Combining Neural Networks and Statistical Predictions to Solve the Classification Problem in Discriminant Analysis*

A number of recent studies have compared the performance of neural networks (NNs) to a variety of statistical techniques for the classification problem in discriminant analysis. The empirical results of these comparative studies indicate that while NNs often outperform the more traditional statistical approaches to classification, this is not always the case. Thus, decision makers interested in solving classification problems are left in a quandary as to what tool to use on a particular data set. We present a new approach to solving classification problems by combining the predictions of a well-known statistical tool with those of an NN to create composite predictions that are more accurate than either of the individual techniques used in isolation.     

The second sex: Women's place in polling language

Recent research suggests that gender nouns such as man, and pronouns such as he,· are not generic forms referring to humanity, but often refer exclusively to males. The use of male terms in a variety of role contexts serves to deny females identification with these contexts, and thereby stereotypes the roles of men and women. Public opinion polling, through its use of the questionnaire, represents an important social context where gender-biased phrasing can have an adverse impact on a large segment of society. A content analysis of polling questions from 1936–1973 was undertaken to ascertain their gender representation. Those questions containing gender referrants constituted ten percent of the sample. Questions were analyzed for their gender content in the areas of social, political, and economic life. The findings show a differential use of male and female terms, with the use of female gender most apparent in contexts traditionally defined as female: home and family. When females are mentioned in non-traditional contexts, they appear most often in competition with men. Male pronouns appear exclusively in contexts traditionally defined as male: work and politics. The methodological and social implications of these findings for survey research are discussed.     

The Other‐Race Effect in a Longitudinal Sample of 3‐, 6‐ and 9‐Month‐Old Infants: Evidence of a Training Effect

We investigated the development of the other‐race effect “ORE” in a longitudinal sample of 3‐, 6‐, and 9‐month‐old Caucasian infants. Previous research using cross‐sectional samples has shown an unstable ORE at 3 months, an increase at 6 months and full development at 9 months. In Experiment 1, we tested whether 9‐month‐olds showed the ORE with Caucasian and African faces. As expected, the 9‐month‐olds discriminated faces within their own ethnicity (Caucasian) but not within the unfamiliar ethnicity (African). In months. In Experiment 2, we longitudinally tested infants at 3, 6, and 9 months by presenting either the Caucasian or the African faces used in Experiment 1. In contrast to previous cross‐sectional studies and Experiment 1, we found that infants discriminated between all stimuli. Hence, we did not find the ORE in this longitudinal study even at 9 months. We assume that the infants in our longitudinal study showed no ORE because of previous repetitive exposure to African faces at 3 and 6 months. We argue that only a few presentations of faces from other ethnic categories sufficiently slow the development of the ORE.     

Metapher als Intervention

The main means trainers and consultants have is language. What they can reach very much depends on the choice of the “appropriate words”. So to find them – what to say and how to say it – plays a key role for the quality of interventions. In contrast to other possible conceptions we understand intervention as the establishment of a self-reflective distance of a system towards itself. So the main purpose of an intervention is to bring a system to a better understanding of itself and the situation the system is in. This concept of intervention is a precondition for the self-steering abilities of systems. The concept is culturally limited, for self-steering to some extent always means questioning the ruling principles and authorities. The article investigates the consciousness raising power of metaphors. It discusses the use of the rhetoric figure of the metaphor and connects this to general considerations concerning the anthropologic function of language. Giving things names contains pretty much emotionality, has a magical component and follows a mimetic principle. The allegedly rational terms themselves turn out to be metaphors, their semantic field is wider than one might expect which can be shown by the recourse to the etymology. The value of using metaphors for describing situations lies in their openness to collective emotions and in the widening of horizons. The article underlines the importance of a sense of language as part of the professional skills.     

