Failures to detect moderating effects with ordinary least squares-moderated multiple regression: Some reasons and a remedy.

[Correction Notice: An erratum for this article was reported in Vol 100(2) of Psychological Bulletin (see record 2008-10954-001). Several errors went uncorrected. On page 283, the second line of the first full paragraph should read "in Equation 3...." On page 284, in the eighth line of the first full paragraph, the power in the equation should be "1/2," not "12." On page 287, in Table 4, the heading for column 6 should read "Adjusted SS for deletion of X?X?," not just "X?." The heading for column 7 should read "H?: β?=0c, partial F," not "β?." Finally, in line 3 of the table note, "X?X?" should read "X?,X?."] Describes a means for determining circumstances when ordinary least squares/moderated multiple regression (OLS/MMR) may be at risk in moderator applications and suggests an alternative regression procedure that can be used to overcome the threat of Type II error posed by these circumstances. Using field study data on job satisfaction of employees at state institutions for the developmentally disabled, parallel analyses are presented to show how markedly different results can be obtained between OLS/MMR and the remedial procedure. It is suggested that investigators who have reached conclusions with the traditional OLS/MMR approach reexamine their data. (23 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Neurobehavioral and life-quality changes after cerebral revascularization": Correction to Baird et al.

Reports an error in "Neurobehavioral and life-quality changes after cerebral revascularization" by Anne Dull Baird, James I. Ausman, Fernando G. Diaz, Manuel Dujovny, Kenneth M. Adams and Mark W. Shatz (Journal of Consulting and Clinical Psychology, 1988[Feb], Vol 56[1], 148-151). The authors' names were incorrectly ordered. The correct order is as follows: Anne Dull Baird, Kenneth M. Adams, Mark W. Shatz, James I. Ausman, Fernando G. Diaz, and Manuel Dujovny. (The following abstract of the original article appeared in record 1988-21086-001.) Neuropsychological and life-quality changes were studied 6 months after carotid endarterectomy, superficial temporal artery to middle cerebral artery bypass, multiple revascularization, and vertebrobasilar revascularization procedures. These changes were compared with those in patients with recent severe spinal complaints and in patients for whom revascularization was considered but not elected. Subjects were enrolled prospectively. We did not see statistically significant differences in neurobehavioral gains in patients undergoing cerebral revascularization versus those considered but not selected for surgery and versus patients with spinal complaints. Small sample size, marked intragroup variability, and lack of randomized subject assignments may have obscured modest-to-moderate treatment effects in some individuals. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Controlling teaching strategies: Undermining children's self-determination and performance": Correction to Flink et al.

Reports an error in "Controlling teaching strategies: Undermining children's self-determination and performance" by Cheryl Flink, Ann K. Boggiano and Marty Barrett (Journal of Personality and Social Psychology, 1990[Nov], Vol 59[5], 916-924). In the article, Figure 1 on page 922 is labeled incorrectly. The labels in both the left and right panels of the figure should be reversed so that No Pressure Condition is the label for the broken lines and Pressure Condition is the label for the solid lines. The following abstract of the original article appeared in record 1991-13808-001.) Examined the hypothesis that students would show performance impairment when they were exposed to teachers who were pressured to maximize student performance level and who used controlling strategies. For this purpose, 4th-grade teachers and their students participated in a field experiment in which teachers either were pressured to maximize student performance or were told simply to help their students learn. In addition, the teaching sessions were videotaped to assess teachers' use of controlling strategies, as rated by blind coders. Following the teaching sessions, student performance on tasks initially taught by teachers as well as on a generalization task was assessed by blind experimenters. As predicted, the data indicated that students evidenced performance impairment during the subsequent testing session only when they were exposed to pressured teachers using controlling strategies. Results are discussed within the context of self-determination theory. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"School readiness and later achievement": Correction to Duncan et al. (2007).

