Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Sample Size”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 37 records · Page 2Linked to original sources
PubMed1997

Exact equivalence test for risk ratio and its sample size determination under inverse sampling.

When data are dichotomous, this paper notes the utility of inverse sampling in establishing equivalence with respect to the risk ratio. This paper develops an exact equivalence test that accounts for the risk ratio under inverse sampling and further discusses the relationship between the exact equivalence test and the exact conditional confidence limits. Also included are an exact and two asymptotic procedures for calculation of the minimum required number of index subjects for a desired power 1--beta at a given alpha-level. Finally, this paper provides a table that summarizes the minimum required number of index subjects for powers equal to 0.90 and 0.80 in application of the proposed exact equivalence test at 0.05-level in a variety of situations.

Confidence Intervals↗
PubMedDate not supplied

Small sample size.

Explore the source record for details and available documents.

Sampling Studies↗
PubMed1991

Sample size and dimensionality in multivariate classification: implications for body surface potential mapping.

This paper presents empirically determined guidelines for specifying the number of features appropriate for multivariate classification studies for given sample sizes. Sample size was considered adequate if the mean distance between two sample sets, taken from the same continuous multivariate distribution and projected onto the best separating direction, remained below a prescribed level. To quantitate the sample size requirement, homogeneity of sample set pairs of equal size. N, taken from the same continuous multivariate distribution was studied as a function of dimensionality. M. Homogeneity was characterized by the maximum absolute distances (Dmax) between the corresponding pairs of empirical cumulative probability distributions on the best separating projection. Computer generated data sets were used to estimate the cumulative probability distribution, P(D)M.N, for sample sizes, N, ranging from 5 to 100 and the dimensionality, M, ranging from 1 to 4. An empirical relationship between the estimated step-polygons and the Kolmogorov type one dimensional limiting distribution L(z) has been established. Based on the sample size data of 34 key papers on clinical body surface potential mapping (BSPM) it is noted that in 30% of the cases only one, and in 6% of the cases only two parameters could be used for statistical group representation to ensure a reasonable reliability (Dmax less than 0.2). In 56% of the published cases the sample sizes could not guarantee this reliability even for one feature or parameter.

Computer Simulation↗
PubMed1995

Sample size determination in epidemiologic studies.

Sample size determination is an important issue in epidemiologic studies. Standard methods for determining sample size in cohort and case-control studies have generally been restricted to dichotomous disease and exposure variables and discrete confounding variables, and are based on simplifying assumptions that could often be unrealistic. Methods for sample size determination that make less restrictive and more realistic assumptions regarding the distribution of disease, exposure and confounding variables and which more closely parallel the analyses that are performed on the data, after the study has been conducted, have been developed in recent years. In this article some recent developments in the methodology for sample size determination in epidemiologic studies are reviewed.

Case-Control Studies↗
PubMed1989

Biostatistics in clinical trials: Part 2. Determining sample sizes for clinical trials.

The appropriate sample size for a clinical trial depends on the objectives of the trial. For Phase II trials, the recommended sample size is generally in the neighborhood of 25-40 patients, regardless of the choice of design. Sample sizes for Phase III trials are much more variable, ranging from less than a hundred to thousands of patients. The methodology of sample size determination for Phase III trials depends on the type of endpoint that is to be the primary focus of analysis. Tables of sample sizes for a variety of situations are presented. Use of sequential designs may reduce the required sample size by permitting early termination of trials whose results become definitive prior to completion of patient entry.

Clinical Trials as Topic↗
PubMed2002

Power and sample sizes for linkage with extreme sampling under an oligogenic model for quantitative traits.

Extreme sampling of sibling pairs has been shown to be efficient in terms of statistical power and sample sizes (in number of sibling pairs needed to genotype) to detect a quantitative trait locus (QTL) when the residual distribution is normal. In the present study, the efficiency of extreme sampling strategies to detect each locus under an oligogenic model is analytically explored with a test statistic based on identical-by-descent (IBD) statuses of independent sibling pairs. In the oligogenic model, the joint effect of oligogenes is the sum of the effects of each locus. Under this model, detecting each single locus will depend, in part, on the allele frequencies and magnitudes of effect of the other loci. Effects of two QTLs with different magnitudes of displacement and acting nonepistatically are considered. Three types of extreme sampling-that is, extreme concordant high (ECH), extreme concordant low (ECL), and extreme discordant (ED)-are primarily considered herein. Among these, ED sampling under the oligogenic model is shown to be most efficient in most situations considered here in terms of allele frequency and mode of inheritance. Differences in results between ECH and ECL sampling are purely arbitrary, brought up mostly by the directions of displacement effects. However, power to detect a locus with the lesser (in magnitude) displacement effect does not necessarily increase with extremity of sampling. Combinations of extreme discordant and extreme concordant sibling pairs are briefly discussed.

