Search

Mixed-effect models are flexible tools for researchers in a myriad of fields, but that flexibility comes at the cost of complexity and if users are not careful in how their model is specified, they could be making faulty inferences from their data. We argue that there is significant confusion around appropriate random effects to be included in a model given the study design, with researchers generally being better at specifying the fixed effects of a model, which map onto to their research hypotheses. To that end, we present an instructive framework for evaluating the random effects of a model in three different situations: (1) longitudinal designs; (2) factorial repeated measures; and (3) when dealing with multiple sources of variance. We provide worked examples with open-access code and data in an online repository. We think this framework will be helpful for students and researchers who are new to mixed effect models, and to reviewers who may have to evaluate a novel model as part of their review., Comment: ~8,000 words body text, 7 figures, 4 tables. Currently posting version 3 responding to comments on previous drafts

Previous research has indicated that Type A's have a higher need to control their environments, and respond more reactively to control loss than Type B's. To clarify the role of information processing in the control relinquishment decisions of Type A's and B's, 147 male and female college students took part in a problem solving experiment. Pairs of subjects worked independently on an initial 25 trial task and received feedback indicating that their partner had performed at a comparable or superior level. On a second task subjects combined their efforts, but only one could work on a given trial, necessitating a decision concerning how many trials each would attempt to solve. One third of the subjects made this decision prior to completing an evaluation of the initial performances. Another third completed the evaluation first without knowing they would subsequently make a control decision. The final third of the subjects comleted their evaluations knowing that a control decision would follow. Results indicated that when the evaluations were completed last, or, when the evaluations were completed first but without knowledge of the impending decision, Type A's relative to Type B's failed to relinquish control to a superior partner. When the evaluations were completed with knowledge of an impending control decision, Type A's and B's did not differ in their decisions. These results suggest that, under appropriate conditions, Type A's engage in an automatic or mindless decision strategy with potentially maladaptive consequences. (Author/JAC)

Examined factors in the decisions of 98 battered women to leave or remain in an abusive relationship. Results showed that both economic dependence and psychological commitment were significantly, and independently, related to the decision, and results held for both objective and subjective measures. (JAC)

According to Fiedler's Contingency Model of Leadership Effectiveness, group performance is a function of the leader-situation interaction. A review of past validations has found several problems associated with the model. Meta-analytic techniques were applied to the Contingency Model in order to assess the validation evidence quantitatively. The 33 tests used by Fiedler to derive the model and 127 subsequent tests of the model's validity were examined. Results of the meta-analytic computed on Fiedler's original studies indicated overall support for Fiedler's derivation. Analysis of Fiedler's 1978 review of the validation evidence for the Contingency Model found overall support for the model when field and laboratory studies were combined. Analysis of other direct tests not included in Fiedler's review also provided overall support for the model. Analyses of extension studies and investigations which examined the model in relation to training and co-acting groups indicated strong support for the model. The model was found to be an excellent predictor of group performance. The use of meta-analytic techniques appears to identify the directions that theory and research should take in the development of the Contingency Model. (NRB)

Research has documented two behavior patterns known as Type A and Type B. Type A behavior is characterized by competitive striving, a sense of urgency, and hostility, while Type B behavior is less aggressive and more relaxed. It has been theorized that individuals exhibiting Type A behavior have a strong need to maintain control over their environment and that they react more strongly than other individuals when they experience threat. Two studies were conducted to investigate nonverbal responses by Type A individuals to interpersonal threat. In the first study, 80 college students participated in a sales transaction and assumed the role of either a customer or a salesperson. The customer role provided a situation wherein students expected a threat to their behavioral freedom, while the salesperson role lacked such a threat. The results indicated that an interpersonal interaction characterized by threat to freedom can produce nonverbal reaction. Overall, the subjects in the customer and salesperson roles used nonverbal behaviors in separate and distinct ways. More importantly, Type A and Type B individuals differed in the use of personal space and interpersonal distance while in the customer role--a role that subjects in the second study confirmed was threatening. (FL)

A study investigated the use of nonverbal cues by male and female subjects during relatively pleasant and unpleasant interactions. Five nonverbal behaviors found to be effective in controlling both the quantity and the quality of interactions were examined: gaze, smiling, arm position, interpersonal distance, and personal space. It was expected that females would exhibit more appropriate use of nonverbal cues, but that this advantage would be limited to a favorable, open interaction. The 40 male and 40 female college students participating in the study were instructed to establish a pleasant or to avoid an unpleasant interaction with a male or female stranger. Results showed that women used nonverbal cues more appropriately and in more congruent patterns than did men, but only when endeavoring to be pleasant. Men were somewhat more appropriate and congruent in their nonverbal cues when avoiding unpleasant interaction. The results qualified previous findings that women were better encoders of nonverbal information and suggested that the purpose of the communication is an important mediator. (RL)

