Factorial design and covariance in the study of individual educational development

This is the report of the application of the principles of factorial design to an investigation of individual educational development. The specific type of factorial design formulated was a 2 × 3 × 3 × 3 arrangement, that is, the effect of sex, grade location, scholastic standing, and individual order, singly and in all possible combinations was studied in relation to educational development as measured by theIowa Tests of Educational Development. An application of the covariance method was introduced which resulted in increased precision of this type of experimental design by significantly reducing experimental error. The two concomitant measures used to increase the sensitiveness of the experiment were initial status of individual development and mental age. Without these statistical controls all main effects and two first-order interactions would have been accepted as significant. With their use only sex (doubtful), scholastic standing, and individual order demonstrated significant effects. The chief beauty of the analysis of variance and covariance as an integral part of a self-contained experiment is demonstrated in the complete single analysis of the data. The statistical utilization of the experimental results has also been developed for purposes of estimation and prediction. The mathematical statistician is being continuously required to develop and analyze experimental designs of increasing complexity since the introduction of the analysis of variance and covariance. The mathematical formulation and solution of the problem of this investigation is carried out. The methods illustrated and explained in this study, and modifications and extensions of them are capable of very wide application. The general principles can be used to various degrees and in a number of ways.For the research grant to finance this study, grateful acknowledgment is given to the Graduate School, the University of Minnesota.     

Runway acquisition and extinction as a joint function of magnitude of reward and percentage of rewarded acquisition trials

To determine the joint effects of partial reward and reward magnitude on acquisition and extinction rates, and on acquisition and extinction asymptotes, 215 Wistar albino rats were trained in a Hunter straight runway. The experimental design was a 4 × 4 × 2 factorial combining four reward magnitudes, four reward percentages, and two experimenters. The data revealed that the acquisition rate was an increasing function of both percentage and magnitude of reward and that neither reward magnitude nor percentage of reward significantly affected acquisition asymptote. For extinction, it was found that, for continuous schedules, the larger the reward magnitude the less the resistance to extinction and, for partial schedules, the larger the reward magnitude the greater the resistance to extinction. These results were interpreted within the framework of the sequential effects hypothesis (Capaldi, 1966).     

The effects of sex and sex-role concept on self-disclosure

Socialization into traditional sex-role concepts has been offered as an explanation of sex differences in self-disclosure. The present experiment tested a derivation of this hypothesis: Androgynous males would self-disclose more intimately than sex-typed males. A 2×2×2×2 factorial design was used with sex of subject, sex-role concept of subject, sex of target person, and disclosure level of target person as the independent variables. The results failed to demonstrate an overall difference between androgynous and sex-typed males, while significant sex differences and a strong reciprocity effect were obtained. With both male and female subjects, androgynous subjects did disclose more than sex-typed subjects when the target person had exhibited high disclosure. The findings are discussed in terms of two hypotheses: (1) compliance to a social norm regarding male disclosure to strangers; (2) flexibility of the disclosing behavior of androgynous subjects.An earlier version of this paper was presented at the meeting of the Eastern Psychological Association, Washington, D.C., March 1978. The authors are indebted to two anonymous reviewers for their suggestions concerning this article.     

Effects of group decision rules on decisions involving continuous alternatives: The unanimity rule and extreme decisions in mock civil juries

Two experiments examined the effects of majority and unanimity decision rules on group decisions involving continuous alternatives. Each experiment used mock civil juries to test three hypotheses related to the claim that when the distribution of group members’ initial preferences is skewed, the group’s decision will be influenced more by the presence of outlier members under unanimity rule than under majority rule. Experiment 1 tested the hypotheses using a punitive damages case and positively skewed preference distributions. Experiment 2 tested the hypotheses using a compensatory damages case and manipulating the skewness of preference distributions in a 2 (group decision rule) × 2 (valence of skewness) factorial design. In both experiments, the extreme (outlier) member’s individual preference was significantly correlated with the group decision only under unanimity rule. A simple thought experiment based on the experimental results is suggested regarding the implications of using unanimity or majority rule for making decisions in civil juries.     

