0.49 0.49 0.26 0.27 0.28 0.29 0.30 0.30 0.31 0.32 0.33 0.34 0.35 0.37 0.38 0.40 0.41 0.42 0.43 0.44 0.45 0.46 0.47 0.47 0.48 0.48 0.49 0.49 0.49 0.50 0.50 0.50 0.26 0.27 0.28 0.29 0.30 0.31 0.32 0.32 0.33 0.34 0.36 0.37 0.38 0.40 0.41 0.42 0.43 0.44 0.45 0.46 0.47 0.47 0.48 0.49 0.49 0.49 0.50 0.50 0.50 0.50 Y ~I 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 Xis the smaller of P2 and (1-P2 )· Y is the smaller of P, and (1 - P,). I~ Table 4 ( continued) (b) Sample size for confidence level 95% d 0.01 0.02 0.03 0.04 0.05 0.10 0.15 0.20 0.25 385 769 1153 1537 1921 2305 2690 3074 3458 3842 4610 5379 6147 6915 7684 8452 9220 9989 10757 11525 12294 13062 13830 14599 15367 16135 16904 17672 18440 19209 97 193 289 385 481 577 673 769 865 961 1153 1345 1537 1729 1921 2113 2305 2498 2690 2882 3074 3266 3458 3650 3842 4034 4226 4418 4610 4803 43 86 129 171 214 257 299 342 385 427 513 598 683 769 854 940 1025 1110 1196 1281 1366 1452 1537 1623 1708 1793 1879 1964 2049 2135 25 49 73 97 121 145 169 193 217 241 289 337 385 433 481 529 577 625 673 721 769 817 865 913 961 1009 1057 1105 1153 1201 16 31 47 62 77 93 108 123 139 154 185 216 246 277 308 339 369 400 431 461 492 523 554 584 615 646 677 707 738 769 8 12 16 20 24 27 31 35 39 47 54 62 70 77 85 93 100 108 116 123 131 139 146 154 162 170 177 185 193 6 7 9 11 12 14 16 18 21 24 28 31 35 38 41 45 48 52 55 59 62 65 69 72 76 79 82 86 5 6 7 8 9 10 12 14 16 18 20 22 24 25 27 29 31 33 35 37 39 41 43 45 47 49 5 5 6 7 8 9 10 12 13 14 15 16 18 19 20 21 23 24 25 26 28 29 30 31 0.30 0.35 0.40 0.45 0.50 V w ~ 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 'Sample size less than 5. * * * en III * 5 6 6 7 8 9 10 11 12 12 13 14 15 16 17 18 18 19 20 21 22 * 3 'tI ii' III 5 6 6 7 7 8 9 9 10 11 11 12 12 13 14 14 15 16 16 N' III Q. .....3 III III 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 5 5 5 6 6 7 7 7 8 8 8 9 9 10 10 5' * * 5 5 5 6 6 6 7 7 7 8 8 8 ...0' III :::I Table 4 ( continued) (c) Sample size for confidence level 90% d 0.01 0.02 0.03 0.04 0.05 0.10 0.15 0.20 271 542 812 1083 1354 1624 1895 2165 2436 2707 3248 3789 4330 4871 5413 5954 6495 7036 7577 8119 8660 9201 9742 10283 10825 11366 11907 12448 12989 13531 68 136 203 271 339 406 474 542 609 677 812 948 1083 1218 1354 1489 1624 1759 1895 2030 2165 2301 2436 2571 2707 2842 2977 3112 3248 3383 31 61 91 121 151 181 211 241 271 301 361 421 482 542 602 662 722 782 842 903 963 1023 1083 1143 1203 1263 1323 1384 1444 1504 17 34 51 68 85 102 119 136 153 170 203 237 271 305 339 373 406 440 474 508 542 576 609 643 677 711 745 778 812 846 11 22 33 44 55 65 76 87 98 109 130 152 174 195 217 239 260 282 304 325 347 369 390 412 433 455 477 498 520 542 6 9 11 14 17 19 22 25 28 33 38 44 49 55 60 65 71 76 82 87 93 98 103 109 114 120 125 130 136 5 7 8 9 10 11 13 15 17 20 22 25 27 29 32 34 37 39 41 44 46 49 51 53 56 58 61 5 5 6 7 7 9 10 11 13 14 15 17 18 19 21 22 24 25 26 28 29 30 32 33 34 0.25 0.30 0.35 0.40 0.45 0.50 V ~I 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 'Sample size less than 5. * 5 6 7 7 8 9 10 11 12 13 13 14 15 16 17 18 19 20 20 21 22 5 5 6 7 7 8 8 9 10 10 11 11 12 13 13 14 14 15 16 * 5 5 6 6 7 7 8 8 8 9 9 10 10 11 11 12 I~ 5 5 5 6 6 6 7 7 7 8 8 8 9 9 5 5 5 5 6 6 6 6 7 7 7 * 5 5 5 5 5 6 6 Table 5. Hypothesis tests for two population proportions For a one-sided test n = {z 1-,j[2P(1- P)] + zl_pj[P 1(1- P tl + P 2(1- P 2)]}2 I(P 1 - P 2 )2 where P=(P 1+P 2 )/2. For a two-sided test n= {z 1-"!2j[2P(1- P)] +Zl -pJ[P 1(1- P tl+ P 2(1- P 2)]}2 I(P 1 - P2)2. For a one-sided test for small proportions n ={z 1 _, + z 1 _p)2 1[0.00061 (arcsinJ P 2 - arcsinJ p 1)2]. For a two-sided test for small proportions n = {Z 1 -2/2 + Z1 _p)2 1[0.00061(arcsinJ P 2 - arcsinJp 1)2]. I~ 3 I'E. CD (a) Level of significance 5%, power 90%, one-sided test P, ~I 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 P2 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1/1 ;:i' CD ..... ..o·= Q. 474 474 153 82 53 38 29 23 19 15 13 11 9 8 7 6 5 748 217 109 67 46 34 26 21 17 14 12 10 8 7 6 5 * Sample size less than 5. 153 748 988 273 131 79 53 39 29 23 18 15 12 10 9 7 6 5 82 217 988 1194 320 150 89 59 42 31 24 19 16 13 10 9 7 6 53 109 273 1194 1365 358 166 96 63 44 33 25 20 16 13 10 8 7 38 67 131 320 1365 1502 388 177 101 66 46 33 25 20 16 12 10 8 29 46 79 150 358 1502 1605 410 185 105 67 46 33 25 19 15 12 9 23 34 53 89 166 388 1605 1674 423 189 106 67 46 33 24 18 14 11 19 26 39 59 96 177 410 1674 1708 427 189 105 66 44 31 23 17 13 15 21 29 42 63 101 185 423 1708 1708 423 185 101 63 42 29 21 15 13 17 23 31 44 66 105 189 427 1708 1674 410 177 96 59 39 26 19 11 14 18 24 33 46 67 106 189 423 1674 1605 388 166 89 53 34 23 9 12 15 19 25 33 46 67 105 185 410 1605 1502 358 150 79 46 29 8 10 12 16 20 25 33 46 66 101 177 388 1502 1365 320 131 67 38 7 8 10 13 16 20 25 33 44 63 96 166 358 1365 1194 273 109 53 6 7 9 10 13 16 19 24 31 42 59 89 150 320 1194 988 217 82 5 6 7 9 10 12 15 18 23 29 39 53 79 131 273 988 748 153 CD CD 5 6 7 8 10 12 14 17 21 26 34 46 67 109 217 748 474 3 5 6 7 8 9 11 13 15 19 23 29 38 53 82 153 474 III = Table 5 ( continued) (b) Level of significance 5%, power 80%, one-sided test \ P2 \ ~I p,i 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 0.05 343 111 60 39 28 21 17 14 12 10 8 7 6 5 5 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 343 111 541 60 157 714 39 79 197 862 28 49 95 231 986 21 34 57 109 259 1085 17 25 39 64 120 281 1159 14 20 28 43 70 128 296 1209 12 16 22 31 46 74 134 306 1233 10 13 8 11 14 18 24 33 49 77 137 306 1209 7 9 11 14 19 25 34 49 76 134 296 1159 6 8 9 12 15 19 25 33 48 74 128 281 1085 541 157 79 49 34 25 20 16 13 11 9 8 7 6 5 • 'Sample size less than 5. 714 197 95 57 39 28 22 17 14 11 9 8 7 6 5 862 231 109 64 43 31 23 18 14 12 10 8 7 6 5 986 259 120 70 46 32 24 19 15 12 10 8 7 5 1085 281 128 74 48 33 25 19 15 12 9 8 6 1159 296 134 76 49 34 25 19 14 11 9 7 1209 306 137 77 49 33 24 18 14 11 8 1233 309 137 76 48 32 23 17 13 10 1233 306 134 74 46 31 22 16 12 17 23 32 48 76 137 309 1233 1209 296 128 70 43 28 20 14 1159 281 120 64 39 25 17 1085 259 109 57 34 21 986 231 95 49 28 0.75 0.80 5 5 6 7 8 10 12 14 18 23 31 43 64 109 231 862 7 8 10 12 15 19 24 32 46 70 120 259 986 862 197 79 39 714 157 60 0.85 5 6 7 8 9 11 14 17 22 28 39 57 95 197 714 541 111 0.90 5 6 7 8 9 11 13 16 20 25 34 49 79 157 541 343 0.95 5 5 6 7 8 10 12 14 17 21 28 39 60 111 343 I~ Table 5 ( continued) (c) Level of significance 5%, power 90%, two-sided test ~I w (Xl 0.01 0.02 0.03 0.04 0.05 0.10 0.15 0.20 0.25 0.05 0.10 0.15 0.20 0.25 0.30 0.35 50654 12711 5669 3199 2053 52126 13053 5809 3271 2095 52546 13131 5832 3278 2095 519 524 231 128 81 5f9 ai? 54 39 28 21 16 62 36 26 19 14 0.40 0.45 0.50 "en III 19753 5143 2376 1386 918 27531 7062 3216 1852 1212 34258 8717 3940 2253 1465 39933 10110 4548 2588 1675 44558 11239 5038 2857 1843 48132 12107 5412 3061 1969 336 393 477 503 181 1. 440 203 97 109 118 47 266 133 82 57 85 36 42 47 28 23 19 16 33 26 21 17 36 28 23 19 72 52 39 30 24 19 10924 2962 1418 854 582 188 101 65 0.30 0.35 0.40 0.45 0.50 x is the smallest of P" (1 - P,). P2 and (1 - P2 ). 217 2T1 231 128 77 125 81 54 40 31 24 19 56 41 31 24 19 57 130 82 68 41 31 24 19 40 30 23 17 82 125 77 3 'tI CD 1/1 iii' CD ..... ..cj' Q. CD CD 3 S· III = ( continued) Table 5 (d) Level of significance 5%, power 80%, two-sided test ~x I 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 8161 2213 1060 638 435 141 76 49 36. 