Scaling Up the TREC Collection

Due to the popularity of Web search engines, a large proportion of real text retrieval queries are now processed over collections measured in tens or hundreds of gigabytes. A new Very Large test Collection (VLC) has been created to support qualification, measurement and comparison of systems operating at this level and to permit the study of the properties of very large collections. The VLC is an extension of the well-known TREC collection and has been distributed under the same conditions. A simple set of efficiency and effectiveness measures have been defined to encourage comparability of reporting. The 20 gigabyte first-edition of the VLC and a representative 10% sample have been used in a special interest track of the 1997 Text Retrieval Conference (TREC-6). The unaffordable cost of obtaining complete relevance assessments over collections of this scale is avoided by concentrating on early precision and relying on the core TREC collection to support detailed effectiveness studies. Results obtained by TREC-6 VLC track participants are presented here. All groups observed a significant increase in early precision as collection size increased. Explanatory hypotheses are advanced for future empirical testing. A 100 gigabyte second edition VLC (VLC2) has recently been compiled and distributed for use in TREC-7 in 1998.     

Doing a science project: Gender differences during childhood

By adolescence, men's participation and achievement in science exceeds women's. This article reports a case study that examined the beginnings of this gender differentiation during a naturally occurring academic activity that was designed to support and guide young children's interest in doing science. Data were collected during 2 successive years of a science fair for children in Grades 1–6 of a progressive private school. A total of 268 projects were characterized in terms of achievement and area of science. Parents provided information about the way children selected and created projects. In both years and in all grade levels, boys tended to choose to work in the physical sciences, and girls in the biological and social sciences. Peer collaborations were exclusively same sex. Achievement and parental involvement were not gender related. Factors are discussed that might lead to an early divergence of boys' and girls' interests in science within a context that promotes its exploration. © 1998 John Wiley & Sons, Inc. J Res Sci Teach 35: 845–857, 1998.     

The prediction of first-grade reading achievement from WPPSI scores of preschool children

The purpose of this study was to measure the relationship between the WPPSI IQs and subtest scaled scores in preschool children and their reading achievement in grade one. Twenty-eight suburban middle class children were given the WPPSI at ages three to five and retested with the Gray Oral Reading Test near the end of first grade. Significant correlations between WPPSI IQs and reading were found. These correlations were similar in magnitude to those obtained between WPPSI IQs of kindergarten children and grade one reading. Further study of the predictive power of the Geometric Designs and Arithmetic subtests was suggested on the basis of correlations obtained between these subtests and later reading scores.     

The effect of familiarity with the examiner on WISC-r verbal,performance, and full scale scores

The Wechsler Intelligence Scale for Children-Revised (WISC-R) was administered to 72 elementary school children (36 first graders and 36 fifth graders). Three methods of pretest rapport establishment were compared for each of two age groups. Treatment Group 1 received less than 5 minutes of pretest rapport establishment immediately preceding testing. Treatment Group 2 received 15 minutes of rapport establishment immediately preceding testing. Treatment Group 3 received 5 minutes of rapport establishment on two of the five school days prior to testing and 5 minutes immediately preceding testing. It was found that time spent in establishment of familiarity with the examiner as determined by the three treatment conditions had no significant effect on WISC-R Verbal, Performance, or Full Scale scores for this sample. Also, no significant difference was found between 6-year-old and 10-year-old students under the various treatment conditions. Finally, no significant differences on WISC-R Verbal, Performance, or Full Scale scores were found as a result of the interaction of age and familiarity conditions.