Exposure and response prevention in the natural environment: A comparison with obsessive-compulsive patients

Thirteen obsessive-compulsive clients were treated with different methods of response prevention. Four experimental conditions were obtained through a 2 × 2 factorial design, in which the two main variables were varied: (1) gradual treatment vs flooding; and (2) modeling vs no-modeling. Clients were treated at their homes, three times weekly. Each session lasted 2 hr. Results at the post-test revealed, that modeling was superior to no-modeling (p < 0.05) on avoidance-other compulsions, therapist-rating only. On anxiety-main compulsions, clients rating the gradual method of treatment had been superior to flooding (p < 0.05). No other significant differences were found.     

Looking time: Experimenter and instruction effects

A 2 by 3 factorial design was used to investigate the effects of the instructions and the presence of E on looking time. Two hundred thirty-nine college students viewed 30 colored stimuli of landscapes, single objects, and arrays of objects under three different sets of instructions, with E either present or absent while S viewed the stimuli. The results demonstrated that E’s presence depressed the time spent looking at the stimuli Evidence was also found suggesting that ambiguity in the typical instructions given Ss in looking-time experiments may result in artifactually long looking times. It was concluded that a more appropriate situation to investigate looking time would (1) utilize sham GSR instructions to reduce S’s uncertainty concerning the experiment’s purpose, and (2) arrange for E’s absence while S is viewing the stimuli     

Escalation behavior in domains related and unrelated to decision makers' academic background

Based on Staw and Ross's (1987) analysis of escalation behavior it was hypothesized that (a) those who are involved with projects related to their academic background will tend to persist in a losing course of action and to judge the situation as reversible more than those who have to make decisions in areas unrelated to their specialty; (b) providing information that the loss is irreversible will decrease the tendency to invest additional resources. A secondary goal of the study was to examine whether there would be any difference in sunk cost decisions when the resource invested is financial as compared to time. Using a 2 × 2 × 2 factorial design, a total of 296 students were given eight different scenarios in which relatedness of decision subject-matter to academic background, reversibility of the situation, and type of resources (time or money) were the independent variables, and commitment to a course of action was the dependent variable. The results confirmed the hypotheses. No significant main effect for type of resources (time vs. money) was found. Theoretical and practical implications of the results are discussed.The authors thank Dr. Michael Hoffman for his valuable contribution in preparing this paper.     

Subclinical inclination toward manic-depression and creative performance on the Remote Associates Test

An experiment examined manic-depressive inclination in college students as an influence on creativity. The Bipolar: Manic Disorder scale of the Millon Clinical Multiaxial Inventory determined manic-depressive inclination or its absence. The Remote Associates Test was the basis for assessing creativity. The experiment was a 2×2 factorial design with manic-depressive inclination (presence or absence) as the first factor, and a mood-enhancing experience as the second. The mood-enhancing experience required that students write an essay about an activity they once had performed superbly well. Students assigned to the control condition wrote an essay about how they organized and sequenced activities in an ordinary day. As hypothesized, students with high Bipolar scores who wrote about a peak performance obtained especially high Remote Associates Test scores in comparison with students in other conditions of the experiment (p<0.02). High Bipolar scores did not result in high Remote Associates Test scores in the control condition, suggesting that mood enhancement is a necessary prerequisite for manic-depressive inclination to potentiate creative performance.     

Sex-role imitation in children: Effects of sex of child,sex of model,and sex-role appropriateness of modeled behavior

The qualifying influence of the sex-role appropriateness of observed behavior on children's same-sex imitation was investigated. In each of two studies employing a 2×2×2 design, girls and boys observed a live male or female model display appropriate (to the child's sex) or inappropriate behavior. Results that were consistent in both studies indicated an interaction between sex of child and appropriateness. Boys imitated less when exposed to the inappropriate than to the appropriate sequence. An interaction between sex of child and sex of observer was obtained on only one measure in Experiment I (girls responded more quickly to a female than to a male model). In this study, girls also displayed greater opposite-sex imitation than boys; in Experiment II, there were no differences between boys and girls in either same- or opposite-sex imitation. The results are discussed with regard to the same-sex hypothesis and the differential impact of sex-role factors on boys and girls.The authors wish to thank the children, parents, and staff of Forest Park Elementary School for their participation and cooperation in the conduct of this study. Special thanks are extended to Mr. Paul H. Daby, Principal. The contribution of the undergraduate models and experimenters is also acknowledged.     

