Monotonic weighted power transformations to additivity

A class of monotonic transformations which generalize the power transformation is fit to the independent and dependent variables in multiple regression so that the resulting additive relationship is optimized. This is achieved by minimizing a quadratic fitting criterion with linear inequality constraints on the parameters. A quadratic programming technique which works reliably and quickly in this application is outlined. Some examples of the analysis of artificial and real data are offered.This work was supported by National Research Council of Canada Grant APA 320 to the author.     

Activism Against Nuclear Weapons Build-up—Student Participation in the 1984 Primary Campaigns1

This study presents longitudinal data concerning factors that influence student participation in political campaigns for candidates opposed to nuclear weapons build-up. Participation in campaigns was stimulated by nervousness about the nuclear arms race and the possibility of nuclear war, a feeling of moral obligation to act to try to prevent nuclear war, and several additional factors. The factors that contributed to participation in political campaigns are quite different from the factors found previously to be associated with other types of activism against nuclear weapons build-up. These findings suggest varied strategies for recruiting people to participate in different types of activism.     

Some deleterious consequences of the act of recollection

Two experiments investigated the impact of responding to recognition test items that do not include a correct alternative. In Experiment 1, subjects who were given exclusively incorrect response alternatives were less likely than control subjects to favor the correct alternatives on a second recognition test. Analysis of subjects’ responses indicated that commitments, rather than distractor familiarity, was the main source of this effect. In Experiment 2, an impairing effect of committing to an incorrect alternative was observed even when the initial distractors were excluded from the final test. Thus, this decreased performance cannot simply be attributed to a bias toward remaining consistent. One interpretation of these results is that committing to a distractor causes subjects to remember a false detail that can interfere with their later ability to access the original information. Other potential theoretical and applied implications of these results are explored.     

Principal components analysis of sampled functions

This paper describes a technique for principal components analysis of data consisting ofn functions each observed atp argument values. This problem arises particularly in the analysis of longitudinal data in which some behavior of a number of subjects is measured at a number of points in time. In such cases information about the behavior of one or more derivatives of the function being sampled can often be very useful, as for example in the analysis of growth or learning curves. It is shown that the use of derivative information is equivalent to a change of metric for the row space in classical principal components analysis. The reproducing kernel for the Hilbert space of functions plays a central role, and defines the best interpolating functions, which are generalized spline functions. An example is offered of how sensitivity to derivative information can reveal interesting aspects of the data.This research was supported by Grant PA 0320 to the second author by the Natural Sciences and Research Council of Canada. We are grateful to the reviewers of an earlier version and to J. B. Kruskal and S. Winsberg for many helpful comments concerning exposition.     

Attitudes toward the division of child-rearing responsibility

Though the theoretical and clinical ramifications of attribution of primary responsibility to mothers for child rearing have been discussed extensively, there is little empirical data in this area. In the present study a sample of parents rated several child-rearing behaviors in terms of ideal maternal and paternal responsibility. Mothers were attributed primary responsibility for the largest proportion of behaviors, fathers for the smallest, and an intermediate number of items were rated as parentally shared. In addition, item analysis indicated a relationship between the directionality of responsibility attribution and traditional sex-role stereotypes. Implications for patterns of intrafamilial blame and clinical intervention are discussed.     

Psychophysiological assessment of imagery in obsessive-compulsives: a pilot study

In a within-groups design, four obsessive-compulsive patients imagined four different types of scenes: (1) exposure to feared situations; (2) exposure to feared situations followed by engaging in ritualistic behavior: (3) engaging in ritualistic behavior without provocation; and (4) neutral situations. Each scene type was presented three times with order of presentation randomized. Beat-by-beat heart rate samples were taken during the visualization of each scene and converted to second-by-second heart rate. An examination of second-by-second heart rate for each scene type suggested that scenes involving exposure to feared situations elicited defensive responses (DR), whereas those that did not elicited orienting responses (OR). The implications of these results for compulsive behavior, the fear-reduction model and Sokolov's (1963) OR-DR model, are briefly discussed.     

Conciliatory strategies and relative power

In an investigation of variations of a conciliatory strategy, 96 subjects, half female and half male, played a modified Prisoner's Dilemma in which their relative power, in terms of control over the other's outcomes, was less than, equal to, or greater than that of a simulated other. After 10 trials on which (s)he was 50% competitive, the other either did or did not send a handwritten note of general intent to be cooperative. In the second block of 10 trials, the other either was 100% cooperative or began with cooperation and stayed with it as long as the subject did not exploit. When the subject did exploit, the other made the competitive choice on the following trial as a form of retaliation, but followed this retaliation by a return to cooperation on the next trial so that there were never two consecutive competitive choices. During this second block, the simulated other always signaled in advance, with an explicit message, which choice (s)he was about to make. In a third block, the other was 100% cooperative in all conditions and sent no messages. As hypothesized, subjects whose power was equal to or less than that of the other were more cooperative in response to conciliation than were subjects whose power was greater than that of the other. Subjects who received the general note of intent were more cooperative than those who did not, but only in the equal and low power conditions. The subjects in the equal power condition who experienced retaliation were more cooperative than those who did not, but those in the unequal conditions tended to react negatively to retaliation. The results were discussed in terms of Osgood's (Peace Research Reviews, 1979, 8(2), 77–92) GRIT strategy and outcome control in interaction.     

The joint analysis of direct ratings,pairwise preferences,and dissimilarities

In studies involving judgments of similarity or dissimilarity, a variety of other variables may also be measured. Examples might be direct ratings of the stimuli, pairwise preference judgments, and physical measurements of the stimuli with respect to various properties. In such cases, there are important advantages to joint analyses of the dissimilarity and collateral variables. A variety of models are described for relating these and algorithms described for fitting these to data. A number of hypothesis tests are developed and an example offered.The research reported here was supported by grant number APA 320 to the author by the National Research Council of Canada.     

Some small sample results for maximum likelihood estimation in multidimensional scaling

Some aspects of the small sample behavior of maximum likelihood estimates in multidimensional scaling are investigated by Monte Carlo. An investigation of Model M2 in the MULTISCALE program package shows that the chi-square test of dimensionality requires a correction of tabled chi-square values to be unbiased. A formula for this correction in the case of two dimensions is estimated. The power of the test of dimensionality is acceptable with as few as two replications for 15 stimuli and as few as five replications for 10 stimuli. The biases in the exponent and standard error estimates in this model are also investigated.The research reported here was supported by grant number APA 320 to the author by the National Science and Engineering Research Council of Canada.