27 22 18 15 12 14756 3842 1775 1036 686 200 100 62 43 32 25 20 16 14 20566 5275 2403 1384 906 251 121 73 49 36 27 22 17 14 25590 6512 2944 1683 1095 294 138 82 54 39 29 23 18 15 29830 7552 3398 1934 1252 329 152 89 33284 8396 3764 2135 1377 357 163 94 61 42 31 24 19 15 35954 9044 4043 2287 1471 376 170 96 62 43 31 24 18 14 37838 9495 4235 2390 1534 388 173 97 62 42 31 23 17 14 38937 9751 4340 2444 1566 392 173 96 61 41 29 22 16 12 39251 9809 4357 2449 1566 388 170 94 I P2 - P,1 0.01 0.02 0.03 0.04 0.05 0.10 0.15 4).20 0.25 0.30 0.35 0.40 0.45 0.50 w co 58 41 31 24 19 15 58 39 27 20 15 11 X is the smallest of Pl' (1 - P,). P2 and (1 - P2 ). "-I = cr iD en (e) Level of significance 5%, power 90%, one-sided test. small proportions "- P, I 0.0001 0.0002 0.0003 0.0004 0.0005 0.0010 0.0025 0.0050 0.0075 0.0100 0.0200 0.0300 0.0400 0.0500 249634 79919 423934 42827 124798 596429 28029 63398 168559 768327 9158 14011 20928 31688 49897 2675 3328 4006 4753 5600 12662 1160 1336 1500 1662 1828 ·2796 9960 728 813 890 963 1035 1412 3182 16856 527 579 624 667 708 912 1703 4957 23657 246 262 276 288 300 362 607 847 1407 2460 159 167 174 180 186 211 278 401 661 784 4135 117 122 126 130 134 149 187 251 326 418 1212 5758 92 96 99 101 104 114 13B 179 222 273 613 1619 7341 P2 0.0001 0.0002 0.0003 0.0004 0.0005 0.0010 0.0025 0.0050 0.0075 0.0100 0.0200 0.0300 0.0400 0.0500 249634 79919 42827 28029 9158 2676 1160 728 527 246 159 117 92 423934 124798 63398 14011 3328 1336 813 579 262 167 122 96 596429 168559 20928 4006 1500 890 624 276 174 126 99 768327 31688 4763 1662 963 667 288 180 130 101 49897 6600 1828 1035 708 300 186 134 104 12662 2796 1412 912 352 211 149 114 9950 3182 1703 507 278 187 139 16856 4957 847 401 251 179 23657 1407 561 326 222 2460 784 418 273 4135 1212 613 5758 1619 7341 Table 5 ( continued) (h) Level of significance 5%, power 80%, two-sided test, small proportions P1 i 0.0001 0.0002 0.0003 0.0004 0.0005 0.0010 0.0025 0.0050 0.0075 0.0100 228767 73239 388498 39247 114367 546575 25686 58099 154470 704104 8392 12840 19179 29040 45727 2451 3050 3671 4356 5132 1063 1224 1374 1523 1675 667 745 815 882 948 11604 2562 9118 1294 2916 15447 483 530 572 611 648 836 0.0200 0.0300 0.0400 226 241 253 264 275 146 153 160 165 171 107 112 116 119 123 85 88 91 93 95 323 465 776 1289 2254 193 255 368 614 719 136 104 127 0.0500 P2 .j::. ..... 0.0001 0.0002 0.0003 0.0004 0.0005 0.OQ10 228767 73239 39247 25686 ().002$ 8392 2451 0.0050 1063 0.0075 667 0.0100 0.0200 0.0300 0.0400 0.0500 483 226 146 107 85 388498 114367 58099 546575 154470 3050 1224 745 630 19179 3671 1374 816 572 241 153 112 88 253 160 116 91 12840 704104 29040 4356 1523 882 611 264 165 119 93 457.27 5132 1675 948 648 275 171 123 95 11604 2562 1294 836 323 193 136 104 9118 2916 1561 465 255 171 127 15447 4542 776 368 230 164 21680 1289 514 299 204 1561 4542 21680 2254 719 383 250 3789 3789 1110 561 5277 1484 171 230 299 383 1110 5277 6728 164 204 250 561 1484 6728 I! Table 6. Estimating an odds ratio with specified relative precision n = zi -a/2 {1 /[Pi( 1 - Pi)] + 1/[P!(1- P!)]} /[loge(1- e)Y (a) Confidence level 95%, relative precision 10% '\oRI P* 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 69912 35313 23785 18025 14572 7691 5429 4326 3692 3296 3043 2884 2797 2769 2797 2884 3296 4326 7691 63061 31923 21550 16367 13261 7078 5052 4071 3513 3172 2961 2838 2783 2786 2846 2968 3469 4655 8462 58494 29664 20061 15263 12389 6672 4806 3908 3403 3101 2922 2827 2798 2827 2914 3067 3651 4990 9235 55232 28051 18998 14476 11767 6385 4634 3798 3333 3062 2907 2835 2828 2880 2993 3175 3838 5327 10010 52786 26842 18201 13886 11302 6172 4510 3721 3288 3041 2908 2856 2869 2942 3078 3288 4030 5667 10786 50883 25901 17582 13429 10941 6009 4416 3665 3259 3033 2919 2884 2916 3009 3168 3405 4223 6009 11563 49361 25149 17087 13063 10654 5880 4344 3626 3242 3034 2937 2919 2968 3080 3262 3525 4419 6351 12340 48116 24535 16683 12765 10419 5776 4288 3597 3233 3042 2960 2958 3023 3154 3357 3646 4615 6694 13117 47079 24023 16347 12516 10225 5691 4244 3576 3230 3055 2987 3000 3081 3230 3454 3769 4812 7037 13894 46202 23589 16063 12307 10061 5621 4209 3563 3233 3071 3017 3044 3141 3308 3553 3893 5011 7381 14672 45449 23219 15819 12128 9921 5562 4181 3554 3239 3090 3050 3090 3203 3387 3652 4017 5209 7725 15450 44798 22897 15609 11974 9800 5513 4159 3549 3248 3112 3084 3138 3265 3467 3752 4143 5408 8070 16228 44228 22616 15425 11839 9695 5470 4141 3548 3259 3136 3120 3187 3329 3548 3854 4269 5608 8414 17006 43725 22369 15263 11720 9603 5434 4128 3549 3273 3161 3157 3237 3394 3629 3955 4395 5807 8759 17784 43278 22149 15120 11615 9521 5403 4117 3552 3288 3187 3195 3288 3459 3711 4057 4522 6007 9104 18562 42878 21952 14991 11522 9449 5376 4110 3558 3305 3215 3234 3340 3525 3794 4160 4649 6207 9449 19340 42518 21776 14876 11438 9384 5353 4104 3565 3323 3244 3274 3392 3591 3876 4262 4776 6408 9794 20118 2 ~I 0.01 0.02 0.03 0.04 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.70 0.80 0.90 For OR < 1, use the colur.m value corresponding to 1lOR and the row value corresponding to Pi. en I» 3 'tI CD III j;j' CD ... .. Q. CD CD 3 5' I» 0' ::I Table 6 ( continued) (b) Confidence level 95%, relative precision 20% \oRI P* 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 15587 7873 5303 4019 3249 1715 1211 965 823 735 679 643 624 618 624 643 735 965 1715 14059 7117 4805 3649 2957 1578 1127 908 784 708 660 633 621 622 635 662 774 1038 1887 13041 6614 4473 3403 2762 1488 1072 872 759 692 652 631 624 631 650 684 814 1113 2059 12314 6254 4236 3228 2624 1424 1034 847 744 683 649 632 631 643 668 708 856 1188 2232 11768 5984 4058 3096 2520 1376 1006 830 733 678 649 637 640 656 687 733 899 1264 2405 11344 5775 3920 2994 2440 1340 985 818 727 677 651 643 650 671 707 760 942 1340 2578 11005 5607 3810 2913 2376 1311 969 809 723 677 655 651 662 687 728 786 985 1416 2751 10727 5470 3720 2846 2323 1288 956 802 721 679 660 660 674 704 749 813 1029 1493 2925 10496 5356 3645 2791 2280 1269 946 798 721 681 666 669 687 721 770 841 1073 1569 3098 10301 5259 3581 2744 2243 1254 939 795 721 685 673 679 701 738 792 868 1117 1646 3271 10133 5177 3527 2704 2212 1240 932 793 722 689 680 689 714 755 815 896 1162 1723 3445 9988 5105 3480 2670 2185 1229 928 792 724 694 688 700 728 773 837 924 1206 1799 3618 9860 5042 3439 2640 2162 1220 924 791 727 699 696 711 743 791 859 952 1251 1876 3792 9748 4987 3403 2613 2141 1212 921 792 730 705 704 722 757 809 882 980 1295 1953 3965 9649 4938 3371 2590 2123 1205 918 792 733 711 713 733 772 828 905 1008 1340 2030 4139 9560 4894 3343 2569 2107 1199 917 794 737 717 721 745 786 846 928 1037 1384 2107 4312 9479 4855 3317 2550 2093 1194 915 795 741 724 730 757 801 865 951 1065 1429 2184 4486 2 ~ w 0.01 0.02 0.03 0.04 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.70 0.80 0.90 For OR < 1, use the column value corresponding to 1/ OR and the row value corresponding to Pi. I~ Table 6 (continued) (c) Confidence level 95%, relative precision 25% ""'p.OR I 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 0.01 0.02 0.03 0.04 0.05 (UO 9378 4737 3191 2418 1955 8459 4282 2891 2196 1779 960 7846 3979 2691 2048 1662 6315 3223 2193 1679 1372 764 6198 3165 2155 1651 1350 6097 3115 2122 1627 1331 6009 3072 2094 1606 1315 5933 3034 2069 1588 1301 5805 2971 2028 1558 1278 5752 2945 2011 1546 1268 5703 2921 1996 1535 1259 1M 566 M8 625 510 499 472 425 467 448 441 438 411 407 398 381 374 374 382 399 392 380 376 380 391 390 381 380 387 402 408 390 383 385 395 413 435 407 4M 416 392 387 392 404 425 394 392 399 414 438 478 434 4'2 7. 5. 476 734 570 741 .1 477 435 415 5865 3001 2048 1572 1288 729 593 492 6621 3374 2292 1753 1429 789 583 487 6454 3291 2238 1713 1398 &46 7081 3601 2442 1863 1516 828 006 6825 3475 2359 1802 1468 806 078 7409 3763 2549 1942 1579 867 822 412 490 670 1239 426 615 715 1343 441 541 457 587 761 1447 1551 473 593 852 1658 2 tl 0.16 0;20 0.28 0. • 0.35 0.40 0.45 0.50 0.55 0.80 0~70 0.80 0.90 1()32 721 M1 443 409 387 376 372 376 387 443 581 1032 ... 