Simple Main Effects Analysis Research Articles

Analysis of Variance Frameworks in Clinical Child and Adolescent Psychology: Issues and Recommendations

Reviewed existing practices of factorial analysis of variance (ANOVA), a major analytic tool used in clinical child and adolescent psychology, in the Journal of Clinical Child Psychology (JCCP) and noted several suboptimal strategies. Issues surrounding the analysis of multiple outcome variables, omnibus F tests, and single degree of freedom contrasts, simple main effects analysis, and single degree of freedom interaction contrasts were considered and recommendations were made about analytic strategies. Among the practices questioned were the use of multivariate analyses of variance (MANOVAs) as a means of controlling Type I errors across multiple outcome variables and the use of simple main effects analysis to elucidate the nature of interaction effects.

Effects of Matching and Mismatching Students' Time-of-Day Preferences

The effects of matching and mismatching students' time-of-day preferences with their performance on an algebra test were examined. A model based on the identification of each student's time-of-day learning-style characteristic and congruent and incongruent testing times was used. A 2-factor analysis of variance was conducted and followed with simple main effects analyses. On a morning-administered test, students who studied in the morning scored higher than (a) morning-preferenced students who took the test in the afternoon, (b) students studying in the afternoon, and (c) students studying in the evening. On an afternoon-administered test, students studying in the afternoon scored higher, at a level approaching statistical significance, than those who studied in the morning. Implications of the findings are suggested.

Effect of Selected Recruiting Strategies on Beginning Instrumentalists' Participation Decisions

The purpose of the study was to explore the effect of aptitude, two methods of introducing wind and string instruments, and the effect of socioeconomic status (SES) on fourth-grade students' attitude toward playing an instrument. Students ( N = 531) in nine different elementary schools were divided into two experimental groups (recorder and demonstration) and a control group. Students in the recorder group were taught to play the recorder, whereas the demonstration group encountered the instruments via films and live student performances. After 10 weeks of instruction, students completed an attitude assessment questionnaire. When low aptitude/attitude correlations were found, ANCOVA procedures were abandoned, and ANOVA procedures revealed a statistically significant interaction among the three groups and three SES levels. Simple main effects analyses indicated that the type of recruiting strategy does not significantly affect low-SES stuents' attitudes, but significant differences in both the middle- and high-SES classifications were examined using the Tukey-Kramer test.

The impact of cognitively impaired patients and shift on nursing assistant stress.

This study compared the stress experienced by nursing assistants (NAs) under four work conditions: high or low proportion of cognitively impaired patients and day or other shift. Five standard measures of caregiver stress served as the dependent variables in this study; burden, reaction to patient behaviors, workload, and two measures of burnout. A 2x2 multivariate analysis of variance found an interaction effect of type and shift on the stress measures. Univariate tests found that Burden and Depersonalization accounted for this effect. A further multivariate analysis of simple main effects found significant differences for each independent variable within each level of the other independent variable. Univariate analysis found that NAs who care for cognitively impaired patients on the day shift show significantly higher scores on specific stress measures. The article concludes with a discussion of how institutions can respond to the stresses faced by NAs who care for cognitively impaired patients.

Analysis of simple main effects in fractional factorial experimental designs of resolution v

The analysis of simple main effects is commonplace in complete factorial experimental designs when interaction effects are present. An analysis of simple main effects also will be important in fractional factorial experimental designs of Resolution V where all two-way interactions are estimable and may be non-zero. Parameter estimation and tests of hypotheses regarding simple main effects are defined for Resolution V fractional factorial experimental designs. These methods are illustrated in an example.

Electromyographic recordings of low back pain subjects and non-pain controls in six different positions: effect of pain levels

Surface electromyographic (EMG) activity recordings of bilateral paraspinal muscle tension were measured twice on 20 non-pain controls and on 46 low back pain subjects (21 individuals with intervertebral disk disorders and 25 subjects with unspecified musculoskeletal backache) during 6 positions: standing, bending from the waist, rising, sitting with back unsupported, sitting with back supported, and prone. Back pain subjects were measured during both low pain and high pain states. Results revealed a non-significant trend for all subjects, regardless of diagnosis, to have higher paraspinal muscle tension levels on the second (or high pain) assessment. A significant diagnosis by position interaction was observed which was similar to the interaction in our previous study which employed only a single measurement session. Analysis of simple main effects revealed this to be due to control subjects during the standing position having lower EMG levels than the back pain groups, and intervertebral disk disorder subjects having higher EMG levels than the other groups during the supported sitting position. As in our previous study, diagnosis was found to be a clinically significant factor, in that controls had much fewer clinically abnormal readings than back pain patients. The lack of a significant effect for pain state is congruent with findings in the headache literature. The importance of clearly defined diagnostic categories in low back pain research and the utility of measuring subjects in various positions is discussed, as are possible explanations for lack of significant pain state findings.

