Life is pleasant--and memory helps to keep it that way!

People's recollections of the past are often positively biased. This bias has 2 causes. The 1st cause lies in people's perceptions of events. The authors review the results of several studies and present several new comparative analyses of these studies, all of which indicate that people perceive events in their lives to more often be pleasant than unpleasant. A 2nd cause is the fading affect bias: The affect associated with unpleasant events fades faster than the affect associated with pleasant events. The authors review the results of several studies documenting this bias and present evidence indicating that dysphoria (mild depression) disrupts such bias. Taken together, this evidence suggests that autobiographical memory represents an important exception to the theoretical claim that bad is stronger than good. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Savings in relearning: II. On the formation of behavior-based trait associations and inferences.

Five experiments based on Carlston and Skowronski's (1994) relearning paradigm suggest that people spontaneously derive trait knowledge about actors from behaviors but that this knowledge may reflect either explicit trait inference processes or implicit actor–trait associations. Experiments 1 and 2 found that inference-instructed and control Ss showed equivalent savings in subsequent efforts to learn actor–trait pairs but not when instructed Ss initially inferred the wrong trait. Experiment 3 showed that savings were equivalent for stimuli from different sources, and Experiment 4 showed that savings effects persisted even when the target was only incidentally associated with a stimulus behavior. Finally, Experiment 5 suggests that after several days, even explicit trait inferences can become inaccessible to intentional retrieval, although the earlier experiments show that they continue to exert an implicit effect on learning. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Inorganic-melt-catalysed hydrogenation of phenanthrene

The hydrogenation of the coal model compound phenanthrene was investigated using a molten hydroxide-carbonate catalyst. The rate of catalysed hydrogenation was found to be approximately an order of magnitude greater than the uncatalysed reaction. Batch experiments were conducted at 20.7 MPa hydrogen pressure in a 1-l top-stirred autoclave at 480 °C for reaction times up to 1 h. Spinning-band distillation and gas-liquid chromatography coupled with mass spectrometry were used to separate and identify reaction products. Principal hydrogenation products were 9,10 dihydrophenanthrene and 1,2,3,4-tetrahydrophenanthrene. In addition, substituted naphthalenes and biphenyls were formed by hydrocracking of these two compounds. A similar product distribution was reported by others for an investigation of the thermal high-pressure hydrogenolysis of phenanthrene. The rate of the thermal reaction, however, was considerably lower.     

Negativity and extremity biases in impression formation: A review of explanations.

This article reviews the research relevant to negativity and extremity biases in impression formation and discusses that research as it relates to the major theories that explain these biases. We also describe a model for these biases that draws on principles of natural object categorization. This model explains negativity and extremity biases in terms of the perceived diagnosticity of different kinds of cues for alternative categorizations of the stimulus. The model not only accounts for existing evidence regarding negativity and extremity biases but also suggests circumstances (a) in which positivity biases should occur and (b) in which single cues might be sufficient to prompt categorizations resistant to counterevidence. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Glycolipids and thylakoid proteins in chloroplasts and streptomycin-bleached lamellae ofEuglena

Thylakoid proteins from (a) the chloroplasts of greenEuglena gracilis cells and from (b) the corresponding concentric lamellar bodies of streptomycin-bleachedEuglena have been extracted in aqueous sodium dodecyl sulfate (SDS) and purified. The molecular weights of the proteins, determined by polyacrylamide gel electrophoresis in SDS, were 9 kilodaltons for (a) and 9 and 11.5 kilodaltons for the 2 polypeptide chains of (b). Each of the purified proteins consists of 55% apolar and 45% polar amino acids. The lipids of the green and bleached cells differ in polarity of components. Neutral galactolipids are abundant in green cells but not in streptomycin-bleached cells. Anionic sulfolipid content is greater in the chlorotic organisms. Polyenoic fatty acids, mainly with 16, 18 and 19 carbon atoms, occur in the green cell galactolipids. Cells grown in the presence of streptomycin are, in contrast, rich in monoenes. The opposite is observed in the sulfolipids: monoenes with 16 and 20 carbons are the main fatty acids in the green cells, while a dienoic fatty acid with 12 carbons is found in large amounts in streptomycin-bleached cultures.     

Noise-Robust Automatic Speech Recognition Using a Predictive Echo State Network

Artificial neural networks have been shown to perform well in automatic speech recognition (ASR) tasks, although their complexity and excessive computational costs have limited their use. Recently, a recurrent neural network with simplified training, the echo state network (ESN), was introduced by Jaeger and shown to outperform conventional methods in time series prediction experiments. We created the predictive ESN classifier by combining the ESN with a state machine framework. In small-vocabulary ASR experiments, we compared the noise-robust performance of the predictive ESN classifier with a hidden Markov model (HMM) as a function of model size and signal-to-noise ratio (SNR). The predictive ESN classifier outperformed an HMM by 8-dB SNR, and both models achieved maximum noise-robust accuracy for architectures with more states and fewer kernels per state. Using ten trials of random sets of training/validation/test speakers, accuracy for the predictive ESN classifier, averaged between 0 and 20 dB SNR, was 81plusmn3%, compared to 61plusmn2% for an HMM. The closed-form regression training for the ESN significantly reduced the computational cost of the network, and the reservoir of the ESN created a high-dimensional representation of the input with memory which led to increased noise-robust classification.