Showing papers on "Encoding (memory) published in 1974"

Remotely Multiplexed Data Acquisition for Media Systems

Abstract: The MEDIA digital process control system providing a comprehensive approach to computer control, centralized display, and operator intervention is briefly reviewed. This system consists of a range of modules for encoding individual measurements, generating individual loop control functions, and adjusting individual regulating elements on a continuous basis without using sampling techniques.

Information science and the special needs of chemical documentation

Abstract: The compulsion for more and more specific searches in fast growing information files has produced a variety of sophisticated methods of documentation and information and given rise to a new discipline, called information science. As chemistry is particularly dependent on documentation and places variety of requirements on its methods and effectiveness, a special relationship developed between chemical documentation and information science. Besides the use of common keywords for characterizing a paper, the encoding of structures e.g. in structural formulas, process flow charts, and syntactical combinations as well as the retrieval of substructures are of great importance. Fortunately, the “hard” formulae language makes possible a variety of machine-generated and -transformed codes for special purposes. But even in the field of the “soft” non-structural data much pioneer work has been done in chemistry which is utilized by the documentalists in other areas. On the other hand, the needs of chemical documentation are steadily growing. Some examples of general and special problems which can be considered solved will be given. A catalog of goals yet to be reached will also be briefly described.

Memory search of words and nonwords

Abstract: Summary.-Classification of single-letter probes as members or nonmembers of a list of letters held in memory is faster if the memory list spells a word than if it does not. In terms of Sternberg's theory of memory search, the results seem to show that the speed of the letter-by-letter memory-comparison stage is increased when the letters in the memory set spell a word. In Sternberg's (5) memory-scanning paradigm S first memorizes a short list (termed the memory set) and then is presented probe irems which may or may not be members of the memory set. Ss rask is to classify the probes. Typically, the reaction time (RT) for both yes and no responses increases linearly wirh the size of the memory set. The slope of the function relating RT to set size is consistent with a model whereby Ss compare the probe co each member of the memory set at, in the case of letters or digits, a rate of about 30 msec. per comparison. The present srudy seeks ro determine whether a memory set which forms a well-integrated unit is searched faster than one which does not. The memory sets were either meaningless strings of letters or strings that spelled a familiar word. It has been shown that visual search for a target letter is faster when the letters in the display form words than when they do not (3), and it is of some interest to determine whether the same effect pertains in memory search. In particular, if words do show an advantage over nonwords in memory search, it will be of interest to determine the locus of the effect in terms of an addicive model of RT in the memory-scanning task. In this model the RT is the sum of times consumed in four separate, sequentially dependent stages: (a) encoding of the probe, (b) comparison of the probe with each item in memory, (c) choice of response, and (d) execution of response (4, 5). Factors that affect stage (b), memory comparison, will affect the slope of the RT vs memory load function, and factors that affect the ocher stages should affect only the intercept of the function. Kreuger's (3, Exp. IV) results seem to show that the constant latencies but not the individual memory-comparison times (stage b) would be smaller for words than nonwords. However, Kreuger's results do not bear on the present question. In his Exp. IV the viszral display had eicher words or nonwords in it, and memory-search rates were derived from the effect of memory load on visual-search rate. The letters in the memory load always formed nonwords.