625 1135 896 806 775 576 493 4M 408 397 397 406 424 451 489 619 4'0 401 403 414 434 464 506 646 898 944 1760 1864 For OR< 1, use the column value corresponding to 1 lOR and the row value corresponding to Pi. ... 4,a 405 409 422 444 477 523 409 415 430 455 490 539 672 699 414 421 438 465 504 556 725 1. 1OIl 2073 1083 2177 - -- -.. - .. 726 m 118 .2 .1 478 419 441 444 448 4D 432 476 514 478 417 421 424 419 428 447 476 517 424 435 456 487 531 429 441 464 498 545 434 448 473 509 558 440 455 482 520 572 014 041 1. 2_ 1t14 ·2611 566 573 590 753 119 1121 2111 1116 2316 1m 2490 (I) III 3 ;- "CI UI jij' CD .... ..o·s· CI. CD CD 3 III = Table 6 ( continued) (d) Confidence level 95%, relative precision 50% pRi 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 1616 816 550 417 337 118 1458 738 498 379 307 164 1352 686 464 353 287 1277 649 439 335 272 1220 621 421 321 262 1176 599 407 311 253 1141 582 395 302 247 1112 567 386 295 241 1088 556 378 290 237 1068 546 372 285 233 1051 537 366 281 230 1036 530 361 277 227 1022 523 357 274 224 1011 517 353 271 222 1000 512 350 269 220 991 508 347 267 219 983 504 344 265 217 143 105 134 100 125 125 124 95 71 72 75 71 75 71 82 76 13 82 76 14 96 83 82 74 128 97 82 76 96 84 130 98 83 75 71 96 84 132 99 83 75 71 129 97 86 139 103 85 76 71 136 101 0.25 0.30 148 108 88 78 71 126 117 155 112 91 79 127 126 100 0.35 0.40 0.45 0.50 0.55 71 67 65 64 65 69 66 65 65 66 68 66 65 66 68 68 66 66 67 70 68 66 67 68 68 68 69 72 76 69 69 70 73 78 70 70 72 75 80 70 71 73 77 83 71 72 72 72 68 67 68 70 74 74 79 85 73 76 81 87 0.60 0.70 0.80 0.90 61 77 100 178 69 81 108 196 71 85 116 214 74 89 124 232 76 94 131 250 79 98 139 268 82 103 147 286 85 107 155 88 112 163 322 93 121 179 357 96 304 90 116 171 339 73 74 77 82 90 99 125 187 375 130 195 393 P*2 0.01 0.02 0.03 0.04 0.05 0~10 ~I 0.15 0.20 95 86 76 71 For OR< 1, use the column value corresponding to 1/ OR and the row value corresponding to P~. 83 75 72 72 76 74 77 95 83 77 75 75 73 75 79 84 92 74 76 80 86 94 75 78 82 88 97 76 79 83 90 99 102 135 203 411 105 139 211 429 108 144 219 447 111 149 227 465 83 I~ Table 6 ( continued) (e) COi'lfidence level 90%, relative precision 10% ~oRI 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 49246 24875 16754 12697 10264 5418 3824 3048 2601 2322 2144 2032 1970 1951 1970 2032 2322 3048 5418 44421 22487 15180 11529 9341 4986 3559 2868 2475 2234 2086 1999 1961 1963 2005 2091 2443 3279 5961 41203 20896 14131 10752 8727 4700 38906 19759 13382 10197 8289 4498 3265 2675 2348 2157 2048 1997 1992 2029 2108 2236 2704 3753 7051 37182 18907 12821 9782 7961 4348 3177 2621 2316 2142 2048 2012 2021 2073 2169 2316 2839 3992 7598 35842 18245 12385 9459 7707 4233 3111 2582 2296 2137 2056 2032 2054 2120 2232 2399 2975 4233 8145 34770 17715 12037 9202 7505 4142 3060 2554 2284 2138 2069 2056 2091 2170 2298 2483 3113 4474 8692 33893 17282 11752 8992 7339 4069 3021 2534 2278 2143 2085 2084 2130 2222 2365 2568 3251 4715 9240 33163 16922 11515 8817 7202 4009 2989 2519 2276 2152 2104 2113 2171 2276 2433 2655 3390 4957 9787 32545 16617 11315 8669 7087 3960 2965 2510 2277 2163 2125 2144 2213 2330 2503 2742 3530 5199 10335 32015 16355 11143 8543 6988 3918 2945 2503 2281 2177 2148 2177 2256 2386 2573 2830 3669 5442 10883 31556 16129 10995 8434 6903 3883 2930 2500 2288 2192 2172 2211 2300 2442 2643 2918 3810 5684 11431 31154 15931 10866 8339 6829 3853 2917 2499 2296 2209 2198 2245 2345 2499 2715 3007 3950 5927 11979 30800 15757 10752 8256 6764 3828 2908 2500 2306 2227 2224 2280 2391 2556 2786 3096 4091 6170 12527 30485 15602 10650 8182 6707 3806 2900 2503 2316 2245 2251 2316 2437 2614 2858 3185 4232 6413 13075 30203 15463 10560 8116 6656 3787 2895 2506 2328 2265 2278 2353 2483 2672 2930 3275 4373 6656 13623 29950 15339 10479 8057 6610 3771 2891 2511 2341 2285 2306 2389 2530 2731 3003 3364 4514 6899 14172 P* 2 ~I 0.01 0.02 0.03 0.04 0.05 0.10 0.15 0.20 0.25 0.30 0.35 O.~ 0.45 0.50 0.55 0.60 0.70 0.80 0.90 33~ 2753 2398 2185 2058 1991 1971 1991 2053 2161 2572 3515 6505 For OR < 1, use the column value corresponding to 1/ OR and the row value corresponding to Pi. en = 3 "CI CD 1/1 i;r CD Q. CD .... .o·== CD 3 = Table 6 ( continued) (f) Confidence level 90%, relative precision 20% ,PRJ 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 10979 5546 3736 2831 2289 1208 853 680 580 518 478 453 440 435 440 453 518 680 1208 9903 5014 3385 2571 2083 1112 794 640 552 499 465 446 437 438 447 467 545 731 1329 9186 4659 3151 2397 1946 1048 755 614 535 487 459 444 440 444 458 482 574 784 1451 8674 4406 2984 2274 1848 1003 728 597 524 481 457 446 445 453 470 499 603 837 1572 8290 4216 2859 2181 1775 970 709 585 517 478 457 449 451 462 484 517 633 890 1694 7991 4068 2761 2109 1719 944 694 576 512 477 459 453 458 473 498 535 664 944 'i816 7752 3950 2684 2052 1673 924 683 570 510 477 462 459 466 484 513 554 694 998 1938 7557 3853 2620 2005 1637 907 674 565 508 478 465 465 475 496 528 573 725 1052 2060 7394 3773 2567 1966 1606 894 667 562 508 480 470 471 484 508 543 592 756 1106 2182 7256 3705 2523 1933 1580 883 661 560 508 483 474 478 494 520 558 612 787 1160 2304 7138 3647 2485 1905 1558 874 657 559 509 486 479 486 503 532 574 631 818 1214 2427 7035 3596 2452 1881 1539 866 654 558 510 489 485 493 513 545 590 651 850 1268 2549 6946 3552 2423 1860 1523 859 651 558 512 493 490 501 523 558 606 671 881 1322 2671 6867 3513 2397 1841 1508 854 649 558 514 497 496 509 533 570 622 691 912 1376 2793 6797 3479 2375 1824 1496 849 647 558 517 501 502 517 544 583 638 711 944 1430 2915 6734 3448 2355 1810 1484 845 646 559 519 505 508 525 554 596 654 730 975 1484 3038 6677 3420 2337 1797 1474 841 645 560 522 510 515 533 564 609 670 750 1007 1538 3160 p*2 0.01 0.02 0.03 0.04 0.05 0.10 0.15 ~I 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.70 0.80 0.90 For OR < 1, use the column value corresponding to 1/ OR and the row value corresponding to Pi. I~ Table 6 ( continued) (g) Confidence level 90%, relative precision 25% \ ~ (Xl 0.01 0.02 0.03 0.04 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.70 0.80 0.90 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 6606 3337 2248 1703 1377 727 513 409 349 312 288 273 265 262 265 273 312 409 727 5959 3017 2037 1547 1253 669 478 385 332 300 280 269 263 264 269 281 328 440 800 5527 2803 1896 1443 1171 631 455 370 322 293 277 268 265 268 276 290 345 472 873 5219 2651 1795 1368 1112 4988 2537 1720 1312 1068 584 427 352 311 288 275 270 271 278 291 311 381 536 1020 4808 2448 1662 1269 1034 568 418 347 308 287 276 273 276 285 300 322 399 568 1093 4664 2377 1615 1235 1007 556 411 343 307 287 278 276 281 292 309 333 418 600 1166 4547 2319 1577 1206 985 546 406 340 306 288 280 280 286 298 318 345 436 633 1240 4449 2270 1545 1183 966 538 401 338 306 289 283 284 292 306 327 357 455 665 1313 4366 2229 1518 1163 951 532 398 337 306 291 286 288 297 313 336 368 474 698 1387 4295 2194 1495 1146 938 526 395 336 306 292 289 292 303 320 346 380 493 730 1460 4233 2164 1475 1132 926 521 393 336 307 294 292 297 309 328 355 392 511 763 1534 4179 2137 1458 1119 916 517 392 336 308 297 295 302 315 336 365 404 530 795 1607 4132 2114 1443 1108 908 514 390 336 310 299 299 306 321 343 374 416 549 828 1681 4089 2093 1429 1098 900 511 389 336 311 302 302 311 327 351 384 428 568 861 1754 4052 2075 1417 1089 893 4018 2058 1406 1081 887 506 388 337 314 307 310 321 340 367 403 452 606 926 1901 604 438 359 315 290 275 268 268 273 283 300 363 504 946 . For OR < 1. use the column value corresponding to 1/ OR and the row value corresponding to Pi. 508 389 337 313 304 306 316 333 359 393 440 587 893 1828 rn DI 3 ~ i' III N' CD .... ..o· CI. CD CD 3 :i' DI :::I Table 6 ( continued) (h) Confidence level 90%, relative precision 50% " ~I 0.Q1 0.02 0.03 0.04 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.70 0.80 0.90 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 1138 