Split-plot designs for robust product experimentation

Genichi Taguchi has emphasized the use of designed experiments in several novel and important applications. In this paper we focus on the use of statistical experimental designs in designingproducts to be robust to environmental conditions. The engineering concept of robust product design is very important because it is frequently impossible or prohibitively expensive to control or eliminate variation resulting from environmental conditions. Robust product design enablesthe experimenter to discover how to modify the design of the product to minimize the effect dueto variation from environmental sources. In experiments of this kind, Taguchi's total experimental arrangement consists of a cross-product of two experimental designs:an inner array containing the design factors and an outer array containing the environmental factors. Except in situations where both these arrays are small, this arrangement may involve a prohibitively large amount of experimental work. One of the objectives of this paper is to show how this amount of work can be reduced. In this paper we investigate the applicability of split-plot designs for thisparticular experimental situation. Consideration of the efficiency of split-plot designs and anexamination of several variants of split-plot designs indicates that experiments conductedin a split-plot mode can be of tremendous value in robust product design since they not only enable the contrasts of interest to be estimated efficiently but also the experiments can be considerably easier to conduct than the designs proposed by Taguchi.     

The role of nonverbal sensitivity in childhood psychopathology

Two samples of boys were compared on their sensitivity to nonverbal communication: boys not in treatment (sampled from elementary school populations) and boys in treatment (sampled from community mental health clinics). Three main decoding tasks assessed the boys' ability to identify, classify, and predict nonverbal displays. Parent and self appraisals on nonverbal sensitivity were also collected. Results indicated that clinic-recruited boys performed less well than school-recruited boys on the three decoding tasks, but there were no significant differences in the self or parent appraisals. Older boys performed better than younger boys on the decoding tasks, but were not differentiated by self or parent appraisals. Further analyses revealed a significant positive correlation between the level of the boys' social incompetence and poor self-control and the number of decoding errors the boys made on the dominant/submissive but not the negative/positive dimension of nonverbal displays. In addition, the interrelationship of the nonverbal skills appeared to differ across clinic- and school-recruited boys. Results are discussed in terms of their relevance for understanding child psychopathology and their implications for child assessment and treatment.     

A case study in venture management

This paper examines the venture concept as a means of stimulating new product development within large companies. The literature reveals a significant difference between the venture idea as originally conceived in North America and as applied in European companies. A case study of a UK venture experiment is then presented which clearly demonstrates the resistance to substantive change which may characterise large organisations.     

Pearl'S Law and the Growth of the Scottish National Party

An examination of the percentage of the total Soottish poll gained by the SNP at general elections between 1951 and 1974 inclusive suggested that a stable, sigmoidal growth process might be identified if elections were regarded as occurring at unit intervals in time. The resulting time service was extremely well-fitted by Pearl's Law, a theoretical growth curve possessing a signoidal trend, and a theoretical upper limit to growth. On the basis of this model, estimated of, and confidence intervals for, both this theoretical upper limit, and of the party's performance in the next general election were obtained.     

Unemployment,union density,and wages

We investigate, both theoretically and empirically, whether long-run industry unemployment rates modify the wage impact of union density on the earnings of members. Our theory suggests that the density effect increases as unemployment increases. Our empirical estimates use wage equations exclusive and inclusive of unemployment and of the interactive effect of unemployment and density in influencing wages. Based on a 1985 sample of manufacturing production workers, our findings indicate that the wage effect of union density for union workers as usually measured is only 41 percent as large as the effect when unemployment is in the model.     

Investigation and improvement of a process for vacuum-formed ceramic fibre composites, using a fractional 2n experimental design

Refractory ceramic fibre products are used extensively as high temperature furnace linings, because of their properties of a lower thermal conductivity, improved resistance to thermal shock, cheaper installation and maintenance costs, and reduced wall thickness when compared with traditional firebrick linings. Vacuum forming of the ceramic fibre enables modular linings for small laboratory furnaces to be produced efficiently and cost effectively. This process involves drawing a slurry of ceramic fibre, binder and water through a porous mould, and subsequently oven drying the finished product. The product mass, which is strongly related to its relevant thermal and structural properties, was identified as a suitable response variable. A resolution IV fractional 2n factorial design was employed to investigate the effects of 15 key process factors on the product mass in just 32 experimental runs. Analysis of the results revealed that only two of the 15 process factors investigated in the experiment significantly affected the mean product mass, while none of these factors significantly affected the variability of the product mass. The experimental design is described, and its results and their implications are presented and discussed.     

Variations in Household Telephone Access: Implications for Telephone Surveys

The growing proportion of households that have telephones and the increasing cost of face-to-face interviewing have led to greater interest in the use of telephone surveys in Australia over recent years. However about 20 per cent of private households do not have a telephone service, which could give rise to significant biases in population estimates derived from telephone surveys if no adjustment is made to take account of the differences between those with and without telephone access. This paper examines these differences and suggests ways in which their possible effects might be overcome.