Premium
Het matchen in de psychologie
Author(s) -
Spitz J. C.
Publication year - 1953
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.1467-9574.1953.tb01057.x
Subject(s) - mathematics , series (stratigraphy) , matching (statistics) , combinatorics , table (database) , permutation (music) , degree (music) , discrete mathematics , statistics , computer science , data mining , paleontology , physics , acoustics , biology
Summary Matching in Psychology In experimental psychology the matching method seems to be one of the most promising methods. This paper deals with the matching of a series of t elements against a series of t other elements. The stated null hypothesis is: for a fixed order of one of the series each permutation of the second series has the same probability. As a test criterion in the case of one matching is defined: the number r of matches, e.g. the number of “right pairs”; for n matchings: . As the distribution of for t=3 cannot be found in literature, a table of is given for n=I(I)30: see table I. The way of calculating and verifying is discussed. For the problem of t=3 and n 31 the degree of approximation by the normal distribution and by the Pearson's Type III distributions is examined. With the application of a correction for continuity the former gives for n=30 a reasonable approximation, the latter a very good approximation: see table 2.