575 388 294 238 126 89 71 61 54 50 47 46 46 46 47 1027 520 351 267 216 116 83 67 58 52 49 47 46 46 47 49 57 76 138 952 483 327 249 202 109 79 64 66 51 48 46 46 46 48 50 60 82 151 899 457 310 236 192 104 76 62 55 50 48 47 47 47 49 52 63 87 163 860 437 297 226 184 101 74 61 829 422 287 219 179 98 12 60 804 410 279 213 174 96 71 59 53 50 48 48 49 51 54 58 72 104 784 400 272 208 170 94 70 59 53 50 49 49 50 52 55 60 76 109 214 767 391 267 204 167 93 70 69 63 50 49 49 51 53 57 62 79 115 227 752 384 262 201 164 92 69 68 63 50 50 50 52 54 58 64 82 121 239 740 378 258 198 162 91 69 68 63 51 50 51 53 56 60 66 85 126 252 730 373 255 195 160 90 68 68 53 51 51 52 54 57 62 68 89 132 265 720 369 252 193 158 90 68 68 54 52 51 52 55 58 63 70 92 137 277 712 365 249 191 157 89 68 58 54 52 52 53 56 60 65 72 95 143 290 705 361 247 190 155 88 68 68 698 358 244 188 154 88 67 68 54 53 53 55 58 62 68 76 102 154 315 692 355 243 187 153 88 67 59 55 63 54 56 59 64 70 78 105 160 328 54 71 126 54 54 50 48 47 47 48 51 54 66 93 176 50 48 47 48 49 52 56 69 98 189 201 For OR < 1, use the column value corresponding to 1/ OR and the row value corresponding to Pi. 54 52 52 54 57 61 67 74 98 149 303 I~ Table 7. Hypothesis tests for an odds ratio n = {Zl-a/2J[2Pi(1-Pi)] +Zl-flJ[Pt(1- Pt)+ Pi(l- P!)]}2 /(Pt -Pi)2 In this formula the term 2PH1 - P~) is used instead of 21'*(1 - 1'*) because the study population is likely to be made up of many more controls than cases, and the exposure rate among the controls is often known with a high degree of precision; under the null hypothesis this is the exposure rate for the cases as well. If the investigator is in doubt about the exposure rate among the controls, however, the formula should be modified and the term 21'*(1- 1'*) used, where 1'*= (Pf+ ~)/2. (a) Level of significance 5%, power 90%, two-sided test len III ORa I P:~ ~I 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 Il 5.00 III 0.01 0.02 0.03 0.04 0.05 35761 18133 12261 9327 7569 9375 4771 3237 2472 2013 4355 2224 1514 1160 948 2554 1308 894 687 563 1700 873 598 461 379 1225 631 434 335 276 932 482 332 257 213 738 383 264 205 170 602 313 217 169 140 503 262 182 142 118 428 224 156 122 102 370 194 135 106 89 0.10 0.1:6 0.20 0.26 0.30 4072 1102 807 666 689 318 161 124 1Q8 101 97 81 79 68 67 61 68 68 72 68 63 61 61 63 61 47 46 61 41 86 101 80 71 86 68 204 186 176 217 1. 143 131 126 126 644 627 392 329 296 276 46 41 42 0.35 0.40 0.45 0.50 0.55 1769 1707 1686 1699 1747 519 509 510 522 544 267 265 269 279 294 172 173 177 185 198 125 127 131 138 149 97 100 104 111 120 80 82 87 93 101 68 70 75 80 88 59 62 66 71 78 52 55 59 64 71 47 50 54 59 66 0.60 0.70 0.80 0.90 .1834 2170 679 704 317 394 660 1076 216 272 393 766 163 132 172 9& 88 118 179 80 108 166 74 101 366 3i4O 2929 2379 2068 1881 2948 982 6421 1861 240 ee 307 266 112 141 221 606 609 445 209 130 197 399 For ORa < 1, use the column value corresponding to 1/ ORa and the row value corresponding to Pi. 166 319 324 170 119 94 79 4$ 287 151 106 83 70 41 3' 37 31 14 35 32: .32 33. 43 46 50 55 61 40 43 46 51 57 37 40 43 48 54 35 38 41 46 51 69 94 141 66 89 139 .62 85 133 303 289 2.'18 42 46 • • 257 135 95 75 63 40 34:. .. ,. ;:i' 231 122 86 68 57 ~ Q. ~ 3 :37' iJ: ..•• 3I:j.:. 3I:j :1' 33 36 39 43 49 ...... .··.1•.·•• 1. ······1··· 69 .82 268 .. S' : III 0' :::I Table 7 ( continued) (b) Level of significance 5%, power 80%, two-sided test 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 26421 13400 9063 6896 5598 6858 3492 2371 1811 1476 3157 1614 1100 843 690 1836 942 644 496 407 1213 624 429 331 272 656 340 235 183 151 516 269 186 145 121 419 219 152 119 99 348 182 127 100 83 295 155 108 85 71 254 134 94 74 62 221 117 82 65 55 195 103 73 58 49 174 92 65 52 44 156 83 59 47 40 "':1 810 231 1$7 868 448 309 239 198 f.f' '''<ORal p*2 0.01 0.02 0.03 0.04 0.05 .. . .> ,. e.· !: . . .~. ······>j:i;\i • .,:., ..- - '. ... .., 1..1 4. ••.._. - .2,. - •••• =! . O?f1J ~I .. 0.:··: ~.I· ~.lD. 0. • 0.35 0.40 0.45 0.50 0.55 ..... ~:'l2 1400 1318 1273 1259 1270 1307 404 387 380 381 391 408 1._ .742 t.lO· ,-",. 1373 0.• . ~, For 1~ t~. ltf ",,,,,"C ... "lj... J4- .«). ... f't: ·21. 2GI 1'37· 110 II 128 129 133 139 149 t4 93 95 98 104 112 12 199 198 202 209 221 73 75 78 84 91 60 62 65 70 238 .12 113 100 1~1 218: 820 ·fT :1.·. a 1 iG 77 51 53 56 61 67 44 46 50 54 60 85 11 n~: '111 OCI U 11 1.' ·1M I~ . 341 ORa < 1, use the column value corresponding to 1/ ORa and the row value corresponding to P~. * 39 42 45 49 54 'fu 1.···· 36 38 41 45 50 32 35 38 42 47 17D 30 32 35 39 44 ;~E 28 30 33 37 42 26 29 31 35 40 25 27 30 33 38 Table 8. Estimating a relative risk with specified relative precision n=zi -~/2[(1-Pl)/Pl +(1-P2)/P2]/[loge(1-e)]2 (a) Confidence level 95%, relative precision 10% ~RRI 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 0.01 0.02 0.03 0.04 0.05 68521 33915 22379 16612 13151 56986 28147 18534 13728 10844 53690 26499 17436 12904 10185 51218 25263 16612 12286 9690 49295 24302 15971 11805 9306 47757 23533 15458 11421 8998 46499 22904 15039 11106 8746 43802 21555 14140 10432 8207 43143 21226 13920 10267 8075 42566 20937 13728 10123 7960 42057 20683 13558 9996 7858 41605 20457 13407 9883 7768 41200 20254 13272 9781 7687 40836 20072 13151 9690 7614 6230 3923 2769 2077 1615 4747 2984 2027 1484 1121 4499 2769 1904 4807 2641 4027 2454 1308 975 4153 2538 1131 1246 923 3Ia4 2326 1511 1119 817 3768 2275 1133 1088 3692 2231 1100 t0l2 36M 3461 2423 1800 1_ 6076 3164 2192 1615 1231 45450 22379 14689 10844 8537 a923 44563 21936 14393 10622 8359 0.10 ~I C:UI 0;20 0.21 61600 30454 20072 14881 11767 1638 1286 1039 846 693 567 1088 866 693 554 441 956 750 590 462 357 862 668 517 396 297 737 558 693 520 462 297 174 847 270 P2 0.30 0.35 0.40 0.45 0.50 0.55 (l.GO 0.70 0;80 0.90 1_ 1039 792 606 462 347 ,.- ,. 1196 1i1 657 198 81 77 Since RR= P,/P2 • RR~ 1/P2 • For RR < 1. use the column value corresponding to 1/ RR and the row value corresponding to P,. 2384 1615 1164 146 2192 1411 1039 1. aE ·21. aas .34JI 2128 1423 2t.~ 1483 14&1 ,. 2077· en DI 3 ~ i" en j;j' .. .o· CD Q. CD CD 3 :i' DI = Table 8 ( continued) (b) Confidence level 95%, relative precision 20% ~RRI 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 0.01 0.02 0.03 0.04 0.05 15276 7561 4990 3704 2932 13733 6790 4475 3318 2624 12705 6275 4132 3061 2418 11970 5908 3887 2877 2271 11419 5633 3704 2739 2161 10990 5418 3561 2632 2075 10647 5247 3447 2546 2006 10367 5107 3353 2476 1950 10133 4990 3275 2418 1904 9935 4891 3209 2368 1864 9766 4806 3153 2326 1830 9490 4668 3061 2257 1775 9377 4611 3023 2229 1752 9276 4561 2989 2204 1732 9186 4516 2959 2181 1714 9104 4475 2932 2161 1698 0.10 13. 815 618 463 .361 1236 112 U)I$ 1003 6 •• ill 618 501 342 488 541 481 1132 103 481 361 9619 4732 3104 2289 1801 .123 DO 243 250 287 232 189 155 127 243 193 155 124 99 216 214 168 132 103 80 P2 <U6 ~I tUO e,a 0:. 0.35 0.40 0.45 0.50 0.55 qQ ..... ,~,.. ~.'.: .. eo' - 1. .,. •• 11 .' 154 *2 331 193 149 116 89 67 9. .. .,--&. .,- .. • 2. ,. 911 426 m 177 136 103 78 it. 218 165 125 206 155 116 372 287 111 147 6' 20 11 Since RR=P,/P2, RR~1/P2' For RR < 1, use the column value corresponding to 1/ RR and the row value corresponding to P" '82 331 131 •• 1" 811 32& 131 412 323.· ...- ...• ttl ·31.3 111 ..1 I! Table 8 ( continued) (c) Confidence level 95%, relative precision 25% ~RRI 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 0.01 0.02 0.03 0.04 0.05 9191 4549 3002 2229 1764 8263 4085 2693 1996 1579 7644 3776 2486 1842 1455 7202 3555 2339 1731 1367 6612 3260 2143 1584 1249 6406 3157 2074 1532 1207 6237 3073 2018 1490 1174 6097 3002 1971 1455 1145 5978 2943 1931 1425 1122 5710 2809 1842 1358 1068 5642 2775 1819 1341 1054 5581 2744 1799 1326 1042 5478 2693 1764 1300 1022 636 527 372 279 217 743 465 325 242 186 661 423 294 217 166 637 576 355 243 176 131 556 341 541 330 224 161 119 527 320 515 312 211 150 110 5876 2892 1897 1400 1101 504 306 5787 2847 1868 1378 1084 0.10 0.15 ~I 0.. 