Alleles↗
PubMedDate not supplied

Sample size determinations using examples drawn from the NCTR Collaborative Behavioral Teratology Study data.

The Collaborative Behavioral Teratology Study (CBTS) introduced the coefficient of detection (CD) as an estimate of the minimum difference needed to detect a significant difference among groups. The CD was used to provide an index of the sensitivity of the tests used. As originally described, the CD is not suitable for estimates of two or more group minimally significant differences because it is based on a one-group formula. Also, the CD operates at a power of 50% in the one-group case (see preceding paper). These drawbacks leave questions about the sensitivity of the tests used in the CBTS unanswered. Using examples drawn from the CBTS data base, a more standard method of approaching this question has been taken here based on power calculations. Using the means and standard deviations from the vehicle controls, alpha = 0.05, 1-beta = 0.80, and 20, 30 or 40% group mean difference sizes, required samples sizes were determined for all the behavioral measures reported in the CBTS final report for Experiment 1. The results showed that most of the measures can detect 30% group mean differences in a two-group, two-tailed t-test situation with group sizes of less than or equal to 20 litters per group. More complex calculations are required when more than two groups are planned. Multigroup designs require additional assumption about the distribution of group differences and accordingly are more difficult to specify.(ABSTRACT TRUNCATED AT 250 WORDS)

Animals↗
PubMed2001

Effects of study duration, frequency of observation, and sample size on power in studies of group differences in polynomial change.

Consider a study in which 2 groups are followed over time to assess group differences in the average rate of change, rate of acceleration, or higher degree polynomial effect. In designing such a study, one must decide on the duration of the study, frequency of observation, and number of participants. The authors consider how these choices affect statistical power and show that power depends on a standardized effect size, the sample size, and a person-specific reliability coefficient. This reliability, in turn, depends on study duration and frequency. These relations enable researchers to weigh alternative designs with respect to feasibility and power. The authors illustrate the approach using data from published studies of antisocial thinking during adolescence and vocabulary growth during infancy.

Adolescent↗
PubMed1997

Planning significant and meaningful research in exercise science: estimating sample size.

Exercise science researchers are familiar with the use of parametric tests to detect significant differences among treatment groups. However, in planning research a question asked with increasing frequency is, "How many participants are needed to detect real and meaningful differences among groups?" In this paper, we provide an overview of the use of alpha, power, and effect size in planning sample sizes that allow tests of real and meaningful differences among groups. Because effect size is the parameter most often missing, we have located meta-analyses in sport and exercise psychology (n = 26), and motor behavior (n = 6). We provide examples and a discussion of how researchers can use these effect sizes along with common estimates of alpha and power to plan for the sample size needed to detect real and meaningful group differences.

Exercise↗
PubMed2001

On sample size calculation in bioequivalence trials.

Sample size calculation plays an important role in bioequivalence trials. In practice, a bioequivalence study is usually conducted under a crossover design or a parallel design with raw data or log-transformed data. In this paper, we discuss the differences in sample size calculation between a crossover design and a parallel design with raw data or log-transformed data. Formulas for sample size calculation under a crossover design and a parallel design with raw data or log-transformed data are derived. A brief discussion for the relationship among these formulas is given.

Clinical Trials as Topic↗
PubMed1998

A simple method of sample size calculation for linear and logistic regression.

A sample size calculation for logistic regression involves complicated formulae. This paper suggests use of sample size formulae for comparing means or for comparing proportions in order to calculate the required sample size for a simple logistic regression model. One can then adjust the required sample size for a multiple logistic regression model by a variance inflation factor. This method requires no assumption of low response probability in the logistic model as in a previous publication. One can similarly calculate the sample size for linear regression models. This paper also compares the accuracy of some existing sample-size software for logistic regression with computer power simulations. An example illustrates the methods.

Humans↗
PubMed1989

Determination of sample sizes for epidemiological surveys using cluster sampling technique.

Cluster sampling often provides a convenient and low cost device, in epidemiological surveys. The sample size needed under cluster sampling is generally larger than that in an individual based scheme due to the intra-class correlation existing in a cluster. This intra-class correlation coefficient is usually not known and some assumptions or estimates are essential. The strengths and weaknesses of cluster sampling over other sampling plans are presented and briefly discussed in this paper with particular reference to leprosy control programmes. One particular model of multistage cluster sampling technique is suggested in the evaluation of a District level programme, which includes determining the effectiveness of Multi-Drug Therapy, monitoring efficiency of paramedical workers and estimating the incidence of leprosy.

