摘要:

Three chief subjects are considered. (1) How accurate should data be for economic analysis via economic models, particularly input-output models? Preference is expressed for measurement of additional aspects of economic phenomena rather than broadside improvements in generally sound series. (2) How can models be developed to utilize imperfect data? Several suggestions are made: (a) examine the concepts and methods lying behind the basic data; (b) question models acutely sensitive to the inclusion of a single observation; (c) evaluate the economic meaning of models after empirical testing; (d) set up check models. (3) How can more adequate data be developed? A proposal is outlined for an integrated set of data on financial aspects of business, and for similar integration of data on employment and on consumer economic behavior.

ISSN:0162-1459    

DOI:10.1080/01621459.1954.10483500

出版商:Taylor & Francis Group    

年代:1954

数据来源: Taylor

ISSN:0162-1459    

DOI:10.1080/01621459.1954.10483501

出版商:Taylor & Francis Group    

年代:1954

数据来源: Taylor

ISSN:0162-1459    

DOI:10.1080/01621459.1954.10483502

出版商:Taylor & Francis Group    

年代:1954

数据来源: Taylor

ISSN:0162-1459    

DOI:10.1080/01621459.1954.10483503

出版商:Taylor & Francis Group    

年代:1954

数据来源: Taylor

ISSN:0162-1459    

DOI:10.1080/01621459.1954.10483504

出版商:Taylor & Francis Group    

年代:1954

数据来源: Taylor

摘要:

* Work done in part as Consultant to Stanford University and in part under grant from the Higgins Foundation.

ISSN:0162-1459    

DOI:10.1080/01621459.1954.10483505

出版商:Taylor & Francis Group    

年代:1954

数据来源: Taylor

摘要:

Since most conventional control chart limits are designed so that the ratio of the expected number of false alarms to the number of samples tested is fixed in advance irrespective of the sample size, the ratio of false alarms to the number ofarticlestested varies with the sample size. In this paper, control charts are designed for which the expected ratio of false alarms to the number of articles tested is independent of the sample size. The power of the control charts in relation to sample size is investigated.

ISSN:0162-1459    

DOI:10.1080/01621459.1954.10483506

出版商:Taylor & Francis Group    

年代:1954

数据来源: Taylor

ISSN:0162-1459    

DOI:10.1080/01621459.1954.10483507

出版商:Taylor & Francis Group    

年代:1954

数据来源: Taylor

摘要:

The present section contains all available abstracts of papers presented at the 1953 national meeting of the American Statistical Association in Washington, D. C. The sequence of presentation here conforms to a grouping of abstracts according to the various sessions at which they were delivered.

ISSN:0162-1459    

DOI:10.1080/01621459.1954.10483508

出版商:Taylor & Francis Group    

年代:1954

数据来源: Taylor

ISSN:0162-1459    

DOI:10.1080/01621459.1954.10483509

出版商:Taylor & Francis Group    

年代:1954

数据来源: Taylor