20 0.25 0.30 6870 3389 2229 1648 1300 604 496 488 2M 461. 206 146 .202 143 .1 290<,.;;;;211 1M· ·;·"1!I~:~:1.11 0.35 0.40 0.45 0.50 0.55 173 140 114 93 76 146 117 93 75 60 129 101 80 62 48 99 75 93 70 0.60 0.70 0.80 0.90 62 47 27 12 37 4.75 5.00 P2 40 24 11 394 272 199 151 116 90 70 54 40 372 256 186 140 107 82 62 47 233 168 124 89 Since RR=P,/ P2' RR ,;;;1/P2 · For RR < 1, use the column value corresponding to 1/ RR and the row value corresponding to P, . 217 155 114 300 196 ~'~~~ 5527 2717 1781 1312 1031 ~;; .....;410. "o,"\W~O",,. o~; :;: ", 140 ' .. , ~ ~, ' .. ~j -" en 3 DI "CI CD 1/1 N' .. .. CD a. CD CD ~ 3 5' DI S' ::;, Table 8 (continued) (d) Confidence level 95%, relative precision 50% "(RRI 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 0.01 0.02 0.03 0.04 0.05 1584 784 518 384 304 1424 704 464 344 272 1317 651 429 318 251 1241 613 403 299 236 1184 584 384 284 224 1139 562 369 273 215 1104 544 358 264 208 1075 530 348 257 203 1051 518 340 251 198 1030 507 333 246 194 1013 499 327 242 190 984 484 318 234 184 962 473 310 229 180 952 468 307 226 178 944 464 304 224 176 144 128 118 104 100 $4 17 $1 88 81 53 12 81 54 37 • 972 478 314 231 182 (UO 0.'11 0'.21. 997 491 322 238 187 16 Sf 34 Qi2$ 48 26 26 24 P2 t1I t1I 91 M to • 42 0.3G 31· 32 0.35 0.40 0.45 0.50 0.55 30 24 20 16 14 11 26 20 16 13 11 0;71. ,.• oi9l·· • 1;. O~.L 8 1 • 73 no 68 51 47 31 35 28 23 18 14 11 9 2.20 16 12 10 7 G4 $2 44 42 M 31 23 ·31 19 14 11 8 18 13 " • a 59 ~ 2$ 22 16 12 21 16 7 ·Sample size less than 5. Since RR=P,/P2 , RR~ 1/P2 • For RR< 1, use the column value corresponding to 1/ RR and the row value corresponding to P,. . 27 .21 26 1$ - • 13 .JQ;. , II·'· I~ Table 8 ( continued) (e) Confidence level 90%, relative precision 10% ~RRI 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 48266 23890 15764 11701 9264 43391 21452 14139 10483 8289 3901 2438 1707 1268 976 767 610 488 391 311 244 140 61 40141 19827 13056 9670 7639 3676 2221 37819 18666 12282 9090 7174 33442067 1428 1045 790 607 471 364 279 209 36078 17796 11701 8654 6826 3169 1961 1341 976 732 558 427 326 244 34724 17118 11250 8316 6555 3034 33640 16577 10889 8045 6338 2926 32754 16133 10593 7823 6161 2837 1729 1175 843 621 463 32015 15764 10347 7639 6013 2763 31390 15452 10139 7482 5888 2701 1a1 110:7 788 576 30855 15184 9960 7348 5781 2647 1602. 3.75 4.00 4.25 4.50 4.75 5.00 29625 14569 9550 7041 5535 29307 14410 9444 6961 5472 24'2 29022 14267 9349 6890 5415 2464. 28765 14139 9264 6826 5363 P2 01 0> 0.01 0.02 0.03 0.04 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.70 0.80 0.90 4388 2763 1961 1463 1138 906 732 596 488 399 326 209 122 1544 1138 867 674 529 416 326 252 190 ,. 1274 921 &87 519 393 1788 1219 878 &51 488 366 65 Since RR=P,/P2' RR~1/P2' For RR< 1, use the column value corresponding to 1/ RR and the row value corresponding to P,. 1660 1138 813 59& 1610: 7&7 30390 29984 14952 14748 9670 9805 7131 7232 5688 5607 2601 2160 1571 . 1544 1f.)$? 1637 732. 7. 2524 1520 10U. '.9· 1003 .,. '... 2411 9~ tn III 3 'C ii' UI j;j' CD .. ..o· Q. CD CD ~ 3 5i' III :::s Table 8 ( continued) (f) Confidence level 90%. relative precision 20% ~RRI 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 0.01 0.02 0.03 0.04 0.05 10761 5326 3515 2609 2066 9674 4783 3153 2337 1848 8949 4421 2911 2156 1703 8432 4162 2738 2027 1600 8044 3968 2609 1930 1522 7742 3817 2508 1854 1462 7500 3696 2428 1794 1413 7303 3597 2362 1744 1374 7138 3515 2307 1703 1341 6999 3445 2261 1668 1313 6879 3385 2221 1639 1289 6776 3334 2186 1613 1269 6685 3288 2156 1590 1250 6605 3248 2130 1570 1234 0.10 0.15 870 798 496 677 416 284 616 375 254 182 133 591 563 346 339 335 247 176 129 241 171 580 351 236 167 571 358 231 164 227 224 221 153 633 386 262 188 139 366 164 653 399 272 196 145 602 254 194 707 436 299 218 6413 3153 2066 1522 1196 544- 381 283 218 746 461 319 233 176 6470 3181 2085 1536 1208 550 330 0.25 0.30 979 616 435 327 254 6534 3213 2106 1552 1220 656 0.35 0.40 0.45 0.50 0.55 202 164 133 109 89 171 136 109 87 70 151 118 93 73 56 136 105 82 63 47 125 96 73 55 116 88 109 82 104 73 55 32 14 43 P2 c..n -...J 0.20 0.60 0.70 0.80 0.90 47 28 544- 345 206 13 Since RR= P,/ P2 • RR~ 1/ P2 • For RR< 1. use the column value corresponding to 1/ RR and the row value corresponding to P,. 327 218 I~ Table 8 ( continued) (g) Confidence level 90%, relative precision 25% ,\RRI 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 0.01 0.02 0.03 0.04 0.05 6474 3205 2115 1570 1243 5821 2878 1897 1406 1112 5073 2504 1648 1220 963 4840 2387 1570 1161 916 4658 2297 1509 1116 880 4513 2224 1461 1079 851 4394 2164 1421 1050 827 4295 2115 1388 1025 807 4211 2073 1360 1004 790 4139 2037 1336 986 776 4077 2006 1316 971 763 4022 1979 1297 957 753 3974 1955 1281 945 743 3931 1933 1267 934 734 3893 1914 1254 925 727 3859 1897 1243 916 720 0.10 0.15 0..20 0.25 0.30 589 298 208 153 117 449 218 192 141 106 426 262 180 131 407 250 171 124 93 393 240 164 381 232 158 113 84 371 226 153 109 80 363 220 149 355 215 145 103 349 211 142 101 344 208 139 99 339 204 137 335 202 135 331 199 133 327 197 131 197 153 524 327 229 171 131 5385 2660 1752 1297 1025 480 0.35 0.40 0.45 0.50 0.55 122 99 80 66 54 44 103 82 66 53 42 33 91 71 56 44 34 82 64 49 38 29 70 53 66 50 29 17 8 19 9 P2 (J1 00 0.60 0.10 0.80 0.90 371 262 99 75 58 44 33 118 88 62 26 Since RR= P,/ P2 , RR";; 1/ P2 . For RR< 1, use the column value corresponding to 1/ RR and the row value corresponding to P,. 106 78 en DI 3 'C CD 1/1 j;j' CD ... ..o· Do CD CD 3 S· DI = Table 9. Hypothesis tests for a relative risk n= {zl-a/zJ[2P(l- P)] +Zl-pJ[P l (l-Pd+Pz(l-Pz)]Y /(P 1 -Pzf where P=(P 1 +P z )/2 (a) Level of significance 5%, power 90%, two-sided test RRa) 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 37411 18492 12185 9032 7140 3357 2095 1465 1086 834 654 519 414 329 261 203 113 46 10378 5122 3371 2495 1969 918 568 393 287 217 167 130 101 5066 2497 1641 1212 955 442 270 185 133 99 75 56 42 30 21 3104 1528 1002 739 582 266 161 109 2149 1056 692 509 400 182 109 72 50 36 25 17 1605 787 515 379 297 133 79 52 35 24 16 11 1261 618 403 296 232 103 60 39 26 17 11 1028 503 328 240 188 82 47 30 19 12 862 421 274 200 156 68 38 24 15 9 738 360 234 171 133 57 32 19 12 643 313 203 148 115 568 276 178 130 101 42 23 13 .7 507 246 159 115 89 37 457 221 143 103 80 416 201 129 93 381 184 118 85 65 33 29 26 19 11 11 1i 7 13 P2 "" ~I 0.01 0.02 0.03 0.04 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.70 0.80 77 58 42 77 56 41 30 21 14 Since RRa=P,/P2' RRa ~1/P2· For RRa < 1. use the column value corresponding to 1/ RRa and the row value corresponding to P" 4$ 27 16 9 & 72 6 I:3 'C ii' 1/1 N' CD .... ..S' Q. CD CD 3 III 0' ;:, Table 9 ( continued) (b) Level of significance 5%, power 80%, two-sided test ""'-,RRal P2 0)1 0.01 0.02 0.03 0.04 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.70 0.80 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 27946 13814 9103 6747 5334 2508 1566 1095 812 623 489 388 309 247 195 152 85 35 7752 3827 2518 1864 1471 686 425 294 215 163 125 97 76 58 44 32 3785 1866 1226 906 714 330 202 138 100 74 56 42 32 23 16 2319 1142 749 553 435 200 121 82 58 42 31 23 16 11 1606 789 517 381 299 136 82 1199 589 385 283 222 100 59 39 27 19 13 8 943 462 302 222 174 769 376 245 180 141 62 36 23 15 10 644 315 205 150 117 51 29 18 12 7 552 269 175 128 100 481 234 152 111 86 37 20 12 7 425 207 134 97 76 32 17 10 6 379 184 119 87 67 28 15 8 342 166 107 78 60 25 13 7 311 151 97 70 54 22 11 6 285 138 89 64 49 20 10 5 54 38 27 19 14 77 45 29 20 13 9 Since RRa= P,/ P2' RRa~1/P2' For RRa< 1. use the column value corresponding to 1/ RRa and the row value corresponding to P,. 