Arm - en - rijk oftewel: leren - de dommekracht onder de inkomenslift*

Summary This address deals with factors contributing to the inequality of income, with particular reference to the Netherlands and to the U.S.A. Sections 2 and 3 survey the shape of the income distribution curve, and the (possible) impact of the length of the reference period and of capital income on income variance. Section 4 shows the modest contribution of intergroup (compared with intra-group) differences in (mean) income, after classification of income recipients by occupation, age and sex. Special attention is paid to estimation of schooling and intelligence effects in multivariate analysis. All systematic factors taken together appear to account for less than 1/3 of the total income variance, so that a major part of it is still unexplained; mainly entrepreneurial income seems to be responsible for it. Finally, possible changes in income distribution - autonomous or resulting from policy measures -are discussed summarily.     

Streuungszerlegung für Paar-Vergleiche

Zusammenfassung Pl?ne für Paar-Vergleich Versuche werden gegeben und die Auswertung solcher Versuche wird beschrieben. Die Arbeit st?tzt sich auf das vonBock (1958) abge?nderte Modell vonThurstone undMosteller (1951a). Die Pl?ne sind so, da? sie zu unabh?ngig verteilten Fehlern führen. Zu ihrer Auswertung kann bedenkenlos die Streuungszerlegung angewendet werden. Die Konstruktion von ausgewogenen Pl?nen (allek(k−1)/2 m?glichen Paare werden verglichen) wird beschrieben und teilweise ausgewogene Pl?ne (nicht alle m?glichen Paare werden verglichen) für bis zu 10 Objekte sind tabelliert. Es wird gezeigt, da? die Auswertung solcher Versuche mit der Auswertung von Versuchen in unvollst?ndigen Bl?cken identisch ist. Zur leichteren Auswertung der teilweise ausgewogenen Versuche sind die Inversen der Koeffizientenmatrix der Normalgleichungen tabelliert. Ein Anwendungsbeispiel aus der Wollforschung wird durchgerechnet.

---

Summary Designs for paired comparisons are given and their analysis is discussed. The paper is based onBock’s modification (1958) ofThurstone andMosteller’s (1951a) model. The designs lead to independently distributed errors and analysis of variance can be applied. The construction of balanced designs (in which allk(k−1)/2 possible comparisons are made) is described and partially balanced designs (in which not all possible comparisons are made) for up to 10 objects are tabulated. It is shown that the analysis of paired comparison experiments is identical to the analysis of incomplete block experiments. To simplify the analysis of the partially balanced designs the inverses of the matrices comprising the coefficients of the normal equations are tabulated. A worked example of application to a problem in wool research is given.

---

---

Die Arbeit wurde im South African Wool Textile Research Institute, Rhodes University, Grahamstown, Südafrika, geschrieben.     

The link between attractiveness of “Extrabrand” attributes and the adoption of innovations

While innovators may rush to purchase many new products, most consumers are more conservative and do not want to buy into fads but purchase only those new products that are viable. How do the majority of consumers make judgments about whether they will adopt an innovation? This article examines the evaluative aspects of adoption as a means for better understanding consumer adoption and the market factors that may influence the success of an innovation. This research introduces a conceptual model that shows how consumers’ evaluation of product category attractiveness affects the adoption decision for really new products. These consumer evaluations are based on the attributes of the product category (“extrabrand” attributes) rather than brand attributes. Results from a test of the model indicate that consumers do use extrabrand attributes to assess the attractiveness of innovative new products.