Showing papers in "Journal of the American Statistical Association in 1953"
1,532 citations
TL;DR: The up-and-down method for quantal data was originally devised for testing the sensitivity of explosives as mentioned in this paper, and it has been shown to be 30 or 40 per cent more efficient than the usual probit analysis method.
Abstract: THE up-and-down method for estimating the 50 per cent response point of quantal data was originally devised for testing the sensitivity of explosives. Dixon and Mood [6] suggested that this method might have advantages in other fields of application. They gave approximate maximum likelihood estimates for the parameters of the (normal) response curve, and approximate formulas for the standard errors of these estimates. They pointed out that, on the basis of formulas for the asymptotic variance, the up-and-down method was 30 or 40 per cent more efficient than the usual probit analysis method. In spite of this efficiency advantage, the up-and-down method does not seem to have been given much consideration in such fields as bioassay [10, p. 909] or fatigue testing of metals. It has suffered from two restrictions which, taken together, have greatly limited its usefulness in many experimental situations:
435 citations
TL;DR: In this paper, a method of estimating the BioAssay with Quantal Response based on the Logistic Function was proposed. But it is not a simple method to estimate the bio-assay with quantal response.
Abstract: (1953). A Statistically Precise and Relatively Simple Method of Estimating the Bio-Assay with Quantal Response, Based on the Logistic Function. Journal of the American Statistical Association: Vol. 48, No. 263, pp. 565-599.
405 citations
TL;DR: The Circular Normal Distribution: Theory and Tables as discussed by the authors has been studied extensively in the literature, see, e.g., the Journal of the American Statistical Association, Vol. 48, No. 261, pp. 131-152.
Abstract: (1953). The Circular Normal Distribution: Theory and Tables. Journal of the American Statistical Association: Vol. 48, No. 261, pp. 131-152.
188 citations
159 citations
TL;DR: In this article, a note on regression when there is extraneous information about one of the coefficients is given, and a discussion of the effect of such information on the performance of regression is given.
Abstract: (1953). A Note on Regression When There is Extraneous Information About One of the Coefficients. Journal of the American Statistical Association: Vol. 48, No. 264, pp. 799-808.
119 citations
97 citations
TL;DR: The report of a committee appointed by S. S. Wilks as discussed by the authors to review the statistical methods used by Kinsey, Wardell B. Pomeroy, and Clyde E. Martin in their Sexual Behavior in the Human Male was published as a monograph by the American Statistical Association in 1954.
Abstract: * This article consists of the main text, but not the appendices, of the report of a committee appointed in 1950 by S. S. Wilks as President of the American Statistical Association, to review the statistical methods used by Alfred C. Kinsey, Wardell B. Pomeroy, and Clyde E. Martin in their Sexual Behavior in the Human Male (Philadelphia, W. B. Saunders Co., 1948). For further details on the appointment of the committee and its charge, see Section 1, p. 676 below. For an outline of the appendices, as well as of this paper, see Section 3, pp. 678–81, Appendix G, “Principles of Sampling,” will appear as an article in the March issue of this Journal. The full report, including both the text given here and the appendices, will be published as a monograph by the American Statistical Association in 1954.
97 citations
TL;DR: In this paper, a probability mechanism for the simultaneous production of the bias of nonresponse and variance of response is proposed, where nonresponse arises from two sources, not at home, and refusal.
Abstract: The author postulates a probability mechanism for the simultaneous production of the bias of nonresponse and for the variance of response. The nonresponse arises from a graded series of classes of the members of the universe to be sampled. The classes range from an impregnable core of no possible response, on up to a class of complete response. Nonresponse arises from two sources, not at home, and refusal. Refusals are of two kinds, permanent and temporary. The variation in the amount of time spent at home, and the variation in the firmness of the temporary refusal, produce the graded series of classes. The bias of nonresponse arises from the variation of any characteristic from one class to another. The variance of response arises from the variation of any characteristics from one member to another within a single class, and from the random variation in the number of responses therefrom. An increase in the size of the initial sample or a more efficient method of selection will decrease the varia...
