Use Of Datasets Research Articles

The Acceptability of Regression Solutions: Another Look at Computational Accuracy

Abstract Longley [10] proposed a set of data for use in testing regression programs. This paper shows that the numerically accurate solution in this example was probably an unreasonable estimate of the regression coefficients. This is true because the accuracy of the data and appropriateness of the model may affect the solution more than the computational method. An easily computed index is derived that can be used to indicate such computational instability. The basic conclusion is that a concern about highly accurate computational methods must be tempered with a concern for whether the data are accurate enough to make the results meaningful.

THE ACCEPTABILITY OF REGRESSION SOLUTIONS: ANOTHER LOOK AT COMPUTATIONAL ACCURACY

ABSTRACTLongley proposed a set of data for use in testing regression programs. This paper shows that the numerically accurate solution is likely to be an unreasonable estimate of the regression coefficients for this problem. This is true because the accuracy of the data and appropriateness of the model may affect the solution more than the computational method. An easily computed index is derived that can be used to indicate such computational instability. The basic conclusion is that a concern about highly accurate computational methods must be tempered with a concern for whether the data are accurate enough to make the results of such computation meaningful.

The Apollo Passive Seismic Experiment

The completed data set obtained from the 4-station Apollo seismic network includes signals from approximately 11,800 events of various types. Four data sets for use by other investigators, through the NSSDC, are in preparation. Some refinement of the lunar model based on seismic data can be expected, but its gross features remain as presented two years ago. The existence of a small, molten core remains dependent upon the analysis of signals from a single, far-side impact. Analysis of secondary arrivals from other sources may eventually resolve this issue, as well as continued refinement of the magnetic field measurements. Evidence of considerable lateral heterogeneity within the moon continues to build. The mystery of the much meteoroid flux estimate derived from lunar seismic measurements, as compared with earth-based estimates, remains; although, significant correlations between terrestrial and lunar observations are beginning to emerge.

Dysplastic growth differentials in patients with psychiatric disorders studies on the morphology of maturity.

This paper is concerned with some aspects of the morphology of man, as they present in relationship to the occurrence of psychiatric disorder. Kallmann in the U.S.A. and Slater in England, among others, have initiated fundamental work, summarized recently by Slater (1950), dealing with the genetic linkages involved in the causation of mental illness, while the sociological, clinical, psychiatric, and especially psycho-analytic literature provides an overwhelming if diffuse body of evidence pointing to the importance of environmental factors. It is not a function of this paper to review this literature; it is necessary only to mention its existence for orientation purposes since the studies which follow will necessarily have to be interpreted in terms of the relative importance of these two fundamental aetiological influences. The present investigation resulted from two assumptions based on the clinical observation of mentally ill patients in a psychiatric hospital. These were firstly, that there is something about the physical appearance of such patients which reflects their relative incapacity to remain emotionally stable under stress, and secondly, that, if this clinical impression be true, then over the length of the patient's life the separate features of his physical appearance on which the impression is based must have resulted from dysplastic or imperfect develop ment at some definite phase of his growth, since such features cannot be assumed to appear sud denly out of the blue. Some of the differentials contributing to the clinical hunch have been measured and analysed by Draper and his co-workers (1944) at Columbia University with respect to physical illnesses such as cholecystitis, peptic ulceration, rheumatoid arthritis, rheumatic fever, and migraine. Features of significance in Draper's data are convincing, not only because of the com prehensive nature of his morphological observations, but also because of their ontogenetic basis. This point will be referred to later, but it is worth remark ing here that, in the field of psychiatry, the scientific strictures of the developmental hypothesis in biology ave been largely neglected. Somatotypology has been widely employed by Kretschmer (1936) a d authors to reveal constitutional dysplasias in psychiatric patients but these studies remain largely descriptive, and therefore sterile, in so far as they do hot lead to an increase in understanding of causation and later of rational therapy in mental disor ers. The studies of Sheldon and others (1941, 1942) suffer from the same purely descriptive faults but do at least add one set of useful data to those of Kretschmer, i.e., an analysis of the predominating germ-layer involved in the resulting dysplasia. Both workers, moreover, have pointed to the concordance of emotional, temperamental, and other psychological features of the personality with the constitutional somatotype of the individual examined, each agreeing that the extreme dysplasias in any orientation of their classifications tend to be ass ciated with overt mental illness of a particular type (Kline and Tenney, 1950).