Reports an error in "School readiness and later achievement" by Greg J. Duncan, Chantelle J. Dowsett, Amy Claessens, Katherine Magnuson, Aletha C. Huston, Pamela Klebanov, Linda S. Pagani, Leon Feinstein, Mimi Engel, Jeanne Brooks-Gunn, Holly Sexton, Kathryn Duckworth and Crista Japel (Developmental Psychology, 2007[Nov], Vol 43[6], 1428-1446). The DOI for the supplemental materials was printed incorrectly. The correct DOI is as follows: http://dx.doi.org/10.1037/0012-1649.43.6.1428.supp. (The following abstract of the original article appeared in record 2007-16709-012.) Using 6 longitudinal data sets, the authors estimate links between three key elements of school readiness--school-entry academic, attention, and socioemotional skills--and later school reading and math achievement. In an effort to isolate the effects of these school-entry skills, the authors ensured that most of their regression models control for cognitive, attention, and socioemotional skills measured prior to school entry, as well as a host of family background measures. Across all 6 studies, the strongest predictors of later achievement are school-entry math, reading, and attention skills. A meta-analysis of the results shows that early math skills have the greatest predictive power, followed by reading and then attention skills. By contrast, measures of socioemotional behaviors, including internalizing and externalizing problems and social skills, were generally insignificant predictors of later academic performance, even among children with relatively high levels of problem behavior. Patterns of association were similar for boys and girls and for children from high and low socioeconomic backgrounds. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Gender and the evaluation of leaders: A meta-analysis": Correction to Eagly et al.

Reports an error in "Gender and the evaluation of leaders: A meta-analysis" by Alice H. Eagly, Mona G. Makhijani and Bruce G. Klonsky (Psychological Bulletin, 1992[Jan], Vol 111[1], 3-22). Some of the numbers in the Value columns of Table 1, page 11, were aligned incorrectly. The corrected version of Table 1 is provided in the erratum. (The following abstract of the original article appeared in record 1992-16290-001.) Reviews research on the evaluation of women and men that occupy leadership roles. While holding the characteristics, except for sex, constant and varying the sex of the leader, these experiments investigated whether people are biased against female leaders and managers. Although this research showed only a small overall tendency for Ss to evaluate female leaders less favorably than male ones, this tendency was more pronounced under certain circumstances. Specifically, women in leadership positions were devalued relative to their male counterparts when leadership was carried out in stereotypically masculine styles, especially when this style was autocratic or directive. Also, the devaluation of women was greater when leaders occupied male-dominated roles and when the evaluators were men. Findings are interpreted from a perspective that emphasizes the influence of gender roles within organizational settings. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Ocular fixation control as a function of age and exposure duration": Correction to Kosnik et al.

Reports an error in "Ocular fixation control as a function of age and exposure duration" by William Kosnik, Donald Kline, John Fikre and Robert Sekuler (Psychology and Aging, 1987[Sep], Vol 2[3], 302-305). In the aforementioned article, the following corrections should be made: 1. The title of Table 1 should be changed to Mean Bivariate Areas (min-arc2) and Mean Horizontal and Vertical Standard Deviations (min-arc) of Fixations of Older and Younger Groups. 2. The equation on page 304 should have used the natural log rather than the log base 10. The corrected equation is provided in the erratum. (The following abstract of the original article appeared in record 1988-01066-001.) In previous work we reported that fixation stability did not deteriorate in older adults over relatively long viewing durations. In the present study we reanalyzed the data to examine potential aging effects on fixational control for viewing durations typically used in psychological experimentation. Monocular eye movements were recorded in 12 older and 12 younger observers using a dual Purkinje image technique, while observers fixated a stationary target. The two-dimensional scatter of eye positions was measured during nine viewing durations ranging from 100 ms to 12.8 s. Fixational control of the two groups was comparable at all of the viewing durations. Both younger and older observers were able to maintain fixation within an area several times smaller than the size of the fovea. Implications for aging studies that use briefly presented visual stimuli are discussed. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Previous experience disrupts atropine-induced stereotyped 'trapping' in rats": Correction to DeVietti et al.

