The effects of collaborative practice on statistical problem solving: Benefits and boundaries

In recent years, understanding the effects of collaboration on learning and memory has emerged as a major topic of investigation. Findings from applied educational research and from basic cognitive research demonstrate a complex view of how collaboration affects learning. The present laboratory study bridged these two domains of research to address the question of how collaborative learning affects statistical problem solving. After viewing a lecture, participants completed two statistics tests. They either completed the tests collaboratively and then individually, or completed both tests individually. Results showed an immediate benefit of collaboration, but this benefit did not persist on a subsequent individual test. Repeated practice by those who worked individually increased performance to the level of those who had previously collaborated. These results were qualified by gender as females showed a consistent benefit from prior collaboration on the post-collaborative test, particularly on conceptual problems. Implications for education are discussed.     

Assumptions of agreement and familiarity on the abbreviated conners teachers rating scale

Implicit in the development and use of the Conners' scales are the assumptions that the raters will agree upon their rating of hyperactivity regardless of a lack of operational defitions of items in such scales, and that the cut-off score (1.5) for ADHD should be higher in cases where the rater has some familiarity with the clinical case. This study describes a method for the evaluation of both assumptions. Seven children were rated by nursing staff and school teachers using the Abbreviated Conners Teachers Rating Scale (ACTRS) daily. Observations were conducted during the 7 a.m.-3 p.m. and 3 p.m.–7 p.m. shifts in order to provide two separate scores across raters. The results supported the assumption of agreement within (for example nurses' ratings during the 7 a.m.–3 p.m. period) and between (i.e. nurses' ratings versus teachers' ratings) settings. The results did not support the assumption of familiarity: scores during early observations were similar to the scores during later observations (when more familiarity with the clinical case is assumed). It was concluded that familiarity with the clinical case does not necessarily lead to higher scores (i.e. a better prediction of hyperactivity as a function of the rater's familiarity with the case) on the ACTRS. Agreements within and between observers could be a function of the raters' agreement in terms of using a specific cut-off point (i.e. 1.5) in their definition of a case as ‘hyperactive’, rather than the raters' agreement on reporting similar scores across observers.     

Differentiating visual and kinesthetic imagery in mental practice.

Through two experiments, the study sought to emphasize the usefulness of the visual and kinesthetic imagery in mental practice. In Experiment 1, it was hypothesized that when the task to be learned through mental practice necessitates the reproduction of a form by drawing, the visual image, which provides a wide span of apprehension, is more suitable than the kinesthetic image. On the other hand, the kinesthetic image that supplies inputs from the muscles' positions and movements should be more appropriate for the acquisition of the duration of the drawing. In Experiment 2, it was hypothesized that the task, transformed into a motor task necessitating minute coordination of the two hands, would benefit more from kinesthetic imagery. To have optimal control over what was actually experienced during mental practice, the participants' imagery skills were measured. The participants also benefited from prior imagery training. The results demonstrate that when using mental practice to initially acquire a task, visual imagery is better for tasks that emphasize form while kinesthetic imagery is better for those tasks that emphasize timing or minute coordination of the two hands.     

Differentiating the effects of self-control and self-esteem on reactions to mortality salience

Whereas many previous studies suggest that self-esteem may buffer against the psychological threat of death, recent research has begun to suggest that self-control also may serve as a buffer. Two studies examined the possibility that dispositional self-control uniquely predicts responses to mortality salience, above and beyond self-esteem. In Study 1, an initial exercise in emotion regulation increased subsequent accessibility of death thoughts. In Study 2, mortality salience increased worldview defense. Both of these effects were moderated by dispositional self-control, such that the effects occurred among participants with low but not high self-control. More importantly, these moderating effects were observed over and above the moderating effects of self-esteem. Findings suggest that self-control may serve as an important and unique buffer against thoughts of death.     

Differentiating the N3 and N4 electrophysiological semantic incongruity effects

An event-related potential termed the N4 has been widely studied due to its sensitivity to semantic incongruity. A recent report (Nobre & McCarthy, 1994) indicates there is also an N3 component that is sensitive to semantic incongruity. To differentiate these two components, an existing data set with 65 electrode sites, 78 subjects, and 120 sentences was examined. Instead of the usual procedure of averaging over the stimuli within distinct categories for each subject, a new approach--averaging over subjects--was employed. In this item average approach, 120 averages (one per sentence) were produced. Correlational analyses indicate that the N3 is equally sensitive to cloze probability and sentential constraint. The N4, by contrast, is more sensitive to sentential constraint and less to cloze probability; it is also correlated with familiarity. We interpret these results as evidence that the N3 is more responsive to semantic fit whereas the N4 is more responsive to semantic expectancy.     

