An alternative to the methodology for analysis of covariance

A general statistical model for simultaneous analysis of data from several groups is described. The model is primarily designed to be used for the analysis of covariance. The model can handle any number of covariates and criterion variables, and any number of treatment groups. Treatment effects may be assessed when the treatment groups are not randomized. In addition, the model allows for measurement errors in the criterion variables as well as in the covariates. A wide variety of hypotheses concerning the parameters of the model can be tested by means of a large sample likelihood ratio test. In particular, the usual assumptions of ANCOVA may be tested.Research reported in this paper has been partly supported by the Swedish Council for Social Science Research under project Statistical methods for analysis of longitudinal data, project director Karl G. Jöreskog, and partly by the Bank of Sweden Tercentenary Foundation under project Structural Equation Models in the Social Sciences, project director Karl G. Jöreskog.     

The effect of punishment and reinforcement procedures on a stutterer's factor II avoidance response

The purpose of this clinical experiment was to investigate the manipulability of a Factor II avoidance response (arm movement). The substantive hypothesis under test was that its suppression would be enhanced by repeated punishment sessions. During the first of nine therapy sessions response-contingent negative stimulation was not delivered. In the next five, after a base-rate period, “wrong” was made consequent upon the target response. In each of these sessions the suppression of arm movement was apparent. However, the magnitude of the suppression decreased from session to session. Evidently, the continued use of punishment did not enhance suppression. Because of this finding procedural changes were introduced in the next three therapy sessions. Verbal praise was added as a consequence of increasingly stringent omission criteria and three consecutive units of praise led to time-out from the reportedly fatiguing task of continuous speech. A marked reduction and then complete extinction resulted. Moreover, learning was evidenced since the frequency of arm movement lessened in the base-rate period of each session.     

The subjective vertical under ‘dry’ and ‘wet’ conditions at clockwise and counterclockwise changed positions and the effect of a parallel-lined background field

Summary The dependence of the subjective vertical (SV, the angle between a subjective vertical line and body median plane) on the gravity vertical (body tilt position, angle ) and on the optical vertical (i.e., a field of parallel lines seen as background to the line to be adjusted) was investigated. The SV was measured under dry and wet conditions at different degrees of body tilt attained in either clockwise (CW) or counterclockwise (CCW) progression.The measured difference in between field-of-lines left and right of the line is smallest at the upright position (=O°) and largest at =150°/165°. All body positions show a -difference between CW and CCW attainment (hysteresis), this too being least at upright and greatest at inverted body positions.These results, and changes of with test time, are discussed relative to the hypothesis that efficiency of the statolith organs decreases with body tilt increase, favouring increase of interference of somatoreceptors and the optical reference.     

A reconstructive-processing interpretation of the recognition failure phenomenon

The generality of the recognition failure phenomenon was the primary basis for the present research. It was hypothesized that the phenomenon obtains whenever an insufficient degree of compatibility exists between the task demands at recognition and those present during study. Experiments 1 and 2 demonstrated that, despite a recognition task that was incompatible to conditions of encoding, subjects could actively reconstruct aspects of the initial situation, and hence recognition failure was minimized. It was assumed that the stimulus material — categorized word pairs — facilitated this reconstructive process. Experiment 3 was designed specifically to examine further the reconstructive process as well as recognition failure. The results indicated that (a) the nature of this process could be mathematically predicted, and (b) subjects seldom failed to recognize words they recalled. The results of the present study strongly suggest an important boundary condition of the phenomenon of recognition failure. Moreover, Tulving and Wiseman's (1975) quadratic function could not predict accurately the obtained magnitude of recognition failure.     

The influence of verbal specifications on ratings of relative properties of pictures

Summary Sixty subjects were shown pictures each containing two objects (A and B). The objects were unobtrusive with respect to relative properties such as long — short, big — small, etc. The pictures were preceded by sentences referring either to objects A and B or only to object A. Object A was dimensionally specified: The same object was called, e.g., long for one and short for a different subject. Ratings during a later test (without presentation of the pictures) revealed that not only were objects A remembered according to the verbal specification provided during the presentation but ratings of objects B also changed in line with the specification of A. This was true even when object B had not been mentioned verbally during the presentation. In this way, verbal specifications had an effect beyond the referential bond. It is concluded that models based on the assumption that the semantic interpretation of an event depends on the application of syntactic rules — rules indicating, e.g., to which object an adjective refers — lack an important clue.     

Some theory and results for metrics for bounded response scales

In an earlier note, a new metric for bounded response scales (MBR) was introduced which resembles the city-block metric but is bounded above. It was suggested the MBR may be more appropriate than minkowski metrics for data obtained with bounded response scales. In this article, some formal properties of the MBR are investigated and it is shown that it is indeed a metric. Empirical predictions are then derived from the MBR and contrasted with those of a “monotonicity hypothesis,” which holds that dissimilarity judgements tend to be biased towards overestimation of larger distances, and with the predictions of the minkowski metrics, which imply additivity of collinear segments. Some empirical results are presented which contradict the monotonicity hypothesis and the minkowski metrics, and favor the MBR. Finally, the logic used to motivate the MBR is invoked to define a subadditive concatenation for bounded norms in the one-dimensional case, which may be useful in psychophysical work where the upper bounds are often real, rather than due to the response scale. This concatenation predicts understimation for doubling and overestimation for halving and middling tasks.     

Isobias curves in some detection tasks

Signal/noise ratio was varied in three tone-detection tasks while other conditions were held constant, generating 9 8-point and 12 12-point isobias plots. These plots were compared with the theoretical isobias curves associated with seven proposed bias measures. Each of the proposed measures covaried with signal/noise ratio in at least half of the data. For two of the plots, none of the measures was invariant. It is clear that different subjects respond quite differently to change in discriminability, and the implications of this for bias measurement are discussed.     

On the logic of adverbs

In the paper I investigate aspects of adverbial modification as an operation applying an adverb or adverbial phrase to a predicate and thereby creating a new predicate. The logic of adverbial modification, on this view, belongs to the logic of predicate modifiers. The theory I present is intended to cover not only adverbial modification but also attributive modification, but problems concerning the latter will not be given any special attention.     

On the conflict between logic and belief in syllogistic reasoning

Three experiments are reported that investigate the weighting attached to logic and belief in syllogistic reasoning. Substantial belief biases were observed despite controls for possible conversions of the premises. Equally substantial effects of logic were observed despite controls for two possible response biases. A consistent interaction between belief and logic was also recorded; belief bias was more marked on invalid than on valid syllogisms. In all experiments, verbal protocols were recorded and analyzed. These protocols are interpreted in some cases as providing rationalizations for prejudiced decisions and, in other cases, as reflecting a genuine process of premise to conclusion reasoning. In the latter cases, belief bias was minimal but still present. Similarly, even subjects who focus primarily on the conclusion are influenced to an extent by the logic. Thus a conflict between logic and belief is observed throughout, but at several levels of extent.     

Affect and performance in simple problems

We investigate the effects of emotional mood states on solving simple pattern comparison tasks. It is assumed that the effects of mood are mediated through response biases as well as effort expenditure. Departing from these assumptions, trend hypotheses are derived for the response biases evaluated by means of a signal detection analysis, for the response latencies, and for an adequate index of the underlying ability. The data support the hypotheses, and in particular, a positive mood state is found to improve performance. The assumption of mood-dependent differences in effort expenditure provides a parsimonious and integrative account of our results, while a reduced-capacity view cannot explain the findings obtained under positive mood induction.