This article examines possible working memory deficits in 7-8-year-olds who are accurate readers but relatively poor comprehenders. In Expt 1, poor comprehenders scored below good comprehenders on a non-linguistic test of working memory (reading series of digits and recalling the last digit in each series) on the more taxing items. Experiment 2 examined the relationship between working memory and text comprehension using an anomaly resolution task. Good and poor comprehenders heard stories describing an adult's anomalous emotional response to a child's action, some of which contained information to resolve the anomaly. The load on working memory imposed by the need to integrate the resolution with the anomaly was varied in two ways: the resolution was presented immediately next to, or two sentences distant from, the anomaly, and appeared either before or after the anomaly. Poor comprehenders were worse than good ones at anomaly resolution only when the anomalous and resolving information were separated. The results support the hypothesis that text processing is influenced by working memory demands and suggest that children's comprehension is related to the efficiency of a general non-linguistic working memory system.

The present research addressed children's understanding of self-presentational display rules: putting on false facial expressions in order to manipulate others' evaluations of the self. A sample of 4-to 6-year-olds was used to test our hypothesis that selfpresentational display rules involve recursive cognition about others' mental states. Children completed a task measuring understanding of various display rules and additionally performed a second-order false-belief task. Results supported the hypothesis that an appreciation of second-order mental representation is associated with understanding self-presentational display rules but not with understanding prosocial display rules (designed to spare others' feelings). We discuss the likely interaction of social processes with the observed changes in mental-state understanding in relation to the development of self-presentation.

Briri.vh Journulo/ Drveloptm~rital Ps.)'c.ho/ogy ( 1984) 2, 73 8 I ('; 1Y84 The British Psychological Society Pritird bi Griwi Brilriiri

No abstract

Previous studies show that 7-8-year-old poor comprehenders differ from good comprehenders, matched in age and decoding skill, primarily in their failure to make highlevel inferences, despite adequate text recall. The impact of inference awareness training on reading comprehension in two such groups was compared with the effects of two other treatments. Inference-trained children were instructed over 4 weeks in making inferences from text and generating questions. Other groups were trained either in rapid decoding or in standard comprehension exercises. Less skilled comprehenders given inference training improved significantly more than those given decoding practice, and slightly, but not significantly, more than those given comprehension exercises. Skilled comprehenders showed little improvement regardless of treatment condition. The results are discussed in relation to possible sources of comprehension failure, and implications for remediation.In a series of studies of children with adequate decoding skills, but poor reading comprehension, Oakhill (1982Oakhill ( , 1984 has shown that such children have specific deficits in constructive memory and inference skills. Poor comprehenders were compared with good comprehenders, who were similar in decoding skill, but superior in comprehension skill. This method of subject selection prevents comprehension skill being confounded with differences in decoding abilities. However, as Campione and Armbruster (1984) point out, when investigating the source of such deficits one cannot totally rule out the possibility that there is some crucial aspect of decoding, not measured by the tests used, that differentiates between the two groups, and might therefore explain the differences found. One way of testing this possibility is to gain convergent evidence using a training study. If less skilled comprehenders can be trained in the skills they are claimed to lack, then they should show improvements in comprehension skill. Campione and Armbruster describe several different possible outcomes of such a study, comparing trained and untrained children in deficient and normal groups. The outcome that points most clearly to the causal role of the trained skills is achieved when trained less skilled comprehenders improve to the level of skilled comprehenders, and skilled comprehenders remain at the same high level of *Correspondence should be addressed to Nicola Yuill, who is now at the MRC Unit on the Development and Integration of Behaviour, Madingley, Cambridge CB3 8AA, England. Jane Oakhill is at the MRC Perceptual and Cognitive Performance Unit, University of Sussex, Brighton BNI 9QG, England. This research was funded by an ESRCproject grant to Alan Parkin, whose help we gratefully acknowledge. We are also very grateful for the cooperation of the staff and pupils of Balfour, Coombe Road, Downs, Fairlight and Moukecoomb Junior Schools.