Reports an error in "Previous experience disrupts atropine-induced stereotyped 'trapping' in rats" by Terry L. DeVietti, Sergio M. Pellis, Vivien C. Pellis and Philip Teitelbaum (Behavioral Neuroscience, 1985[Dec], Vol 99[6], 1128-1141). Three lines were erroneously repeated. In the first paragraph on p. 1138, the third sentence should read as follows: "Further, at the high doses that result in stereotypic trapping there are reliable changes in both cortical and hippocampal activity (Schallert et al., 1980; Shoham, Chen, DeVietti, & Teitelbaum, 1985; Vanderwolf, 1975)." (The following abstract of the original article appeared in record 1986-13974-001.) Three experiments with 64 Long-Evans hooded rats investigated atropine-induced stereotypic trapping reported by T. Schallert et al (see record 1981-07548-001). Exps I-II showed that such trapping was disrupted by previous experience with the specific trapping task or the test context alone. Exp III showed that, in response to the test context, specific behaviors were altered in Ss that were experienced with the context. Inexperienced Ss treated with intraperitoneal atropine (60 mg/kg) moved slowly and showed a strong thigmotaxis to surfaces with the body, particularly the snout. The hindquarters did not cooperate well with the movements of the forequarters. In contrast, atropine-treated Ss familiar with the context moved with medium-speed, coordinated movements, and showed a reduced tendency for surface contact with body and snout. These reactions of drugged Ss were exaggerated forms of those of undrugged Ss to the unfamiliar and familiar context, respectively. Results indicate that stereotypic trapping develops as a consequence of an interaction between the adaptive responses of the rat to a novel environment and atropine. (37 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Stable negative social exchanges and health": Correction to Newsom et al. (2008).

Reports an error in "Stable negative social exchanges and health" by Jason T. Newsom, Tyrae L. Mahan, Karen S. Rook and Neal Krause (Health Psychology, 2008[Jan], Vol 27[1], 78-86). In the aforementioned article, there is an error in the Appendix. The labels for the factors Unsympathetic or insensitive behavior and Rejection or neglect were transposed. The revised table, listing the items correctly under each factor, is presented in the erratum. (The following abstract of the original article appeared in record 2008-00647-011.) Negative social exchanges with family, friends, and neighbors are known to be an important source of stress in daily life, and chronic stress is theorized to have especially potent impacts on health. Little is known about the health effects of stably high levels of negative social exchanges, however. In a national, longitudinal study of older adults (N = 666), we examined the association between stable negative social exchanges and health over a 2-year period. Trait-state-error models indicated that higher levels of stable negative social exchanges were significantly predictive of lower self-rated health, greater functional limitations, and a higher number of health conditions over 2 years after controlling for initial levels of health and sociodemographic variables. These results highlight the importance of examining continual and recurring interpersonal problems in efforts to understand the health effects of social relationships. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"How good is the evidence for a production deficiency among learning disabled students?": Correction to Shepherd et al.

Reports an error in "How good is the evidence for a production deficiency among learning disabled students" by Margaret J. Shepherd, Lynn M. Gelzheiser and Roberta A. Solar (Journal of Educational Psychology, 1985[Oct], Vol 77[5], 553-561). Figures 1 and 2 (p. 557 and 559, respectively) are reversed. The captions are correct, but Figure 1 should be above the caption for Figure 2 and Figure 2 should be above the caption for Figure 1. (The following abstract of the original article appeared in record 1986-14779-001.) Investigated the spontaneous use of mnemonic strategies by learning disabled (LD) and non-LD children and adolescents to examine whether LD Ss can be distinguished from their non-LD peers on the basis of strategy use and recall. In Exp I, 105 LD and 105 non-LD 9-15 yr olds were administered a picture study/recall task, in which the strategies of interest were categorical organization during study and clustering during recall. In Exp II, 140 LD and 140 non-LD 11-17 yr olds were administered a paired-associate recall task, in which the strategy of interest was elaboration. In both studies, LD Ss earned lower mean recall scores than did the non-LD Ss. As a group, LD Ss did not differ from non-LD Ss in the use of categorical organization during study but showed less categorical clustering at recall. Fewer LD Ss used elaboration. Despite these differences, recall and strategy use were not useful predictors of classification as LD or non-LD and were only weak to moderate correlates of academic achievement. (24 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Child and Parent Weight Loss in Family-Based Behavior Modification Programs": Correction to Epstein et al.

