AN EXAMINATION OF THE CONTEXT EFFECT IN ITEM SAMPLING

Item sampling and/or multiple matrix sampling techniques have been recommended for a variety of purposes. For some of these purposes, it must be assumed that examinee performance on a set of items is unaffected by the conditions under which the items are taken (i.e., no context effect exists). In this paper factors that may lead to a context effect among high school students are discussed. The net effect of such factors on examinee scores for an English test and a mathematics test is investigated empirically. For the English test there was little support for the existence of a context effect, However, a definite context effect was found for the mathematics test.     

The evaluation of faculty development programs

This article reviews literature pertinent to the evaluation of faculty development programs and presents data from several studies conducted at two institutions. These data were consistent with those previously reported in that faculty participants consistently expressed satisfaction with development services. In addition, one study found that most faculty members voluntarily took some action to improve their instructional effectiveness, though only a minority pursued these efforts in depth. Volunteers who worked intensively with a faculty development consultant improved more on objective measures of effectiveness than did those who were only superficially involved in improvement efforts; those who received no consultative assistance failed to improve significantly. Evidence from a final study provided a control for faculty motivation and led to the conclusion that improvement was contingent both on faculty desire to improve and on the availability of professional assistance.     

PREPARING THE EDUCATIONAL TECHNOLOGIST

The author describes the course for the Diploma in Applied Educational Technology at the University of York. with particular reference to the programme of practical attachments and visits.     

Concurrent validity of the test of nonverbal intelligence with learning disabled children

Results of the TONI, WISC-R, and WRAT were compared for a sample of 66 learning disabled children: 51 males (32 white, 19 black) and 15 females (9 white, 6 black) whose mean age was 9–5 (SD = 1–10). The mean score of the TONI was significantly different from the Performance IQ. Nonsignificant differences were found between the TONI and Full Scale IQ and between the TONI and Verbal IQ. Correlation coefficients between the TONI and WISC-R ranged from a low of .35 for the Verbal IQ to .44 for both the Full Scale and Performance IQs. The correlation coefficients between the TONI and standard scores of the WRAT were .38, .27, and .23, for Reading, Spelling, and Arithmetic, respectively. Implications of these findings are discussed.     

Associations among Adult Attachment Representations, Maternal Sensitivity, and Infant-Mother Attachment in a Sample of Adolescent Mothers

Associations among adolescent attachment organization, maternal sensitivity, and infant attachment organization were examined prospectively in 74 teenaged mother-infant dyads. Pregnant teenagers' attachment organizations predicted both sensitivity and infant-mother attachments. Mothers classified autonomous (F) in the prenatal period showed higher levels of sensitivity at both 3 and 9 months than mothers classified dismissing (Ds), preoccupied (E), or unresolved (U). Correspondence between maternal attachment (F vs. Ds/E/U) and infant attachment (secure [B] vs. avoidant [A]/resistant [C]/ disorganized [D]) was observed in 58 of 74 (78%) dyads. Exact 4-group (Ds/E/F/U and A/B/C/D) agreement was observed in 50 of 74 (68%) families. In contrast, associations between maternal sensitivity and infant attachment were not significant, leading to questions about the processes that link attachment representations, maternal behavior, and infant attachment in adolescent mothers.     

Criminal justice outcomes of prosecution of child sexual abuse: A case flow analysis

This study examined the progress and disposition of child sexual abuse cases referred for prosecution in four urban jurisdictions. Most cases were accepted for prosecution. The vast majority of cases carried forward for prosecution resulted in guilty pleas. Only 9% of the total sample went to trial. A majority of the small number that went to trial were convicted. Over 3/4 of those convicted were incarcerated. The rate at which cases were carried forward for prosecution and convicted was comparable to that of a national sample of felony arrests overall, but child sexual abuse cases were more likely to go to trial (p < .005) and received more severe sentences (p < .005). Our results are strikingly consistent with those from previous studies of prosecution of child sexual abuse. A new understanding of prosecution of child sexual abuse is recommended that takes into account the large proportion of cases that do not go to trial.