Differentiating the effects of status and power: a justice perspective

Few empirical efforts have been devoted to differentiating status and power, and thus significant questions remain about differences in how status and power impact social encounters. We conducted 5 studies to address this gap. In particular, these studies tested the prediction that status and power would have opposing effects on justice enacted toward others. In the first 3 studies, we directly compared the effects of status and power on people's enactment of distributive (Study 1) and procedural (Studies 2 and 3) justice. In the last 2 studies, we orthogonally manipulated status and power and examined their main and interactive effects on people's enactment of distributive (Study 4) and procedural (Study 5) justice. As predicted, all 5 studies showed consistent evidence that status is positively associated with justice toward others, while power is negatively associated with justice toward others. The effects of power are moderated, however, by an individual's other orientation (Studies 2, 3, 4, and 5), and the effects of status are moderated by an individual's dispositional concern about status (Study 5). Furthermore, Studies 4 and 5 also demonstrated that status and power interact, such that the positive effect of status on justice emerges when power is low and not when power is high, providing further evidence for differential effects between power and status. Theoretical implications for the literatures on status, power, and distributive/procedural justice are discussed.     

Imagination inflation: a statistical artifact of regression toward the mean

In the imagination inflation procedure of Garry, Manning, Loftus, and Sherman (1996), subjects rated a list of events in terms of how likely each was to have occurred in their childhood. Two weeks later, some of the events were imagined; control events were not. The subjects then rated the likelihood of occurrence for each event a second time. Garry et al. (1996) reported that the act of imagining the target events led to increased ratings of likelihood. This finding has been interpreted as indicating that false events can be suggestively planted in memory by simply having people imagine them. The present study tests and confirms the hypothesis that the results that have been attributed to imagination inflation are simply a statistical artifact of regression toward the mean. The experiment of Garry et al. (1996) was reproduced (with some procedural changes), using younger and older adults. The results of Garry et al. (1996) were replicated; likelihood ratings for events initially rated low in likelihood did increase from Time 1 to Time 2. However, ratings for events initially rated high in likelihood decreased under the same conditions, and these results were consistent with the imagined target events, the target events not imagined, and the nontarget events.     

Irrelevant speech effects and statistical learning

Immediate serial recall of visually presented verbal stimuli is impaired by the presence of irrelevant auditory background speech, the so-called irrelevant speech effect. Two of the three main accounts of this effect place restrictions on when it will be observed, limiting its occurrence either to items processed by the phonological loop (the phonological loop hypothesis) or to items that are not too dissimilar from the irrelevant speech (the feature model). A third, the object-oriented episodic record (O-OER) model, requires only that the memory task involves seriation. The present studies test these three accounts by examining whether irrelevant auditory speech will interfere with a task that does not involve the phonological loop, does not use stimuli that are compatible with those to be remembered, but does require seriation. Two experiments found that irrelevant speech led to lower levels of performance in a visual statistical learning task, offering more support for the O-OER model and posing a challenge for the other two accounts.     

Range and regression effects in magnitude scaling

The relation between power law exponents obtained by magnitude estimation and magnitude production was studied for both loudness and perceived distance. While the results confirm the usual finding of higher values for production for relatively large stimulus ranges, just the opposite occurs when the stimulus range is short, necessitating a revision of the Stevens-Greenbaum regression principle. The relation between range and exponent was explored, both for the case in which several intensities are presented for judgment and for the simpler case of only two intensities. In both cases, a power relation was described relating stimulus ratios to judgmental ratios, with exponents containing both range-dependent and range-independent components.     

Unpacking outcome dependency: Differentiating effects of dependency and outcome desirability on the processing of goal-relevant information

Outcome dependency has been shown to have a substantial impact on impression formation. Perhaps because the outcome dependency variable is a composite variable (including variation in amount of dependency, desirability of outcome, or both), many different effects have been described by researchers. The current research unpacks the outcome dependency variable by separately examining effects of dependency (when outcome desirability is held constant) and effects of outcome desirability (when amount of dependency is held constant). Using research methods adapted from those used in the attitudes domain, variation in the amount of dependency on a target person influenced the level of objective processing of target information (when outcomes were not substantial, Experiment 1), but variation in the desirability of the outcomes influenced biases in the processing of information (when dependency was relatively high, Experiment 2). Furthermore, personal motivation mediated effects of desirable outcomes on positive biases in processing (Experiment 3).     

Differentiating the effects of anger and sadness: A perspective of spontaneous trait inference

In the present research, three experiments were conducted to examine the effects of anger and sadness on spontaneous trait inferences (STIs). Using a probe recognition paradigm, Experiment 1 revealed that angry participants made more errors in response to probes following trait‐implying behaviours than sad participants did. Using a false recognition paradigm, Experiments 2 and 3 revealed that angry participants made more errors in response to systematic pair trials than sad participants did. The three experiments provided convergent evidence that angry individuals were more inclined to form STIs than sad individuals were. The current research first demonstrated the different effects of specific negative mood states (anger vs. sadness) on STIs, providing further insight into the relationship between mood and STIs.     