Children’s eyewitness memory: A comparison of two interviewing strategies as realized by forensic professionals

A critical issue for developmental psychology is how to obtain accurate and complete eyewitness memory reports from preschoolers without offering suggestions that might result in false allegations. We examined effects of two interviewing strategies (police/verbal interviews and clinician/prop-assisted interviews) on young children’s reports about a medical examination. A total of 58 4-year-olds participated in the study, which conformed to a 2 (Interview Type) × 2 (Number of Interviews) factorial design. Analyses revealed that interviewers spent less time off topic and asked more free recall questions in the police/verbal interviews than in the clinician/prop-assisted interviews. Compared with police/verbal interviews, clinician/prop-assisted interviews resulted in significantly more correct rejections and commission errors in children’s memory reports. However, on a final free recall test, error rates were comparable across conditions. Higher child verbal intelligence predicted memory accuracy in police/verbal interviews, and greater parental attachment anxiety predicted children being asked a higher number of misleading questions. The study provides new insights into interview techniques that promote preschoolers’ accurate eyewitness reports.     

结构、任务类型和导航对超文本信息搜索的影响

通过两个实验考察了结构、任务类型和导航对超文本信息搜索绩效的影响。实验一采用2（层次结构,混合结构）×2（特定任务,关系任务）的被试内设计,探讨了超文本结构和任务类型对信息搜索的影响;实验二在实验一的基础上,采用2（层次结构,混合结构）×2（有导航图,无导航图）的被试内设计,考察超文本结构和导航对关系任务信息搜索的影响。结果表明：（1）超文本结构和任务类型对信息搜索绩效有显著的交互影响,就关系任务而言,混合结构超文本优于层次结构超文本,但就特定任务而言,两者无显著差异;（2）导航对信息搜索行为有指导作用,尤其对层次结构超文本有利;（3）两种主观指标的结果与客观指标存在一定程度的一致性。     

An exploratory field experiment of the effect of subconscious and conscious goals on employee performance

The effect of both a primed and a conscious goal on performance in the workplace was investigated. The primed goal was pilot tested. People on their way to work (n = 52) shown a photo of a woman winning a race performed better on a brainstorming task within a 2-min time period than people in the control group. A laboratory experiment (n = 71) revealed that this primed goal significantly increased the participants’ subconscious need for achievement, as measured by a projective test, relative to those in the control group. A 2 (primed goal vs. control group) × 2 (conscious goal vs. do best goal) factorial design involving call center employees (n = 81) revealed a main effect for this primed goal, as well as a main effect for a conscious goal for money solicited from donors.     

Analysis of Variance Models with Stochastic Group Weights

The analysis of variance (ANOVA) is still one of the most widely used statistical methods in the social sciences. This article is about stochastic group weights in ANOVA models – a neglected aspect in the literature. Stochastic group weights are present whenever the experimenter does not determine the exact group sizes before conducting the experiment. We show that classic ANOVA tests based on estimated marginal means can have an inflated type I error rate when stochastic group weights are not taken into account, even in randomized experiments. We propose two new ways to incorporate stochastic group weights in the tests of average effects – one based on the general linear model and one based on multigroup structural equation models (SEMs). We show in simulation studies that our methods have nominal type I error rates in experiments with stochastic group weights while classic approaches show an inflated type I error rate. The SEM approach can additionally deal with heteroscedastic residual variances and latent variables. An easy-to-use software package with graphical user interface is provided.     

Demonstration of an aftereffect occurring in the tactile-kinesthetic domain

Summary The aim of this work is the demonstration of a tactile-kinesthetic aftereffect. A sequence of stimuli was offered to adult subjects whose task was to compare two weights presented simultaneously to both hands by means of a Piéron gravimeter. In Experiment 1 the inspection stimuli consisted of two successive presentations of unequal weights for a period of 2 × 10 s. The test stimuli consisted of two equal weights rapidly following an inspection stimulus. Control stimuli consisting of two equal weights not preceded by an inspection stimulus were interspersed in the sequence. The results obtained confirm the existence of a contrast effect after presentation of an inspection stimulus. In Experiment 2 the inspection stimuli consisted of a single presentation of two unequal weights for a period of 20 s. The results confirm those of the first experiment and provide data enabling the contrast effect obtained to be interpreted as a Köhler-type aftereffect. In order to avoid confusion with the kinesthetic-figural aftereffect, we propose to call the effect demonstrated here the Gravimetric Aftereffect.     