Catchment Area, Health↗
PubMed1995

Volume-weighted mean nuclear volume and nuclear area in advanced ovarian carcinoma. An investigation of sampling methods, sample size and reproducibility.

The influence of sampling issues on the reproducibility of volume-weighted mean nuclear volume (mean v) and mean nuclear area (MNA) assessments in patients with International Federation of Gynecology and Obstetrics stage III and IV ovarian carcinoma was evaluated. Ten cases representing the whole range of MNA values were selected from a population of 131 cases. The MNA and mean v of the same tumor cell nuclei were determined in one session by switching between the stereologic module and the morphometric module of the video overlay program used. For both MNA and mean v in one series of measurements, tumor nuclei were sampled from the whole tumor area and in a second series from the most poorly differentiated part (the measurement area) in each section, thus giving four series of measurements per case. For all four series, 500 nuclei were point sampled from approximately 100 systematically randomly selected fields of vision, using the automated scanning stage controlled by the morphometry program. These large samples, containing 500 nuclei for each case, were regarded as representative in each case. To investigate the susceptibility of MNA and mean v to variance at lower sampling levels (fields, nuclei), a nested analysis of variance was performed. Then the influence of sample size and sampling method was evaluated by drawing subsets from these 500 nuclei in each case in three different ways (cluster, systematic or random) with four different sample sizes (50, 100, 125, 250). It was shown that for MNA assessed in the measurement area, the variance between patients contributed the most to the total variance.(ABSTRACT TRUNCATED AT 250 WORDS)

Carcinoma↗
PubMed1999

Sample size estimation for the sorcerer's apprentice. Guide for the uninitiated and intimidated.

OBJECTIVE: To review the importance of and practical application of sample size determination for clinical studies in the primary care setting. QUALITY OF EVIDENCE: A MEDLINE search was performed from January 1966 to January 1998 using the MeSH headings and text words "sample size," "sample estimation," and "study design." Article references, medical statistics texts, and university colleagues were also consulted for recommended resources. Citations that offered a clear and simple approach to sample size estimation were accepted, specifically those related to statistical analyses commonly applied in primary care research. MAIN MESSAGE: The chance of committing an alpha statistical error, or finding that there is a difference between two groups when there really is none, is usually set at 5%. The probability of finding no difference between two groups, when, in actuality, there is a difference, is commonly accepted at 20%, and is called the beta error. The power of a study, usually set at 80% (i.e., 1 minus beta), defines the probability that a true difference will be observed between two groups. Using these parameters, we provide examples for estimating the required sample size for comparing two means (t test), comparing event rates between two groups, calculating an odds ratio or a correlation coefficient, or performing a meta-analysis. Estimation of sample size needed before initiation of a study enables statistical power to be maximized and bias minimized, increasing the validity of the study. CONCLUSION: Sample size estimation can be done by any novice researcher who wishes to maximize the quality of his or her study.

Humans↗
PubMed1990

Power and sample size calculations. A review and computer program.

Methods of sample size and power calculations are reviewed for the most common study designs. The sample size and power equations for these designs are shown to be special cases of two generic formulae for sample size and power calculations. A computer program is available that can be used for studies with dichotomous, continuous, or survival response measures. The alternative hypotheses of interest may be specified either in terms of differing response rates, means, or survival times, or in terms of relative risks or odds ratios. Studies with dichotomous or continuous outcomes may involve either a matched or independent study design. The program can determine the sample size needed to detect a specified alternative hypothesis with the required power, the power with which a specific alternative hypothesis can be detected with a given sample size, or the specific alternative hypotheses that can be detected with a given power and sample size. The program can generate help messages on request that facilitate the use of this software. It writes a log file of all calculated estimates and can produce an output file for plotting power curves. It is written in FORTRAN-77 and is in the public domain.

Case-Control Studies↗
PubMed2000

Interim analysis and sample size reassessment.

This article deals with sample size reassessment for adaptive two-stage designs based on conditional power arguments utilizing the variability observed at the first stage. Fisher's product test for the p-values from the disjoint samples at the two stages is considered in detail for the comparison of the means of two normal populations. We show that stopping rules allowing for the early acceptance of the null hypothesis that are optimal with respect to the average sample size may lead to a severe decrease of the overall power if the sample size is a priori underestimated. This problem can be overcome by choosing designs with low probabilities of early acceptance or by midtrial adaptations of the early acceptance boundary using the variability observed in the first stage. This modified procedure is negligibly anticonservative and preserves the power.

Animals↗