43 24 15 9 I~ Table 9 ( continued) (c) Level of significance 5%, power 50%, two-sided test ""!1 P2 Ol N 0.01 0.02 0.03 0.04 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.70 0.80 R a I 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 13675 6760 4455 3302 2611 1228 767 536 398 306 240 191 152 122 96 75 42 18 3794 1873 1233 913 721 337 209 102 66 49 38 17 11 9 168 82 53 39 30 13 7 153 75 48 35 27 12 6 140 68 44 32 25 11 6 8 5 7 6 186 91 59 43 34 15 8 5 28 22 16 12 9 462 227 148 109 86 39 23 15 10 7 5 236 115 75 55 43 19 37 588 289 189 139 110 50 30 20 14 10 7 5 271 133 86 63 50 22 13 62 49 38 29 22 787 387 254 187 147 67 41 27 19 14 10 316 155 101 74 58 26 15 10 7 81 1136 559 367 271 214 98 60 .41 29 22 16 12 9 6 377 185 121 89 70 31 145 106 1853 914 601 444 350 162 100 69 50 209 7 18 12 8 6 17 'Sample size less than 5. Since RRa=P,/P2• RRa~1/P2. For RRa < 1. use the column value corresponding to 1/ RRa and the row value corresponding to P,. * • en III 3 'C CD III jij' CD .... ..o· Q. CD CD 3 S· III :::I Table 10. Accepting a population prevalence as not exceeding a specified value The value of n is obtained by solution of the inequality x~o Cx d* M (N-M) C(n-x) I Cn<a N where M = N P, for a finite population; or Prob {d :::;;d*} < a i.e. d* I Prob(d)<a d=O or d* I nCdpd(l- p)"-d < a d=O I~... for an infinite population. ~I CD 10 (a) Confidence level 95%, d*=O 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.05 0.025 0.0125 100 200 1 000 2000 2500 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 4 4 4 4 4 5 5 5 5 5 6 6 6 6 6 9 9 9 9 9 13 13 14 14 14 25 27 29 29 29 45 51 57 58 58 82 90 112 115 116 5000 1Qooo 15000 20000 25000 2 2; 2 :3 3 3. 4 4 4 4 9 9 "9 9 9 14 14 14 14 14 29 29 29 29 4 5 5 5 5. 5 6 2 2 2 58 69 59 59 59 118 118 118 111 119 96 140 212 225 228 234 286 4 4 5 5 9 9 14 14 59 59 119 119 P N 50000 Infinite 2; 2 2 2 2 2 2 2 3. 3. 3 3 6 6 6 6 6 6 29 29 29 237 236 238 239 239 Table 10 ( continued) (b) Confidence level 95%, d* = 1 P 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.05 0.025 0.0125 3 3 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 4 4 4 5 5 5 5 5 5 5 5 5 5 5 5 6 6 6 6 6 6 6 6 6 6 6 6 8 8 8 8 8 8 8 8 8 8 8 8 10 10 10 10 10 10 10 10 10 10 10 10 14 14 14 14 14 14 14 14 14 14 14 14 20 21 22 22 22 22222222 2222 22 38 42 45 46 46 46 46 46 46 46 46 46 64 95 127 174 181 183 186 187 187 188 188 188 188 100 191 324 348 356 367 312 314 316 319 379 379 N 100 200 1 000 2000 2500 5000 10000 15000 20000 25000 50000 Infinite 77 90 92 92 93 83 93 93 93 94 94 ~I (I) 11/ 3 "CI i' III p:j' CD .... aCD CD I :::I~, (c) Confidence level 95%, d* = 2 P 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.05 0.025 0.0125 4 5 5 5 5 5 5 5 5 5 5 5 6 6 6 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 7 7 8 8 8 8 8 10 11 11 11 11 11 11 11 11 11 11 11 13 14 14 14 14 14 14 14 14 14 14 14 18 19 19 19 19 19 19 19 19 19 19 19 27 28 29 30 30 30 48 54 60 61 61 61 61 61 61 62 62 62 77 100 155 227 238 242 246 "246 100 200 417 455 467 N 100 200 1000 2000 2500 5000 10000 15000 2()000 25000 50000 Infinite 8 9 9 9 9 9 9 30 30 30 30 30 30 98 118 122 122 123 123 124 124 124 125 125 .,- ..,201 251 251 486 43 502 502 502 11/ Ig' Table 10 ( continued) (d) Confidence level 95%, d* = 3 P 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.05 0.025 0.0125 6 6 6 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 7 7 9 9 9 9 9 9 9 9 9 9 9 9 10 11 11 11 11 11 11 11 11 11 11 11 13 13 13 13 13 13 13 13 13 13 13 13 16 17 17 17 17 17 17 17 17 17 17 17 22 23 24 24 24 24 24 24 24 24 24 24 32 34 36 37 37 37 37 37 37 37 37 37 58 66 74 75 75 75 75 75 75 76 76 76 88 116 145 150 150 152 152 152 153 153 155 155 100 176 275 291 297 303 305 306 307 307 309 309 100 200 501 552 571 596 607 610 614 618 619 619 N 100 200 1 000 2000 2500 5000 10000 15000 20000 25000 50000 ~I Infinite Ie (e) Confidence level 95%, d* = 4 p 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.05 0.025 0.0125 7 7 7 7 7 7 7 7 7 7 7 7 8 9 10 10 10 10 10 10 10 10 10 10 10 10 12 13 13 13 13 13 13 13 13 13 13 13 15 16 16 16 16 16 16 16 16 16 16 16 19 20 20 21 21 21 21 21 21 21 21 21 26 27 28 28 28 28 28 28 28 28 28 28 38 41 43 43 43 43 44 66 77 87 88 89 89 89 95 132 170 176 177 179 180 180 180 181 181 181 100 191 321 342 349 357 361 361 362 363 363 364 100 200 578 643 669 701 715 720 724 728 728 730 N 100 200 1 000 2000 2500 5000 10000 15000 20000 25000 50000 Infinite 9 9 9 9 9 9 9 9 9 9 Ii... 44 44 44 44 44 89 89 90 90 90 Table 10 ( continued) (f) Confidence level 90%, d* = 0 P 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.05 0.025 0.0125 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 4 4 4 4 4 5 5 5 5 5 7 7 7 7 7 10 11 11 11 11 20 21 22 22 22 37 41 44 45 45 78 78 87 89 90 94 120 168 175 177 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 4 4 4 4 4 5 5 5 5 5 7 7 7 7 7 11 11 11 11 11 22 22 2 3 3 3 3 3 22 23 45 45 45 45 45 91 91 91 91 92 181 182 182 184 184 2 2 2 2 2 2 3 3 4 4 5 5 7 7 11 11 23 23 46 46 92 92 184 184 N 100 200 1 000 2000 2500 5000 10000 15000 20000 25000 50000 Infinite 22 (J) (J) en QI 3 "C iD 1/1 j;j' ~ .... CI. CD CD I~, (g) Confidence level 90%, d* = 1 :::I 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.05 0.025 0.0125 100 200 1 000 2 000 2 500 3 3 3 3 3 4 4 4 4 4 4 4 5 5 5 5 5 7 7 7 7 7 8 9 9 9 9 11 12 12 12 12 17 18 18 18 18 32 35 37 38 38 56 65 74 76 76 93 112 145 149 151 100 188 274 290 296 5000 10000 15000 20000 25000 . 3 3 3 3 3 4 4 4 4 4 4 4 5 5 5 5 5 7 7 7 7 7 9 9 9 9 9 12 12 12 12 12 18 19 19 19 19 38 38 38 38 38 76 76 76 76 77 153 154 154 154 155 303 306 308 311 311 50000 Infinite 3 3 5 5 7 7 9 9 12 12 19 19 38 38 77 77 155 155 311 311 N ",P I 4 4 4 4 4 4 4 4 4 4 It Table 10 ( continued) (h) Confidence level 90%, d* =2 0.90 0.80 0.70 0.60 0.50 OAO 0.30 0.20 0.10 0.05 0.025 0.0125 100 200 1000 2000 2500 4 4 4 4 4 5 5 5 5 5 6 6 6 6 6 7 7 7 7 7 9 9 9 9 9 12 12 12 12 12 16 16 16 16 16 23 24 25 25 25 43 47 51 52 52 71 86 101 104 104 100 141 195 203 206 100 199 366 391 401 5000 10000 15000 20000 25000 4 4 4 4 4 5 5 5 5 5 6 6 6 6 6 7 9 7 9 7 7 8 9 9 9 12 12 12 12 12 16 16 16 16 17 25 25 25 25 25 52 52 52 52 52 105 105 105 105 105 209 ·210 211 211 212 414 418 42'0 426 427 50000 4 4 5 5 6 6 8 8 9 9 12 12 17 17 25 25 52 52 106 106 212 212 427 427 P N ~l Infinite I (i) Confidence level 90%, d* = 3 0.60 0.50 0.40 0.30 0.20 0.10 0.05 0.025 0.0125 8 8 8 8 8 9 9 9 9 9 11 12 12 12 12 14 15 15 15 15 19 20 21 21 21 29 30 32 32 32 52 58 64 65 65 82 104 126 130 130 100 164 241 253 258 100 200 449 484 500 6 6 6 6 6 8 8 8 8 8 9 12 12 12 12 12 15 15 15 15 15 21 21 21 21 21 32 32 32 32 32 65 65 65 65 66 131 132 132 132 132 262 10 10 10 10 265 265 267 518 526 527 531 535 7 7 8 8 10 10 12 12 15 15 21 21 32 32 66 66 135 135 267 267 535 535 0.90 0.80 0.70 100 200 1000 2000 2500 5 5 5 5 5 6 6 6 6 6 5000 10000 15000 20000 25000 5 5 5 5 5 50000 5 5 P N Infinite Ii... 264 Q Table 10 ( continued) ( i) Confidence level 90%, d* = 4 ~I ~I 0.90 0.80 100 200 1 000 2000 2500 7 7 7 7 7 8 8 8 8 8 5000 10000 1&cOOO JOcOOO 25cOOO 7 7 7 7 7 8 8 50000 7 7 Infinite 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.05 0.025 0.0125 9 9 9 9 9 11 11 11 11 11 14 14 14 14 14 17 18 18 18 18 23 24 25 25 25 34 36 38 38 38 60 69 77 78 78 90 121 150 155 156 100 180 285 302 308 100 200 527 572 595 11 12 12 12 12 14 14 14 14 14 18 18 18 18 25 25 38 25 79 157 158 168 168 159 314 316 316 317 318 619 628 628 18 38 38 38 38 78 78 78 8 9 9 10 10 10 8 8 10 10 12 12 14 14 18 18 25 25 39 39 79 79 159 159 318 318 I:3 "C 8 8 25 25 78 637 637 637 638 CD en iii' CD CI. CD CD ... ..c)' 3 S' DI ~ I~_I"" .,. _ ...................................... • .... J .............. ~ ............. n= [Zl-aJ {P o(1- Po)} +Zl-pJ {P a(1- P a n]2 /(P o - Pa)2 d* = [nPO-z1-aJ {nP o(1- PO)}] The value of d* is always rounded down to the nearest integer (for example 5.8 would become 5). (a) Level of significance 5%, power 90%, one-sided test (Pa< Po) \Pol Pa 0.10 ~ •. 