93 citations
TL;DR: In this article, confidence and tolerance intervals for the normal distribution are presented for the various cases of known and unknown mean and standard deviation, and the relationship between the two types of intervals is described.
Abstract: Confidence and tolerance intervals for the normal distribution are presented for the various cases of known and unknown mean and standard deviation. Practical illustration and interpretation of these intervals are given. Tables are presented permitting a comparison among the intervals. Finally the relationship between the two types of intervals is described. * Presented at the annual meeting of the American Statistical Association, Boston, December 1951.
89 citations
TL;DR: A conference on sampling conducted by the Institute of Statistics, University of North Carolina at the Blue Ridge Assembly, 21 July 1952 as discussed by the authors, was held at the University of South Carolina.
Abstract: * Delivered at a conference on sampling conducted by the Institute of Statistics, University of North Carolina at the Blue Ridge Assembly, 21 July 1952.
TL;DR: This paper is a revision of a paper presented at the 1952 annual meeting of the American Statistical Association and aims to clarify the aims and methodology of this work and provide a basis for future research.
Abstract: * Revision of a paper presented at the 1952 annual meeting of the American Statistical Association
TL;DR: In this article, a method of estimating the parameter of a Poisson distribution which has been truncated at the lower end is presented, and application is made to a number of actual examples.
Abstract: This paper gives a method of estimating the parameter of a Poisson distribution which has been truncated at the lower end. Application is made to a number of actual examples.
TL;DR: In this article, it is shown that for certain types of runs, the sample size n = 1 turns out to be very economical, so that time saving methods of control by gauging may be introduced without appreciable loss of efficiency.
Abstract: For quality control charts controlling the mean of a normal population, either small samples are taken out at frequent intervals or large samples at less frequent intervals. It will be shown that in order to detect small changes of the population mean, the amount of inspection is greatly reduced by the selection of large samples. However, if for other reasons small samples are desirable, a control by runs of sample means above or below certain control limits makes it possible to use small samples and yet maintain the advantage of a reduced amount of inspection. For certain types of runs, the sample size n = 1 turns out to be very economical, so that time saving methods of control by gauging may be introduced without appreciable loss of efficiency.
TL;DR: In this paper, a simple technique is developed for determining the best single sampling inspection plans when the sample size is fixed, and two cases are considered: (1) single sampling plans for placing a lot into one of two categories and (2) single sampled plans to place a lot in one of three categories.
Abstract: A simple technique is developed for determining “best” single sampling inspection plans when the sample size is fixed. Two cases are considered: (1) single sampling plans for placing a lot into one of two categories and (2) single sampling plans for placing a lot into one of three categories. For case (1), the “best” plans are based upon a criterion of minimizing the sums of the producer's and consumer's risks. A similar criterion is employed as a basis for choosing “best” plans for case (2). The more general result for m categories is given also. Tables 1 through 8 are presented to enable the user to choose the appropriate sampling plans.
TL;DR: The methods used in this paper might be applicable to related problems, such as finding bounds to the errors in solving simultaneous equations, but it seemed best to limit this paper to errors in elements of the inverse matrix.
Abstract: (1953). On Errors in Matrix Inversion. Journal of the American Statistical Association: Vol. 48, No. 262, pp. 289-319.
TL;DR: In this paper, the mean and variance of Type III populations from singly and doubly truncated samples with known truncation points were estimated by using moments and maximum likelihood methods.
Abstract: The problem considered is the estimation of the mean and variance of Type III populations from singly and doubly truncated samples with known truncation points when the number of unmeasured observations is (i) unknown for each tail, (ii) known separately for each tail, (iii) known jointly for the two tails. For singly truncated samples, estimation of all three parameters has also been considered. Estimating equations are obtained by the method of moments as well as by the method of maximum likelihood. Previous results for the normal and Type III populations are particular cases of the results obtained in this paper. Numerical examples are given.