Ein Experiment zum Einfluss der Situation auf physische und psychische N?he

Zusammenfassung In einem Experiment (N = 90) wurde der Einfluss situativer Bedingungen auf den Sitzabstand und auf die per Fragebogen erfasste subjektiv erlebte N?he zu einer anderen Person überprüft. Es wurden eine Kooperationsbedingung, eine neutrale Bedingung und eine Wettbewerbsbedingung realisiert. Es zeigte sich, dass die neutrale Bedingung die geringste r?umliche N?he hervorrief, w?hrend die Kooperationsbedingung die h?chste emotionale N?he ausl?ste. Zur Interpretation wird die Annahme vertreten, dass die neutrale Bedingung sich von der Wettbewerbsbedingung darin unterscheidet, dass sie von den Probanden als irrelevant empfunden wird und dadurch keine Aufmerksamkeitszuwendung auf die andere Person ausl?st. Weiterhin wird angenommen, dass emotionale N?he das Ergebnis von positiver Atmosph?re und Relevanz der Interaktion ist. Eine weitere Fragestellung bezog sich auf den Zusammenhang zwischen (geringem) Sitzabstand und emotionaler N?he, der in übereinstimmung mit Studien zur Einstellungs-Verhaltens-Konsistenz relativ gering ausfiel.

---

The influence of the social situation on the physical distance and, as measured by a questionnaire, emotional intimacy between participants was experimentally studied (N = 90) in a cooperative, a competitive, and a neutral social situation. Results indicate that the physical distance was greatest under neutral conditions. The cooperative condition produced the strongest feeling of intimacy. It is argued that the neutral condition is irrelevant to the participants’ behaviour and level of attention to other persons. Further, intimacy can only arise if an interaction is both positive and relevant. The correlation between physical and emotional distance turned out to be relatively small, in accordance with the range usually given for attitudebehaviour correlations.

---

---

PD Dr. Ina Grau, Dipl. Psych., Universit?t Bielefeld, Fakult?t für Soziologie, Lehrstuhl Sozialpsychologie, Prof. Dr. Hans-Werner Bierhoff, Dipl. Psych., Ruhr-Universit?t Bochum, Fakult?t für Psychologie, Lehrstuhl Sozialpsychologie     

The Value of Children,Parenting Strains,and Depression Among Urban African American Mothers

This exploratory analysis examines the relationship between two dimensions of the parenting experience—values regarding children and parenting strains—and depression in a sample of 285 urban, economically disadvantaged African American women. An examination of this relationship among economically disadvantaged African American women is important in light of the high stress exposure of these mothers, their limited opportunity to engage in rewarding employment, and in view of the centrality of motherhood in this cultural context. Our central finding is that the impact of parenting strains on depression is substantially moderated by beliefs about the costs and emotional fulfillment associated with having children. African American mothers who regard the costs of children as low and the potential of children for providing emotional fulfillment as high are significantly more depressed by high parenting strains than African American mothers who hold the opposite beliefs about the costs and benefits of children. Variation in values regarding children and parenting strains accounts for more than 40% of variation in depression. These findings suggest the importance of understanding the social–psychological factors that contribute to the quality of parenting, not only because maternal mental health is important in and of itself, but because maternal well-being has a direct impact on child well-being.     

A new approach to analyzing the Achilles' heel of multisource feedback programs: Can we really trust ratings of leaders at the group level of analysis?

This research addresses the Achilles’ heel of all multi-source leadership feedback programs regardless of whether they use 360°, 270°, or 180° ratings. Namely, should all feedback be universally aggregated by feedback groups and reported to respondents? A sample of raters from the Center for Creative Leadership (N = 15,500 including subordinates, peers, and others) was grouped to align with each raters’ matching focal subject (J = 1,550). Two scales from CCL’s Benchmarks instrument (Building and Mending Relationships and Career Management) were used to: (1) predict a derailment factor (Problems with Interpersonal Relationships) and (2) determine the operative level of analysis and its aggregation characteristics using Within and Between Analysis (WABA). In many cases, the variables under consideration did not aggregate optimally. Using a technique based upon each group’s coefficient of variation as a measure of internal consensus, between 3% and 25% of groups should not have their averages reported. Alternative approaches are discussed.