0.05 0.10 0.15 0.20 .; 0.25 0.20 0.15 0.30 0.35 0.45 0.40 0.50 d* n d* n d* n 18 33 66 161 2 5 13 38 13 22 39 75 1 4 8 19 10 16 26 44 1 3 6 12 8 12 19 29 1 2 4 8 6 10 14 20 0 2 3 6 686 1. 180 753 25 242 4S 81. 206 11 31 80 312 31 10 10 11 38, n d* n d* n d* n d* n 239 16 76 378 6 45 40 109 498 3 14 84 25 54 137 601 2 8 25 132 F: 62 till ,. 13 ee 218 804 ..",' d* 137 Oi. d* n • ' 211 813 t3 402 , - - - - , - _ . _ - I -I III ffi I \pJ Pa 0.55 ~ n 0.05 0.10 0.15 0.20 5 8 11 15 d* 0 2 3 5 d* n 1,. 'i\".a'• • " ,22 d* n t 6 8 11 0.70 0.65 0.60 5 7 9 1 2 3 I 12 1:1 U I t ~,2I.: ,'1. I:~ 21 It II 7 11 18 48 11 27 852 444 210 834 114 477 91 203 798 51 120 496 0.50 0.55 0.60 0.65 0.70 I'" t Sample size less than 5. 1. d* n t 1 2 3 0.75 d* n 5 7 1 2 5 2 I t2 1• 22 3.1 4 7 I: I I 3 4 17 49 87 191 746 29 53 123 501 I n , 1, 15 21 6 8 11 12 14 30 46 80 176 676 18 29 53 122 488 19 27 42 72 156 • d* n t t t t t 0.85 0.80 t t t t , . a 3. &: 4 12 18 29 52 116 d* n .I 0.90 d* n t t t t t t t t t t t t I 10 I 6 7 13 17 24 36 62 8 12 17 27 48 9 11 14 20 30 6 8 10 15 24 131 104 101 362 18 ---- t t ,I. I. 6 I • d* n t t t t .I I~ 0.95 t 4 4 -to 311 t 5 7 8 11 15 3 5 6 9 12 21 11 3D I» M 17 221 2t4 ( continued) Table 11 (b) Level of significance 5%, power 80%, one-sided test (Pa < Po) 0.10 Pol 0.15 0.20 0.30 0.25 0.35 0.40 0.45 0.50 Pa 0.05 0.10 0.15 0.20 0,2$ 0.30 0.35 0.40 0.45 n d* n d* n d* n d* n d* n d* n d* n d* n 184 11 60 283 4 32 32 83 368 2 10 60 21 42 103 441 1 5 18 95 15 26 50 119 501 1 3 9 27 133 11 18 30 56 133 1 2 6 13 37 173 8 13 20 33 61 142 683 0 2 4 8 18 47 213 7 10 14 22 35 65 149 0 1 3 6 10 22 57 252 5 8 11 15 23 37 67 548 606 d* 0 1 2 4 7 13 26 66 288 163 617 len I» 3 ~ CD 251 ~ 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.60 0.55 0.75 0.70 0.65 0.80 0.85 0.90 1/1 i:r 0.95 CD ... Q. n d* 5 6 8 11 16 24 38 68 154 615 0 1 2 3 5 9 15 30 74 317 t Sample size less than d* n 1 2 2 4 6 10 17 33 80 340 5 7 9 12 16 23 37 65 145 573 1 2 3 5 7 11 19 35 84 353 d* n t t t 5 7 9 12 16 24 38 67 151 600 d* n 1 2 4 5 8 11 20 37 86 356 5 6 8 11 15 21 32 57 125 483 d* n t t t t t t t 5 7 9 11 16 22 35 62 136 534 d* n t 5 6 8 .10 13 19 29 50 109 419 2 3 4 5 8 12 19 35 80 321 2 4 5 8 11 18 32 71 279 5 6 7 10 13 20 33 69 253 t t t t 3 4 4 7 9 15 26 68 219 t t t t t t t 5 7 9 14 22 44 CD CD 3 If t t t t 150 5. d* n t t t t t t t 5 7 9 12 16 25 43 91 342 d* n t t t t t t t t 2 2 3 5 8 12 19 37 85 346 d* n 3 5 7 11 19 39 138 ( continued) Table 11 (c) Level of significance 5%, power 50%, one-sided test (Pa < Po) " pJ Pa ~ 0.05 0.10 0.15 0.20 0.10 0.15 0.25 0.20 0.30 0.35 0.40 n d* n d* n d* n d* n d* n 98 4 35 138 13 20 44 174 4 26 13 23 51 203 0 2 7 40 10 15 26 57 0 1 3 11 7 10 16 28 0 1 2 5 6 8 11 17 0 0 1 3 228 57 62 247 15 74 29 65 260 7 19 91 0.25 0.30 0.35 0.40 0.45 d* 0.45 d* n 0.50 d* n 5 6 8 11 0 0 1 2 17 4 9 23 107 30 67 268 d* n t 5 6 8 0 0 1 11 17 2 5 10 27 121 31 68 271 tSample size less than 5. -...II Pa ~I 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 I~....... CD d* n d* n t t t 5 6 0 1 5 1 0.25 0.30 0.35 0.40 0.45 8 11 17 30 67 2 3 5 12 30 6 8 11 17 29 1 2 3 13 0.50 0.55 0.60 0.65 0.70 268 134 65 260 32 143 t 0.75 0.80 0.85 0.90 tSample size less than 5. 6 6 7 10 16 1 2 4 7 28 62 247 14 34 148 5 7 10 15 26 57 228 1 2 4 7 14 34 148 5 6 9 13 23 51 203 4 7 13 33 142 t t t t 2 3 6 13 30 174 130 t t t t t t t t t t t t t t t t t t t t t t t t t t t 6 9 16 3 5 11 35 138 26 110 t t 7 11 25 98 d* n t t t t 5 7 11 20 44 d* n t t t t t t 2 2 d* n t t t t t t t t d* n t t t t t t t d* n t t t t t t d* n t 0.05 0.10 0.15 0.20 t d* n 4 8 20 83 6 . 13 52 t 4 11 46 Sample size determination Table 12. Estimating an incidence rate with specified relative precision n = (Zl_~/2/f.)2 Confidence level E 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 99% 95% 90% 66358 16590 7374 4148 2655 1844 1355 1031 820 664 461 339 260 205 166 138 38417 9605 4269 2402 1537 1068 785 601 475 385 267 197 151 119 97 80 67 57 50 43 38 34 30 27 25 22 20 19 11 16 27061 6766 3007 1692 1083 152 553 423 335 211 188 139 106 84 68 56 41 41 35 116 99 85 74 65 58 52 46 42 38 35 32 29 27 72 31 27 24 21 19 17 16 14 13 12 11 Table 13. Hypothesis tests for an incidence rate F or a one-sided test n =(Zl-aAO + ZI_pAa)2 /{}-o - )~a)2. For a two-sided test n = (Zl-a/2AO + Zl-pAaf /(A o - Aa)2. (a) Level of significance 5%, power 90%, one-sided test ,d 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 21 10 57 7 21 109 6 13 37 179 5 10 21 57 266 5 8 15 31 81 369 5 7 12 21 43 109 490 5 7 10 16 29 57 142 629 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 6 9 13 21 37 6 8 12 17 27 46 90 220 956 5 7 10 14 21 33 57 109 266 1146 5 7 9 13 18 26 41 68 130 315 1352 5 6 8 11 15 21 31 48 81 154 369 1576 5 6 8 10 13 18 25 37 57 94 179 428 1817 5 6 7 9 12 16 21 30 43 66 109 206 490 2075 5 6 7 9 11 14 19 25 34 50 76 125 235 557 2351 5 5 7 8 10 13 16 21 29 39 57 86 142 266 629 2643 0.95 Aa -...J w 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 18 8 6 5 * • 51 18 11 8 7 6 5 5 102 33 18 13 10 8 7 6 6 5 5 5 * * . . . • Sample size less than 5. * . . 169 51 27 18 14 11 9 8 7 7 6 6 5 5 5 254 74 38 25 18 14 12 10 9 8 7 7 6 6 356 102 51 33 23 18 15 13 11 10 9 8 7 474 133 66 42 29 23 18 15 13 12 10 9 610 169 83 51 36 27 22 18 16 14 12 72 179 784 764 209 102 62 43 33 26 21 18 16 934 254 122 74 51 38 30 25 21 1121 303 145 88 60 45 35 29 1326 356 169 102 70 51 40 1548 413 196 117 80 59 1786 474 224 133 90 2042 540 254 151 2315 610 286 2606 685 2913 ~ 5 6 8 9 12 15 19 24 33 45 64 97 160 298 704 2952 CT CD .... Co) Table 13 ( continued) (b) Level of significance 5%, power 80%, one-sided test Ao 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 18 9 44 7 18 83 6 11 29 135 5 9 18 44 199 5 7 13 25 62 5 7 10 18 34 6 9 14 23 6 8 11 18 5 7 10 14 5 7 9 12 5 6 8 11 5 6 7 10 5 6 7 9 5 5 7 8 5 6 8 5 6 7 5 6 7 275 83 44 108 29 56 135 578 22 36 69 165 704 18 27 15 21 32 53 99 13 9 11 44 20 28 14 18 23 8 10 12 16 62 10 13 18 24 34 10 12 15 38 11 15 21 29 116 275 1155 72 135 318 1330 51 83 155 364 1518 . 39 58 95 176 412 31 44 66 108 199 26 35 50 75 121 465 223 Aa 0.05 0.10 0.15 0.20 0.25 0.30 0.36 ~I 0.40 0.45 0.50 12 5 . . " . " 34 12 7 5 . " 69 21 12 8 6 5 " 0.55 0.60 0.65 0.70 0.75 0•• 0,85 6,;90' " " " 117 34 18 12 9 7 177 6 5 " " " • Sample size less than 5. " " 50 249 25 16 12 69 364 465 34 21 334 92 45 431 117 540 9 7 6 5 5 15 12 9 8 7 27 19 15 12 10 56 34 24 18 14 145 69 42 29 21 . 6 5 5 8 7 6 662 177 84 50 34 44 83 199 842 796 211 100 60 12 17 25 40 10 14 20 9 12 16 30 23 942 249 117 69 47 34 235 992 1101 290 136 80 64 18 25 1272 334 155 92 1456 381 177 171' 1652 431 1930 1860 .It - 20 520 2154 ... en III 3 "CD III j:r CD ... ..cr Q- CD CD 3 :i" III := Table 13 ( continued) (c) Level of significance 5%, power 90%, two-sided test \ AO I 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 28 13 10 28 137 8 17 47 223 7 13 28 7 11 20 40 102 459 6 10 16 28 55 137 608 6 9 13 21 37 72 178 779 6 8 12 17 28 47 91 223 970 6 8 11 15 22 35 59 113 274 1182 6 7 10 13 19 28 43 72 137 331 1416 5 7 9 12 16 23 33 52 86 163 392 1670 5 7 9 11 15 20 28 40 61 102 192 459 1946 5 7 8 10 13 17 24 33 47 72 119 223 531 2242 5 6 8 10 12 16 21 28 38 55 83 137 257 608 2560 5 6 8 9 11 14 18 24 32 44 63 95 157 293 691 2898 5 6 7 9 11 13 5 6 7 8 10 12 15 19 24 31 42 57 81 122 200 371 872 3639 Aa ~l 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 21 9 6 5 72 61 21 12 9 7 6 5 5 5 • Sample size less than 5. 