Predictions from lexical-marking theory and a scanning model of recognition memory were compared in a continuous-recognition paradigm. The results were at least consistent with the scanning model and in two cases were at odds with lexical-marking theory. In particular, the probability of making a false-recognition error to the synonym of a previously presented word remained constant over a lag of 120, whereas a recognition-control condition showed a large decrement. Lexical-marking theory predicts equal decay rates. In addition, the latency for correctly rejecting a synonym of a previously presented word monotonically increased over the same lag range, whereas lexical-marking theory predicts a monotonic decrease.

No abstract

1985a), R. S. Bogartz (1990a), and others have questioned forgetting comparisons based on tests of interaction between retention and a second variable. A method for comparing forgetting between conditions on the basis of an article by N. H. Anderson (1963) is presented here. The shape method compares the underlying "shapes" of the performance curves. Anderson's shape method is outlined for memory studies, and its inherent assumptions are stated. A statistical test is developed to apply the shape method to realistic situations in which zero retention interval or asymptotic performance data are not available. A power analysis varied the memory curve shape, sample size, and standard deviation and demonstrated the method's ability to detect shape differences. Application to a recent study claiming a forgetting difference yielded the opposite conclusion. Theoretical and practical limitations of Loftus's and Bogartz's methods are discussed.1 want to thank R. K. ("Skip") Schutz and Ed Israelski for their careful reading of this article and for their useful comments. I also acknowledge Brian Bancroft's computer support in the development of the MS-DOS-based shape method statistical test. I am indebted to the valuable feedback on drafts of this article from D. Bamber, R. Bogartz, G. Loftus, J. Myers, J. Wixted, and an anonymous reviewer. I particularly want to acknowledge my wife, Laure, for her unflagging support throughout the years it took to develop this concept.

Scintillation of radio signals from an artificial earth satellite has been analyzed using the spaced receiver technique in order to gain a greater understanding of the latitudinal and diurnal variation of the height of the electron irregularities responsible for the observed scintillation on the ground. a r e also noted.nighttime hours than during the midday hours. Also, the average i rregularity height increases towards the north. This increase appears to be due to the existence of small patches of irregularities occurring a t higher heights to the north. A large portion of the data falls in the 300-450 km range and thus confirms the observations of others. Other prominent features of the scintillationThe irregularity heights tend to be greater during the A regular type of fading has been observed on many occasions. This regular scintillation is compared t o that observed by others and an argument is presented which tends to refute use of the sharp edge diffraction model a s an explanation for this phenomenon. Also, irregularity patches a t the same sub-ionosphere latitude but a t different heights were observed and this phenomenon is discussed.

BackgroundThe use of meta-analysis to aggregate multiple studies has increased dramatically over the last 30 years. For meta-analysis of homogeneous data where the effect sizes for the studies contributing to the meta-analysis differ only by statistical error, the Mantel–Haenszel technique has typically been utilized. If homogeneity cannot be assumed or established, the most popular technique is the inverse-variance DerSimonian–Laird technique. However, both of these techniques are based on large sample, asymptotic assumptions and are, at best, an approximation especially when the number of cases observed in any cell of the corresponding contingency tables is small.ResultsThis paper develops an exact, non-parametric test based on a maximum likelihood test statistic as an alternative to the asymptotic techniques. Further, the test can be used across a wide range of heterogeneity. Monte Carlo simulations show that for the homogeneous case, the ML-NP-EXACT technique to be generally more powerful than the DerSimonian–Laird inverse-variance technique for realistic, smaller values of disease probability, and across a large range of odds ratios, number of contributing studies, and sample size. Possibly most important, for large values of heterogeneity, the pre-specified level of Type I Error is much better maintained by the ML-NP-EXACT technique relative to the DerSimonian–Laird technique. A fully tested implementation in the R statistical language is freely available from the author.ConclusionsThis research has developed an exact test for the meta-analysis of dichotomous data. The ML-NP-EXACT technique was strongly superior to the DerSimonian–Laird technique in maintaining a pre-specified level of Type I Error. As shown, the DerSimonian–Laird technique demonstrated many large violations of this level. Given the various biases towards finding statistical significance prevalent in epidemiology today, a strong focus on maintaining a pre-specified level of Type I Error would seem critical.