Past research suggests that Type A's have a higher need for control than Type B's, and empirical evidence documents their greater reactivity to control loss. To extend investigation to control decisions and examine the hypothesis that Type A's would be less willing than Type B's to relinquish control to another person, 160 male undergraduates were classified as Type A's or Type B's based on the Jenkins Activity Survey, Form T. On an initial 20-trial reaction-time task, Type A and B subjects received feedback about their own performance, and that of a partner. The feedback indicated equal or superior performance by the partner. On a subsequent replication of the task where only one person could work on any one trial, Type A's, relative to Type B's, were found to relinquish fewer trials to their partners, particularly when the partner had exhibited a superior initial performance. Attribution data indicated that Type A's were less convinced of the diagnosticity of their partners' initial performance with respect to its likelihood of replication, justifying lower relinquishment. Thus, in a situation where competence information and control needs conflicted, Type A's appeared to be less rational decision-makers than Type B's; their control needs appeared to override the most effective use of available information. Additional attribution data suggested, however, that Type A's made decisions which they believed would lead to successful task performance. Results indicated clear differences in control relinquishment by Type A's and B's as a function of feedback about their partner's performance, but no such differences emerged with respect to feedback about the subject's own performance. (Author/JAC)

Past research has established clear behavioral differences between Type A and Type B individuals. The manner in which these behavioral differences are represented in the self-definitions of Type As and Bs was examined in a group of 42 college students. The existence of Type A and B self-schemata was investigated using two tasks designed to measure the influence of these hypothetical structures on speed of processing and memory interference. During an initial task, Type As and Bs made self-relevant decisions (Like Me, Not Like Me) in response to trait adjectives previously scaled as Type A, Type B, or neutral in content. Reaction times for the decisions were measured and results indicated that both Type As and Bs made faster decisions for schema-compatible responses than for schema-incompatible responses. On a second task, Type As and Bs were tested for recognition memory after attempting to memorize half of the aforementioned trait list. Memory errors were examined and indicated that Type As and Bs made more errors compatible with their respective self-schemata. These results indicate that a Type A and B distinction forms a reliable organizing framework for the self-definitions of Type As and Bs. (Author/NRB)

Researchers have begun to recognize that because Type A and Type B behavior patterns represent pervasive lifestyles, they may have general implications for various psychological phenomena. One area of interest concerns the influence of Type A and B behavior on the quality and functioning of interpersonal relationships. A study was conducted to investigate the influence of Type A and B behavior by members of dating couples on their relationship satisfaction, commitment, and long-term intentions. Thirty-one dating couples, all university students, completed questionnaires measuring Type A or B behavior patterns, partner's behavior patterns, quality of relationship, and dating history. Subjects also completed four biweekly questionnaires measuring conflict and conflict resolution, attribution of blame, quality of communication, and relationship satisfaction and commitment. Results indicated that partners with contrasting behavioral orientations were more committed to their relationships, and had higher intentions to marry their partners, than did couples where partners had similar behavior styles. One contributing factor to this difference appeared to be the manner in which couples resolved conflict. Couples with contrasting behavioral styles were less likely than couples with similar behavioral styles to use a confrontational style of conflict resolution. Type A females were more likely than Type B females to blame partners for relationship conflict, suggesting that the Type A self-serving bias may be disruptive in interpersonal relationships. (Author/NB)

Past research has produced conflicting results concerning the manner in which Type As and Bs make attributions following success and failure. Some studies find that Type As are more likely than Type Bs to blame themselves for all outcomes, particularly failure. Other research indicates that Type As are more self-serving in their post-performance attributions. Research indicating more self-blame by Type As typically uses a causal source approach, that is, subjects are asked to indicate the extent to which common causal sources (e.g., ability, effort) produced the outcome. Assumptions are then made about the causal dimensions (e.g., internality, stability) that underlie the attributions. The causal dimension approach, used in studies that find Type As to be more self-serving than Type Bs, assesses the dimensional structure of causes directly. One possible explanation for the divergence in results is that Type As and Bs differ in their interpretations of the dimensional structure underlying common causal sources, with Type As interpreting those causes in a more self-serving manner. To test this hypothesis, 76 undergraduates classified as either Type As or Type Bs were induced to succeed or fail at anagram solving problems and then completed an attribution scale where the causal sources of ability and effort were rated on three dimensions (internality, stability, generality). Results provided support for the hypothesis in that Type As interpreted the dimensional structure of ability and effort differently than did Type Bs and in ways that were more self-serving. (Author/NB)

The recently proposed self-appraisal model of Type A behavior holds that Type As desire an accurate appraisal of abilities, and respond to uncertainty about abilities with exaggerated attempts to generate diagnostic information. The model also predicts that the major emotional response to uncertainty for Type As is anxiety. A study was conducted in which 80 college students classified as either Type As or Type Bs were induced to succeed or fail on a task where they were either very certain or very uncertain of their subsequent ability levels. Subjects completed questionnaires, performed anagram problems, received performance feedback indicating either success or failure, and completed other questionnaires measuring mood and performance attributions. The results indicated that task failure, compared to success, produced greater depression, anger, and frustration for all subjects. Consistent with the self-appraisal model prediction, uncertainty (regardless of success or failure) led to increased anxiety for Type As, but not for Type Bs. In fact, increased uncertainty produced lower anxiety for Type Bs, suggesting that Type As and Type Bs differ dramatically in their emotional responses to uncertainty about abilities. (Author/NB)

Describes two QuickBASIC programs that provide students direct experience with interpreting correlation scatter-plots. Maintains that the programs can be used in classroom exercises to highlight factors that influence the size of a Pearson correlation coefficient. (CFR)