122 38 21 14 11 9 8 7 6 6 5 5 5 204 61 32 21 16 12 10 9 8 7 7 6 6 5 5 72 331 306 89 45 29 21 16 13 11 10 9 8 7 7 6 430 122 61 38 27 21 17 14 12 11 10 9 8 575 160 79 49 34 26 21 17 15 13 12 11 741 204 99 61 42 32 25 21 18 16 14 928 252 122 74 51 38 30 25 21 18 1136 306 147 89 61 45 35 29 24 1365 366 174 104 71 53 41 33 1615 430 204 122 83 61 47 1886 500 236 140 95 69 2179 575 270 160 108 2492 656 306 181 2826 741 345 3181 832 17 21 28 37 50 72 108 178 331 779 3258 3557 ~ DI C' iD .... c..I Table 13 ( continued) (d) Level of significance 5%, power 80%, two-sided test \ ,d 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 23 12 58 9 23 108 8 15 38 174 7 12 23 58 256 6 10 17 33 81 353 6 9 14 23 44 108 466 6 8 12 18 30 58 139 595 6 8 10 15 23 38 73 174 739 6 7 10 13 19 29 48 69 213 899 5 7 9 12 16 23 35 56 5 7 8 11 14 20 28 42 69 128 302 1267 5 6 8 10 13 17 23 33 50 81 150 353 1474 5 6 8 9 12 15 20 27 38 58 94 174 407 1697 5 6 7 9 11 14 18 23 31 44 67 108 199 466 1936 5 6 7 8 10 13 16 20 27 36 51 76 123 227 528 5 6 7 8 10 12 16 16 23 30 41 58 86 139 256 595 5 6 7 8 9 Aa 0.05 0.10 0.15 0.20 0.25 0.30 ~I 0.36 0.40 14 6 • 42 14 8 6 0.46 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.65 0.90 0.95 86 26 14 9 7 6 5 * • * • * • Sample size less than 5. * 146 42 21 14 10 8 6 6 5 221 62 31 19 14 11 9 7 6 6 5 5 • 312 66 42 26 18 14 11 9 8 7 6 6 5 419 114 65 34 23 17 14 11 10 8 7 7 541 146 70 42 29 21 17 14 12 10 9 680 181 86 52 35 26 20 16 14 12 834 221 104 62 42 31 24 19 16 1003 265 124 74 50 36 28 23 108 256 1075 1188 312 146 86 58 42 32 1389 364 169 99 67 48 1606 419 194 114 76 1839 478 221 129 2190 2087 541 250 2460 2350 609 2630 (I) III 3 'C CD 11 III 13 16 21 26 34 46 65 97 156 286 665 2746 . j;j' CD a.. CD CD ~ 3 ..o· :I III :I Table 14. Hypothesis tests for two incidence rates in follow-up (cohort) studies (study duration not fixed) F or a one-sided test n 1 = {Zl-,j[(1 + k)I2] + Z l-pj(U i + ;"~)}2/kVl - )"2f· For a two-sided test n 1 = {Zl-,/2j[(1 +k)I2]+zl-pj(Ui+}"nV/ k()'1-A 2f where A=(A 1 +)"2)/2 and k is the ratio of the sample size for the second group of subjects (n 2 ) to that for the first group (n 1 ). For Tables 14 a-d, k= 1. (a) Level of significance 5%, power 90%, one-sided test A, -....I -....I .le2 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 41 19 109 14 41 212 12 26 71 349 11 19 41 109 521 10 16 29 60 157 726 9 14 23 41 83 212 966 9 13 19 31 55 109 277 1240 9 12 8 11 15 22 33 52 89 174 431 1891 8 11 14 19 28 41 64 109 212 521 2268 8 10 13 18 24 34 50 78 132 254 619 2680 8 10 13 16 21 29 41 8 10 12 15 19 26 35 49 71 109 183 349 842 3605 8 9 11 14 18 23 30 41 57 83 127 212 403 966 4119 8 9 11 13 8 9 11 13 16 19 24 31 41 55 76 109 167 277 521 1240 5250 8 9 10 12 15 18 22 28 36 47 63 86 124 189 312 585 1390 6867 '\ 0.05 0.10 0.15 0.20 0.25 0.30 0.36 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.86 0.90 0.95 41 19 14 12 11 10 9 9 9 8 8 8 8 8 8 8 8 8 109 41 26 19 16 14 13 12 11 11 10 10 10 9 9 9 9 212 71 41 29 23 19 17 15 14 13 13 12 11 11 '1 10 349 109 60 41 31 26 22 19 18 16 15 14 13 13 12 521 157 83 55 41 33 28 24 21 19 18 17 16 15 726 212 109 71 52 41 34 29 26 23 21 19 18 966 277 140 89 64 50 41 35 30 27 24 22 1240 349 174 109 78 60 49 41 36 31 28 1549 431 212 132 93 71 57 48 41 36 17 26 41 71 140 349 1549 1891 521 254 157 109 83 66 55 47 2268 619 300 183 127 96 76 63 2680 726 349 212 146 109 86 3125 842 403 243 167 124 60 93 157 300 726 3125 3605 966 460 277 189 4119 1099 521 312 4667 1240 585 17 21 27 35 48 66 96 146 243 460 1099 4667 5250 1390 5867 Ii.01:0 Table 14 ( continued) (b) Level of significance 5%, power 80%, one-sided test ).,1 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 29 14 79 10 29 153 8 18 51 252 8 14 29 79 376 7 12 21 43 113 524 7 10 16 29 60 153 697 6 9 14 22 39 79 199 895 6 8 12 18 29 51 101 252 1118 6 8 11 16 24 37 64 126 311 1365 6 8 10 14 20 29 46 79 153 376 1638 6 7 9 13 17 24 36 56 95 183 447 1934 0.70 0.75 0.80 0.85 0.90 0.95 6 5 7 8 11 14 18 25 35 51 79 132 252 608 2602 5 7 8 10 13 16 22 29 41 60 92 153 291 697 2974 5 6 8 10 12 15 19 25 34 48 69 106 176 332 793 3369 5 6 8 9 11 14 5 6 7 9 11 13 16 20 26 34 45 62 90 136 225 422 1004 4235 )'2 ~I 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 a.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 29 14 10 8 8 7 7 6 6 6 6 6 6 5 5 5 5 5 79 29 18 14 12 10 9 8 8 8 7 7 7 7 6 6 6 153 51 29 21 16 14 12 11 10 9 9 8 8 8 8 7 252 79 43 29 22 18 16 14 13 12 11 10 10 9 9 376 113 60 39 29 24 20 17 15 14 13 12 11 11 524 153 79 51 37 29 24 21 18 16 15 14 13 697 199 101 64 46 36 29 25 22 19 17 16 895 252 126 79 56 43 35 29 25 22 20 1118 311 153 95 67 51 41 34 29 26 1365 376 183 113 79 60 48 39 34 1638 447 216 132 92 69 55 45 1934 524 252 153 106 79 62 2256 608 291 176 120 90 7 9 12 15 21 29 43 67 113 216 524 2256 2602 697 332 199 136 2974 793 376 225 3369 895 422 3790 1004 17 22 29 39 55 79 120 199 376 895 3790 4235 (I) al 3 'C ii' III j;j' ID ... ..o· 0- lD ID 3 5' al := Table 14 ( continued) (c) Level of significance 5%, power 90%, two-sided test \ . 11.2 U3i \ ,1,1 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.46 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0;85 0.90 0.95 0.05 50 24 17 14 13 12 11 11 10 10 10 10 9 9 9 9 9 9 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 50 24 134 17 50 260 14 31 87 428 13 24 50 134 638 12 20 35 73 192 891 11 17 28 50 101 260 1185 11 15 24 38 67 134 339 1521 10 14 21 31 50 87 171 428 1900 10 13 19 27 40 63 109 213 528 2320 10 13 17 24 34 50 78 134 260 638 2783 10 12 16 21 29 41 61 9 12 15 20 26 35 50 73 114 192 368 891 3834 9 11 14 18 24 31 42 69 87 134 225 428 1033 4422 9 11 14 17 22 28 37 50 70 101 156 260 494 1185 5053 9 11 13 16 20 25 33 43 58 81 117 179 298 564 1348 6726 9 11 13 15 19 24 30 38 50 67 93 134 205 339 638 1521 6440 9 10 12 15 18 22 27 34 44 57 76 106 152 231 382 718 1705 7197 134 50 31 24 20 17 15 14 13 13 12 12 11 11 11 11 10 260 87 50 35 28 24 21 19 17 16 15 14 14 13 13 12 428 134 73 50 38 31 27 24 21 20 18 17 16 15 15 638 192 101 67 50 40 34 29 26 24 22 20 19 18 891 260 134 87 63 50 41 35 31 28 26 24 22 1185 339 171 109 78 61 50 42 37 33 30 27 1521 428 213 134 95 73 59 50 43 38 34 1900 528 260 162 114 87 70 58 50 44 2320 638 311 192 134 101 81 67 57 2783 759 368 225 166 117 93 76 3287 891 428 260 179 134 106 95 162 311 759 3287 3834 1033 494 298 205 152 4422 1185 564 339 231 6053 1348 638 382 5726 1521 718 6440 1705 7197 -I III r:T ....ca CD Table 14 ( continued) (d) Level of significance 5%, power 80%, two-sided test A,I 0.05 0.10 0.15 0.20 0.25 17 13 37 194 10 23 64 320 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 9 9 14 26 54 143 665 8 13 21 37 75 194 885 8 11 17 28 50 100 253 1136 7 10 15 23 37 64 128 320 1419 7 10 14 20 30 47 81 159 394 1733 7 9 13 17 25 37 58 100 194 477 2078 7 9 12 16 22 31 45 7 9 11 14 19 26 37 54 85 143 274 665 2863 7 8 10 13 17 23 31 7 8 10 13 16 21 27 37 52 75 116 194 369 885 3774 7 8 10 12 15 19 24 32 43 60 87 134 222 421 1007 4277 6 8 9 11 14 17 22 28 37 50 69 100 153 253 477 1136 4811 6 8 9 11 13 16 20 25 32 42 57 79 113 173 285 536 1274 5376 A2 ~I 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 37 37 100 13 10 100 37 23 9 17 17 9 8 8 7 7 7 7 7 7 7 7 6 6 14 13 11 10 10 9 9 9 8 8 8 8 8 194 64 37 26 21 17 15 14 13 12 11 10 10 10 9 9 320 100 54 37 28 23 20 17 16 14 13 13 12 11 11 477 143 75 50 37 30 25 22 19 17 16 15 14 13 17 37 100 477 665 194 100 64 47 37 31 26 23 21 19 17 16 885 253 128 81 58 45 37 31 27 24 22 20 1136 320 159 100 71 54 44 37 32 28 25 1419 394 194 120 85 64 52 43 37 32 1733 477 232 143 100 75 60 50 42 2078 567 274 168 116 87 69 57 2455 665 320 194 134 100 79 71 120 232 567 2455 2863 771 369 222 153 113 3303 885 421 253 173 44 64 100 168 320 771 3303 3774 1007 477 285 4277 1136 536 4811 1274 5376 (I) III 3 'C iD 1/1 jij' CD .... . o· CI. CD CD ~. ::I III ::I
© Copyright 2024