Reports an error in the original article by Leonard H. Epstein, Rena R. Wing, Randi Koeske, Frank Andrasik, and Deborah J. Ossip (Journal of Consulting and Clinical Psychology, 1981, 49, 674-685). There are several errors in Table 3. The corrected table is provided. (The following abstract of this article originally appeared in record 1982-01694-001.) Overweight preadolescents and parents from 76 families were assigned to 1 of 3 behavioral treatment groups: parent/child target, child target, or nonspecific target. Percent overweight changes at the end of the 8-mo treatment and 13-mo follow-up were equivalent for children in the 3 treatment groups, but parents in the parent/child group lost more weight during treatment. Weight changes for parents and children increased through the 8-mo treatment. Patterns of maintenance showed that 40% of the children were within 20% of their ideal weight by the end of treatment, achieving nonobese status. 100% of the newly nonobese children in the parent/child group maintained nonobesity during follow-up, whereas only 30% in the child-alone group and 33% in the nonspecific group maintained nonobesity. In addition, weight losses for parents and children in the same family correlated positively at the end of treatment but not at follow-up. Results suggest that parental modeling may be important during treatment, but long-term results probably are due to parental reinforcement of child self-regulation. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Perceived parental social support and academic achievement: An attachment perspective": Correction to Cutrona et al.

Reports an error in "Perceived parental social support and academic achievement: An attachment theory perspective" by Carolyn E. Cutrona, Valerie Cole, Nicholas Colangelo, Susan G. Assouline and Daniel W. Russell (Journal of Personality and Social Psychology, 1994[Feb], Vol 66[2], 369-378). This article, which appeared in the Personality and Individual Differences section, was accepted for publication by Guest Editor Irwin Sarason. We wish to thank Professor Sarason for his help and to apologize for our oversight in not acknowledging this contribution. (The following abstract of the original article appeared in record 1994-31441-001.) Tested the extent to which parental social support predicted college GPA among undergraduates. A sample of 418 undergraduates completed the Social Provisions Scale--Parent Form (C. E. Cutrona; see record 1990-01422-001) and measures of family conflict and achievement orientation. American College Testing (ACT) Assessment Program college entrance exam scores (American College Testing Program, 1986) and GPA were obtained from the university registrar. Parental social support, especially reassurance of worth, predicted college GPA when controlling for academic aptitude (ACT scores), family achievement orientation, and family conflict. Support from parents, but not from friends or romantic partners, significantly predicted GPA. Results are interpreted in the context of adult attachment theory. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Visual analysis of multiple baseline across participants graphs when change is delayed": Correction to Lieberman et al. (2010).

Reports an error in "Visual analysis of multiple baseline across participants graphs when change is delayed" by Rebecca G. Lieberman, Paul J. Yoder, Brian Reichow and Mark Wolery (School Psychology Quarterly, 2010[Mar], Vol 25[1], 28-44). There were several errors in the text. The corrected text is provided in the erratum. (The following abstract of the original article appeared in record 2010-05458-003.) A within-subjects group experimental design was used to test whether three manipulated characteristics of multiple baseline across participants (MBL-P) data showing at least a month delayed change in slope affected experts’ inference of a functional relation and agreement on this judgment. Thirty-six experts completed a survey composed of 16 MBL-P graphs. Graphs with steep slopes, once change began, were far more likely to be judged as showing a functional relation. Generally, experts disagreed with each other regarding functional relation judgments. Implications for the types of dependent variables that fit the requirements of MBL-P are discussed. (PsycINFO Database Record (c) 2011 APA, all rights reserved)     

"Double-blind photoarray administration as a safeguard against investigator bias": Correction to Phillips et al. (1999).

Reports an error in "Double-blind photoarray administration as a safeguard against investigator bias" by Mark R. Phillips, Bradley D. McAuliff, Margaret Bull Kovera and Brian L. Cutler (Journal of Applied Psychology, 1999[Dec], Vol 84[6], 940-951). On page 947, the graphs within Figure 1 were incorrectly labeled. The corrected figure appears in this erratum. (The following abstract of the original article appeared in record 1999-15531-009.) This experiment examined whether a photoarray administrator's knowledge of a suspect's identity increased false identification rates. Fifty participant–administrators (PAs) presented 50 participant–witnesses (PWs) two perpetrator-absent photoarrays following a live staged crime involving two perpetrators. For one photoarray per trial, the experimenter revealed the suspect's identity to the PA. Each PA presented the photoarrays sequentially or simultaneously in the presence or absence of an observer. When the observer was present, PA knowledge of the suspect's identity had a biasing effect in sequential photoarrays only. This pattern did not emerge when the observer was absent. The experimental manipulations did not affect PAs' and PWs' ratings of photoarray fairness or PWs' ratings of pressure to make an identification. These data suggest that only administrators who are blind to the suspect's identity should present sequential photoarrays. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Relation between mastery behavior in infancy and competence in early childhood": Correction to Messer et al.