Crossed categorization outside the lab: Findings from a factorial survey experiment

Crossed categorization studies investigate intergroup attitudes in situations when two categorization dimensions are simultaneously salient, often looking at artificial groups in laboratory settings. The current study tests (a) patterns of evaluation in crossed categorization scenarios when more than two real-life categorization dimensions are crossed, and (b) the moderating role of identity importance. We conduct a factorial survey experiment with a diverse sample (N = 524), crossing eight categorization dimensions. The results provide strong support for the additive pattern of crossed categorization, challenging the view that with an increased number of categories category-based information processing will not be used. Identity importance predicts the strength of intergroup bias only on the dimension of religion, which was the dominant dimension in this sample. The study contributes to multiple and crossed categorization literature by testing some of its key assumptions using a design that increases the ecological validity of the findings.     

The Influence of Indirect Knowledge of Previous Performance on Ratings of Present Performance: The Effects of Job Familiarity and Rater Training

In a 2 × 2 × 2 crossed factorial design, trained or untrained subjects viewed a videotape and evaluated performance on either a familiar (college lecturer) or unfamiliar (salesperson) job. Prior to viewing the videotape, some subjects reviewed positive information about the ratee's prior performance, whereas other subjects did not review any prior performance information. To determine whether assimilation or contrast effects occurred, we compared ratings provided by subjects who reviewed positive information about prior performance with ratings provided by subjects who did not review any prior performance information. A three-way interaction was obtained. Ratings of performance on the familiar job by untrained or trained subjects revealed only a small assimilation effect. However, when rating performance on the unfamiliar job, a large assimilation effect was observed among untrained subjects, whereas a large contrast effect was observed among trained subjects. The results indicate that rater error training may reverse, rather than reduce or eliminate, rating errors that arise from knowledge of a ratee's prior performance. Implications for further understanding and reducing assimilation and contrast effects are discussed.     

Anticipated evaluation and audience presence in the enhancement of dominant responses

Recent studies have shown that the presence of an audience enhances the emission of dominant responses in individual performance. The present study tested the hypothesis that anticipated evaluation is essential to this enhancement of dominant responses. Audience presence (absent or present) and anticipated evaluation (absent or present) were varied in a 2 × 2 factorial design. In each condition 18 subjects performed a pseudorecognition task, using responses based on habits of varying strengths established in prior training. The hypothesis was supported in that anticipated evaluation of performance produced greater emission of dominant responses than no anticipation of evaluation. The presence or absence of an audience did not significantly affect the emission of dominant responses.     

Effect of intentionality on willingness to reciprocate a favor

The purpose of this study was twofold: (1) to examine the influence of deliberate vs accidental help, and, to a lesser extent, high vs low help, on the magnitude of reciprocity; and (2) to clarify the role of obligation and attribution of motivation as mediators of reciprocity. Ss were assigned to one of four cells in a 2 × 2 factorial design. They were run on two tasks, the first to manipulate the two independent variables, and the second to measure magnitude of reciprocity. Both main effects were highly significant in the predicted direction. Data were obtained supporting the role of both obligation and attribution of motivation as mediating intentionality of help and magnitude of reciprocation.     

Mental rotation: effects of dimensionality of objects and type of task

The original studies of mental rotation estimated rates of imagining rotations that were much slower when two simultaneously portrayed three-dimensional shapes were to be compared (R. Shepard & J. Metzler) than when one two-dimensional shape was to be compared with a previously learned two-dimensional shape (Cooper and her associates). In a 2 X 2 design, we orthogonally varied dimensionality of objects and type of task. Both factors affected reaction times. Type of task was the primary determiner of estimated rate of mental rotation, which was about three times higher for the single-stimulus task. Dimensionality primarily affected an additive component of all reaction times, suggesting that more initial encoding is required for three-dimensional shapes. In the absence of a satisfactory way of controlling stimulus complexity, the results are at least consistent with the proposal that once three-dimensional objects have been encoded, their rotation can be imagined as rapidly as the rotation of two-dimensional shapes.