TL;DR: In this paper, some procedures for the rejection of suspected data were discussed. But none of the procedures were discussed in this paper, except for the procedure for rejecting the rejected data.
Abstract: (1953) On Some Procedures for the Rejection of Suspected Data Journal of the American Statistical Association: Vol 48, No 263, pp 531-533
TL;DR: In this article, two approximate methods are given for construction of these confidence intervals and it appears that the second approximation, which allows for correction for discontinuity in the observations, agrees quite closely with the exact computations.
Abstract: IN A number of different sampling situations, we encounter a problem which may be stated as follows: From a population of N objects, each of which possesses or does not possess a certain characteristic, we select without replacement a random sample of n objects. Observing that m of these possess the characteristic in question, we wish to estimate the total number M having the property. This problem arises in sample surveys of human populations and in sampling inspection of manufactured articles as well as in other, less obvious, situations. It is well known (see e.g., F. F. Stephan's solution, quoted by Deming [2, p. 294]) that the maximum likelihood estimate for M is the largest integer in (m/n) (N+ 1). The problem of confidence interval estimation for M, however, seems to have escaped systematic investigation in the literature. In this paper, the problem is considered and two approximate methods are given for construction of these confidence intervals. On the basis of some preliminary investigations, it appears that the second approximation, which allows for correction for discontinuity in the observations, agrees quite closely with the exact computations.
TL;DR: A non-technical description of one electronic computer known to the authors and an account of some of its applications in the field of economic statistics is given and some attention is given to the question of relative costs.
Abstract: (1953). Electronic Computation in Economic Statistics. Journal of the American Statistical Association: Vol. 48, No. 263, pp. 414-428.
TL;DR: In this article, a theory of estimation and of tests of significance is sketched for the case of controlled experiments when the inherent relation between two variables is non-linear, and the results are applied to the study of a constructed example.
Abstract: Joseph Berkson has indicated that when the error of one of two variates is “controlled” in a sense which he defines, and when the true relation between the variates is linear, the classical regression formula can validly be used for the unbiased estimation of the coefficients. In the present paper, Berkson's results are discussed and a theory of estimation and of tests of significance is sketched for the case of controlled experiments when the inherent relation between two variables is non-linear. The results are applied to the study of a constructed example.
TL;DR: In this paper, the first three moments of the logarithmic-normal distribution were used to approximate the unknown distribution by equating the first two moments of an unknown distribution.
Abstract: A COMMON statistical problem is that of testing a null hypothesis us< ing a statistic drawn from some unknown distribution. The general configuration of the probability density function is known from empirical evidence. It is suggested that for certain applications, the logarithmic-normal distribution be used to approximate the unknown distribution by equating the first three moments. It would then be convenient to have a table of critical values of the log-normal distribution, standardized for the first two moments and tabulated for various values of the skewness.
TL;DR: Bogue as discussed by the authors described the procedure used in making a functional grouping of the counties of the United States, Bureau of the Census, Washington, D.C., 1951, with the assistance and cooperation of the Bureau of Population and Economic Research, University of Virginia.
Abstract: * Review article on Donald J. Bogue, State Economic Areas—A description of the procedure used in making a functional grouping of the counties of the United States, Bureau of the Census, Washington, D. C., 1951. This paper was done with the assistance and cooperation of the Bureau of Population and Economic Research, University of Virginia.
TL;DR: In this paper, the Census Tract Conference, American Statistical Association (ASA) meeting in Chicago, December 28, 1952, the author was affiliated with the University of Rochester.
Abstract: * Presented before the Census Tract Conference, American Statistical Association, meeting in Chicago, December 28, 1952. At the time of presentation the author was affiliated with the University of Rochester.
TL;DR: The authors presented at the American Statistical Association Meetings in Chicago, December 27-28, 1952, the first edition of the AASA Conference on Distributed Information System (DIS).
Abstract: * Presented at the American Statistical Association Meetings in Chicago, December 27, 1952.