Premium
Systematische en toevallige fouten bij het aflezen van de stand van een wijzer op een school *
Author(s) -
Homaker H. C.
Publication year - 1949
Publication title -
statistica neerlandica
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.52
H-Index - 39
eISSN - 1467-9574
pISSN - 0039-0402
DOI - 10.1111/j.1467-9574.1949.tb00404.x
Subject(s) - mathematics , statistics , estimator , random error
Summary Systematic and random errors in reading the position of a pointer on a scale In a prèvious article 1) V. Varangot has recorded the readings by 40 different persons of ten different positions of a pointer, the actual positions being known. He has pointed out certain discrepancies in determining according to different criteria who is the “best” estimator. In the present paper Varangot's data are subjected to a further analysis. It is shown that the discrepancies are mainly the result of the existence of both systematic and random errors. If these are not clearly distinguished discrepancies as observed may always be expected. When, on the other hand, both types of error occur together it is very difficult, if not impossible, rigorously to decide who is the “best” estimator. Spearman's criterion based on correlation coefficients is shown to be unsatisfactory. Finally the order of magnitude of systematic and random errors in Varangot's data are deduced by a‘ regression analysis. A brief abstract of some further papers on this subject is given at the end.