Factorial Design: Binocular and Monocular Depth Perception in Vertical and Horizontal Stimuli

This article describes a factorial experiment that is useful as a laboratory exercise in a research methods course. In the Howard–Dolman depth perception apparatus, two vertical rods are adjusted, using binocular or monocular vision, so they appear equidistant from the observer. The two rods can also be oriented horizontally, which allows a factorial design combining the factors of Viewing Condition (binocular and monocular) and Rod Orientation (vertical and horizontal). The exercise illustrates the nature of an interaction and the necessity of an additional analysis of simple main effects. It also provides a basis for understanding a perceptual problem in the real world—the difficulty of localizing horizontally extended stimuli such as power lines.

Electromyographic recordings of 5 types of low back pain subjects and non-pain controls in different positions

Surface EMG recordings of bilateral paraspinal muscle tension were measured on 207 subjects (29 non-back pain controls, 20 individuals with spondyloarthritis, 52 with intervertebral disk disorders, 66 with unspecified musculoskeletal backache, 17 with some combination of the above 3 groups and 23 subjects with other types of back pain, including unknown, scoliosis and psychogenic) in 6 positions: standing, bending from the waist, rising, sitting with back unsupported, sitting with back supported and prone. Results of both individual and group analyses revealed a significant main effect of diagnosis. Post hoc analyses (Duncan's) revealed controls to have significantly lower overall EMG levels than the intervertebral disk disorders and unspecified musculoskeletal backache groups. A significant diagnosis by position interaction was observed. Analysis of simple main effects revealed this to be due primarily to control subjects during the standing position having lower EMG levels than all other groups, and intervertebral disk disorder subjects having higher EMG levels than all other groups during the supported sitting position. The importance of clearly defined diagnostic categories in low back pain research and the utility of measuring subjects in various positions are discussed.

Dual-Career Couples and Geographic Transfer: Executives' Reactions to Commuter Marriage and Attitude Toward the Move

This study investigated four factors predicted to influence executive decision making concerning geographic transfers when married employees are involved. One hundred and fifty-one executives completed a series of "inbasket" decision-making tasks which described geographic transfer situations. Their ratings of geographic transfers were found to be significantly affected by the couple's attitude toward the move and the presence of commuter marriage (a marital arrangement in which spouses maintain separate households to maximize career possibilities, and reunite when schedules permit). In contrast to prior research, significant differences were not obtained for sex of employee. A number of significant interactions were observed; however, both of the above main effects held up in post hoc analyses of simple main effects. The results were discussed in terms of their implications for companies and for employees considering a commuter marriage. It was recommended that companies should show a greater awareness of transferrelated problems to reduce possible bias against employees planning a commuter marriage.

The Influence of Patellofemoral Pain on Electromyographic Activity during Submaximal Isometric Contractions

During fherepeutic exercise sessions, clinicians commonly observe inhibition d muscle activity in patients with orthopedic injury, seemingly due to protective mechanisms 50 avoid pain. As a result of these clinical obsewatjons, testing was conducted to determine if the muscle efficiency d the quadriceps fernoris muscle group differed between limbs in 37 subjects with unilateral patellofemora! pain. Efficiency was assessed by measuring the integrated elecZromyographic (IEMG) activity during submaximal isometric muscle contraction trials with 0.56 and 2.72 kg weights at 140 and 180' af extension. Analysis with a repeated measures ANOVA imcated significant inferaction of weight, joint position, and limb (p < 0.035). Post hoc analysis of fhe simple main effects of lEMG activity between the limbs with weight and joint pasition held constant demonstrated that the lEMG activity of the sxperimentaf limb was significantly higher than the control limb (p < 0.006). The study results demonstr~teth at the lEMG activrty of the painfu! limb is significantly higher and less efficient than the control limb. J Othro Sports Phys Ther 1987;9(6):211-216.