Reliability and concurrent relations between the teacher version of the child behavior profile and the conners revised teacher rating scale

A total of 104 disturbed boys aged 6–11 were assessed with the teacher version of the Child Behavior Profile and the Conners Revised Teacher Rating Scale (TRS). A subsample of 55 boys were reassessed with both measures at an interval of 1 week. For the teacher Profile, test-retest correlations averaged r=87 for eight behavior problem scales and r=91 for four adaptive functioning scales. Reliabilities were slightly higher for Externalizing scales (average r=.91) than for Internalizing scales (r=.80). Reliabilities averaged .94 for the Revised TRS. Correlations between the two measures indicated a strong correspondence between the three Externalizing scales of the Profile and the factor-based TRS scales. Specifically, Profile scales labeled Inattentive, Nervous-Overactive, and Aggressive correlated highly (p < .001) with TRS scales labeled Inattentive-Passive, Hyperactivity, and Conduct Problem, respectively. The Conners Hyperkinesis Index correlated highly with all three Externalizing scales of the Profile, particularly Aggressive (r=.82). Overall, these results support the reliability and concurrent validity of both measures.This research was supported by an NIMH Research Scientist Development Award (MH00403) to the first author and by NIMH grant number MH37372. The support of the William T. Grant Foundation is gratefully acknowledged. The authors are indebted to the students, teachers, and staff of PACE School for their cooperation. Tom Achenbach provided valuable comments on an earlier draft.     

Computerized practice tests and effects on in-class exams

Twenty-eight introductory psychology college students volunteered to learn a technique that might help improve their in-class, multiple-choice exam performance. Two 12-question multiple-choice tests with detailed feedback were presented on a personal computer. There were two types of feedback.Separate feedback consisted of a brief statement indicating why a specific answer was either right or wrong.Combined feedback consisted of the four separate feedbacks combined into a single paragraph. Eighteen students from another introductory class, who did not get to participate in these practice tests, served as a control. A within-group comparison of the two feedback formats showed differences in degree of improvement dependent on the level of achievement by students on the first in-class exam. When compared with data from the control group, the results show significant improvement by the low-level participants on the second in-class exam (p<.01).     

Test reliability and practice effects

The purpose of this paper is to develop mathematical formulations of test reliability when the estimation of reliability requires successive observations and these observations are not independent of the process (or act) of measurement. The specific application of concern arises in performance testing where performance varies as a function of practice and where practice accrues as a result of measurement. Repeated measures are experimentally nonindependent in such cases. Formulations for the estimation of reliability are developed for each of two cases. The first of these assumes a linear model for acquisition; the second assumes a nonlinear (quadratic) model. Illustrative data are presented whereby insights may be gained as to the inadequacy of the models.This investigation was supported by research grant (G14437) from National Science Foundation to Oregon Research Institute.The analysis of much of the data referred to in this report was made possible through the facilities of the Western Data Processing Center at the University of California, Los Angeles.     

The effects of statistical training on thinking about everyday problems

People possess an abstract inferential rule system that is an intuitive version of the law of large numbers. Because the rule system is not tied to any particular content domain, it is possible to improve it by formal teaching techniques. We present four experiments that support this view. In Experiments 1 and 2, we taught subjects about the formal properties of the law of large numbers in brief training sessions in the laboratory and found that this increased both the frequency and the quality of statistical reasoning for a wide variety of problems of an everyday nature. In addition, we taught subjects about the rule by a “guided induction” technique, showing them how to use the rule to solve problems in particular domains. Learning from the examples was abstracted to such an extent that subjects showed just as much improvement on domains where the rule was not taught as on domains where it was. In Experiment 3, the ability to analyze an everyday problem with reference to the law of large numbers was shown to be much greater for those who had several years of training in statistics than for those who had less. Experiment 4 demonstrated that the beneficial effects of formal training in statistics may hold even when subjects are tested completely outside of the context of training. In general, these four experiments support a rather “formalist” theory of reasoning: people reason using very abstract rules, and their reasoning about a wide variety of content domains can be affected by direct manipulation of these abstract rules.     

Testing overall and moderator effects in random effects meta‐regression

Random effects meta‐regression is a technique to synthesize results of multiple studies. It allows for a test of an overall effect, as well as for tests of effects of study characteristics, that is, (discrete or continuous) moderator effects. We describe various procedures to test moderator effects: the z, t, likelihood ratio (LR), Bartlett‐corrected LR (BcLR), and resampling tests. We compare the Type I error of these tests, and conclude that the common z test, and to a lesser extent the LR test, do not perform well since they may yield Type I error rates appreciably larger than the chosen alpha. The error rate of the resampling test is accurate, closely followed by the BcLR test. The error rate of the t test is less accurate but arguably tolerable. With respect to statistical power, the BcLR and t tests slightly outperform the resampling test. Therefore, our recommendation is to use either the resampling or the BcLR test. If these statistics are unavailable, then the t test should be used since it is certainly superior to the z test.     

Sex differences and practice effects on two visual-spatial tasks

Sex differences on visual-spatial tasks have been assumed to be present in young children, and performance on visual-spatial tasks has been assumed to be resistant to modification. Third and fifth graders were pretested on embedded and successive figures. Half of the 110 children then received limited feedback after which both groups were posttested. Grade, time of testing, and type of task significantly affected visual-spatial performance. Additional study might indicate whether amount of practice influences final level of performance for girls and boys and the asymptote on visual-spatial tasks is similar.