Constrained classification: The use of a priori information in cluster analysis

In many classification problems, one often possesses external and/or internal information concerning the objects or units to be analyzed which makes it appropriate to impose constraints on the set of allowable classifications and their characteristics. CONCLUS, or CONstrained CLUStering, is a new methodology devised to perform constrained classification in either an overlapping or nonoverlapping (hierarchical or nonhierarchial) manner. This paper initially reviews the related classification literature. A discussion of the use of constraints in clustering problems is then presented. The CONCLUS model and algorithm are described in detail, as well as their flexibility for use in various applications. Monte Carlo results are presented for two synthetic data sets with appropriate discussion of the resulting implications. An illustration of CONCLUS is presented with respect to a sales territory design problem where the objects classified are various Forbes-500 companies. Finally, the discussion section highlights the main contribution of the paper and offers some areas for future research.We wish to thank C. Mallows and J. D. Carroll for some helpful technical discussion and L. Clark and D. Art for their valuable computer assistance. We also wish to thank H. Pollak, R. Gnanadesikan, and J. Kettenring for their thorough reviews of a previous draft of this paper. Finally, we acknowledge helpful comments of the editor and two anonymous reviewers.     

Rehearsal grouping and hierarchical organization of serial position cues in short-term memory

Lists of 8, 9, or 10 digits were presented at the rate of 1 digit/sec. to subjects instructed to rehearse silently the digits in non-overlapping groups of 1, 2, 3, 4, or 5 digits, after hearing each digit. Subjects were instructed not to rehearse any digits outside the group currently being presented. Rehearsing in 3's was optimal, irrespective of list length. Both recall of items and recall of the correct positions of items improved from 1's to 2's to 3's. Recall of items declined very little from 3's to 4's to 5's, but recall of position declined sharply. Errors in positioning digits tended, above chance, to be in the same group or the same position in a different group. The results suggest that both item-to-item associations and serial positin-to-item associations are formed in short-term memory, that only two or three serial position cues are used, but that these serial position cues can be hierarchically organized into a beginning, middle, and end group and a beginning, middle, and end position within a group.     

The pastor and his privacy

Pastoral Psychology -     

Experimental design symbolization and model derivation

A method of design symbolization (DS) is given for a class of designs in which all factors are related by complete crossing or by nesting. Characterization of acceptable DS's, and logical relational properties are given. It is shown how the design model can be obtained from the DS.The preparation of this article was supported by NIH Grant GM-11128, and by grants to the Institute of Human Learning by NSF and NIH.     

Factor analysis by minimizing residuals (minres)

This paper is addressed to the classical problem of estimating factor loadings under the condition that the sum of squares of off-diagonal residuals be minimized. Communalities consistent with this criterion are produced as a by-product. The experimental work included several alternative algorithms before a highly efficient method was developed. The final procedure is illustrated with a numerical example. Some relationships of minres to principal-factor analysis and maximum-likelihood factor estimates are discussed, and several unresolved problems are pointed out.The authors wish to thank the Factor Analysis Work Group (supported, in part, by ONR) for valuable criticisms and suggestions made in the course of a discussion of the present work in April, 1965.     

Short-term recognition memory for single letters and phonemic similarity of retroactive interference

Copying 12 letters produces more retroactive interference in recognition memory for a single letter when the interference letters possess a vowel sound in common with the letter to be remembered than when they do not. Compared to interference lists that do not include the presented letter, inclusion in the interference list of the letter to be remembered improves recognition memory when the other interference letters have no vowel sound in common with the letter to be remembered, but not otherwise. False recognition rates are greater when the test letter contains a vowel sound in common with the presented letter than when the vowel sounds of these two letters are different. The findings are in complete accord with analogous findings for short-term recall and indicate that short-term recognition memory uses the same phonemic-associative memory system as short-term recall.