Several methods, mostly based on a consideration of the eigenvalues, have been previously used to determine the number of factors to retain in a factor analysis. This paper shows how split-half factor comparabilities, based on factor scores, can provide a direct and unambiguous method of determining the number of reliable factors that should be retained, and of assessing the appropriate rotation that should be used. The method is particularly appropriate for taxonomic factor analysis, where the factor scores are to be used as summary or classificatory measures. It is shown that, for respondents from non-homogenous populations, split-halves based on the sub-populations should be considered, as well as random split-halves. The discussion is supported by a number of principal components analyses, using synthetic data sets of known factor structure and an actual semantic response data bank.

The geometrical fit of the continents now separated by oceans has long been discussed in relation to continental drift. This paper describes fits made by numerical methods, with a 'least squares' criterion of fit, for the continents around the Atlantic ocean. The best fit is found to be at the 500 fm. contour which lies on the steep part of the continental edge. The root-mean-square errors for fitting Africa to South America, Greenland to Europe and North America to Greenland and Europe are 30 to 90 km. These fits are thought not to be due to chance, though no reliable statistical criteria are available. The fit of the block assembled from South America and Africa to that formed from Europe, North America and Greenland is much poorer. The root-mean-square misfit is about 130 km. These geometrical fits are regarded as a preliminary to a comparison of the strati graphy, structures, ages and palaeomagnetic results across the joins.

No abstract

No abstract

DR. JOHN WTXSON IS A SENIOR Lecturer IN the Department of Accounting and Finance and Dr. Jim Everett an Associate Professor in the Department of rinformation Management and Marketing, both at the University of W'esterin Australia, Australia. Results froin pre v ious studies examining the incidence of smiall business failure have reported significant variations in failure rates betxveen industry sectors. Indeed, the results from some studies are in direct coniflict. For example, Lowe, McKenna and Tibbits (1991) reported that the manufacturing sector had the highest failure rate while Bruderi, Preisendorfer H/ (d! (1992) and Phillips and Kirchoff (1989) fotund that the manufactuiring sector had the highest survival rate. Th-e significarnt variations in reported failul-e rates and the apparent conflict between th-ie findings of some studies must surelv be a souti-ce of some confusion for policx- tiiakers and others with an interest in the smiall businiess sector. The results of this sttldx su-ggest that reported failure rates iiiay (lepend heavily on the definition of tfailtui-e adopted. A better understanding of the effect that choice of failuire definiitioni may have on reported failur e rates sh(ould lead to improved policy de isions bv governments, financial instituitions an1d othtie grotips xsNitll mt11 interest in smIiall l)usilICss.