Reports an error in "Relation between mastery behavior in infancy and competence in early childhood" by David J. Messer, Mary E. McCarthy, Susan McQuiston, Robert H. MacTurk, Leon J. Yarrow and Peter M. Vietze (Developmental Psychology, 1986[May], Vol 22[3], 366-372). In the article, an incorrect copyright note has been given. The copyright note has been corrected and is included in the erratum. (The following abstract of the original article appeared in record 1986-24138-001.) 53 infants were observed at 6 and 12 mo of age during 2 24-min play sessions. The Bayley Scales of Infant Development (BSID) were given at 6 and 12 mo and the McCarthy Scales of Children's Abilities (MSCA) at 30 mo of age. Results reveal that measures of competence in infancy (successful task completion during play and the BSID scores) were not strongly correlated with the 30-mo MSCA scores. In contrast, infant mastery behavior during play strongly predicted MSCA scores: The time spent investigating toys at 6 mo and persistence in solving tasks at 12 mo of age were behaviors significantly positively correlated with the MSCA scales. It is suggested that infant behaviors that predict later competence do not remain static but change with age and that infants' mastery behavior is a more effective predictor of later development than their competence with either toys or developmental tests. (23 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Task planning and energy expended: Exploration of how goals influence performance": Correction to Earley et al.

Reports an error in "Task planning and energy expended: Exploration of how goals influence performance" by P. Christopher Earley, Pauline Wojnaroski and William Prest (Journal of Applied Psychology, 1987[Feb], Vol 72[1], 107-114). Table 1 on page 109 appeared with reversed labels for the high and low conditions. The corrected Table 1 is provided in the erratum. (The following abstract of the original article appeared in record 1987-17299-001.) Although many studies demonstrate the effectiveness of goal setting in organizations, it is unclear how goals actually influence performance. The present studies examined the effect of assigned goals and task information on performance, energy expended, and task planning or organizing. In Study 1, a 2×2 (Information×Goal) design was used. Subjects were 72 undergraduates working on a business simulation. In Study 2, 129 male and female workers from a service organization and a moving company responded to a survey assessing an individual's goal setting, job training, energy expended during a typical task performance, and task planning undertaken prior to performance. The results of both studies demonstrated that goal setting and task training influenced the dependent variables. In addition to influencing an individual's energy expended (effort and persistence), a specific goal led an individual to plan and organize more than did an individual given a general goal (i.e., "do your best"). The results of both studies suggest that goal setting and task-relevant information influence performance, in part, through their influence on energy expended and planning. (22 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Random monotone data fit simple algebraic models: Correlation is not confirmation": Correction to Parker et al.

Reports an error in "Random monotone data fit simple algebraic models: Correlation is not confirmation" by Scott Parker, Jay Casey, John M. Ziriax and Alan Silberberg (Psychological Bulletin, 1988[Nov], Vol 104[3], 417-423). Information was inadvertently left out of the author note on page 417. The author note should include the following information: Portions of this work were reported at the 1986 meeting of the Eastern Psychological Association. We thank Robert Jernigan for telling us about Filiben's and Stigler's work. We thank Willard Larkin, John A. Nevin, and several anonymous reviewers for helpful reviews of earlier versions. John M. Ziriax is now at the Department of Radiation Biology and Biophysics, University of Rochester School of Medicine, Rochester, New York. This research was supported in part by grants from the National Science Foundation. (The following abstract of the original article appeared in record 1989-10685-001.) Evaluation of the fit of a data set to an algebraic model often relies on the squared correlation coefficient (r-2). Algebraic models commonly prescribe functions between two variables that are monotonic. We created data sets that are monotonic but otherwise random and computed their r-2 values. These values are shown to far exceed those for data sets not constrained to be monotonic. We tabulated selected cumulants of the distribution of r-2 for monotone data fit to two common models, linear and power functions, for several conditions of randomness for data sets comprising 3, 9, and 15 points. Random monotone data fit all these functions well. The common experimental practices of averaging several sets of data and using regular spacing in the values of the independent variable both tend to produce slight additional improvements in fit. Consideration of these results reveals that correlation alone is inadequate to test the fit of monotone data to algebraic models. Some helpful auxiliary techniques are recommended. (PsycINFO Database Record (c) 2010 APA, all rights reserved)