EFFECTS OF ACHIEVEMENT EXPECTATIONS AND HANDWRITING QUALITY ON SCORING ESSAYS

A search of the literature reveals numerous studies dating from the 1920s which have investigated the relationship between the quality of handwriting and scores given to essays. These studies have consistently found that essays get higher grades when written in good handwriting (e.g., James, 1927; Sheppard, 1929; Chase, 1968; Marshall and Powers, 1969; Briggs, 1970; Markham, 1976; Bull and Stevens, 1979). A few studies have investigated the influence on essay grades of marker expectations based on information regarding the writers' achievement or intelligence. These have consistently reported significant expectation effects (Bonniol, Caverni and Noizet, 1972; Dash, 1975; Petrilla, 1978; Chase, 1979). The study by Chase (1979) investigated the joint effect of marker expectation and handwriting quality on essay grades. Fabricated essay responses to questions on test theory were copied out in both very poor and very good handwriting. These responses were given to graduate students for scoring with cover sheets providing fictitious information on the supposed writer's undergraduate courses and grades for the last semester. While identical courses were listed for each writer, half the writers had uniformly high grades and half uniformly low grades to establish scorer expectations. Chase found a main effect for expectation (p < .01) but no main effect for handwriting quality. While the interaction between quality of handwriting and expectation was not significant, an analysis of simple main effects suggested the possibility that where handwriting is very poor, the effects of high expectation are accentuated. The present study has as its focus Chase's suggestion that, in essay scoring, an interaction exists between writer handwriting quality and scorer achievement expectations.

An Investigation of Fourth-Graders' Comprehension of Pronoun Constructions in Ecologically Valid Texts

An Investigation of Fourth-Graders' Comprehension of Pronoun Constructions in Ecologically Valid Texts

Note: Release from Proactive Interference with Television News Items

An adaptation of the Brown-Peterson short-term memory procedure outlined by Wickens (1, 2) for studying the encoded attributes of simple verbal items has been devised to identify the taxonomic dimensions into which individuals categorize and possibly encode television news materials. Instead of word triads, subjects are presented on each trial, with three short television news items recorded on videotape from actual network newscasts. These are carefully edited to control for length and visual format effects and shown to the subjects over a small monochrome television monitor. Following presentation of each triad of news items, the subject performs a distracting task (crossword puzzle) for 1 mia, and then is asked to free recall the news irems. During recall the subject is required to write short accounts of each item in 2 min. and is scored for the number and accuracy of these responses on each trial. All subjects had four trials: for Group 4-0 the taxonomy of the news items is held constant over all trials, but for Group 3-1 the taxonomy is constant for Trials 1-3 only, then on Trial 4 is switched to a different category. In an initial experiment, 48 subjects were shown items about political news events or items about sport. Half Group 4-0 subjects received political items across all trials and half received sports items. Likewise in Group 3-1, critical fourth trial news-category switches from political to sports items or vice versa were balanced across subiects. Responses were scored out of two points for accuracy. Raw scores were converted to proportions and transformed for computation. A two-way (Groups x Trials) splitplot factorial analysis gave a significant decrement in recall across Trials I to 4 occurring in Group 4-0 (P3.1~ = 22.84. P < 0.001), indicating a buildup of proactive inhibition over a series of news items in which taxonomy was held constant.' In Group 3-1, a decrement in performance occurred over Trials 1-3. followed by substantial improvement on Trial 4 after the critical news-category switch. Analysis of simple main effects showed that recall on Trial 4 in Group 3-1 was significantly better than that on the equivalent trial in Group 4-0 (F1.]% = 24.02, fi < 0.001), indicating a significant release from proactive interference. These early findings are consistent with prior ones (1,2) and suggest that the Brown-Peterson technique may be usefully generalized to television news as a means of identifying taxonomic categories into which individuals classify news items. The broad theoretical and practical implications of these and other findings will be discussed later.

Sex Differences in Reversal and Nonreversal Shift Acquisition

The mediational hypothesis of reversal and nonreversal shift acquisition suggests that mediational behavior is equivalent to covert verbalization and thus related to verbal ability. This coupled with the general observation that males and females manifest significant performance differences in measures of verbal fluency leads to the prediction of an interaction between sex and type of shift. 48 male and female undergraduates were tested using a simultaneous discrimination procedure with the shift manipulation occurring during the first task. A 2 × 2 factorial analysis of variance yielded significant F ratios for sex, type of shift, and the interaction of sex × shift. Analysis of simple main effects showed significant comparisons between males and females on the nonreversal shift and between the reversal and nonreversal shift performance of females. These data were interpreted as supporting a mediational hypothesis of concept-shift behavior.