Papers reviewed and determined not to be word norm studies. Use the flag icon to report errors or suggest re-inclusion.
18625 papers
A LINC-8 computer is used with a Tektronix 611 CRT for presenting verbal information, and a slide projector for displaying facial photographs in studies of interpersonal referential communication.
Computer processing made it feasible for us to base our study of the vocabulary of six-year-old Navajo children on as large a corpus as we could collect in the time available. Our difficulties, as so often in computational linguistics, were with matters of linguistic theory rather than of computing. What we ran into was the as-yet unsolved question of the nature of the word in Navajo: how many affixes should be written as part of the verb and how many as separate words; and how does one handle the unbelievably complex morphophonemics in choosing headwords. The computer once again showed its ability, not just as an aid in handling large bodies of data, but as a heuristic device that makes clear to the researcher the limitations of his understanding of the material he is working with.
Language model (LM) plays an important role in natural language processing (NLP) systems, such as machine translation, speech recognition, learning token embeddings, natural language generation, and text classification. Recently, the multilayer long short-term memory (LSTM) models have been demonstrated to achieve promising performance on word-level language modeling. For each LSTM layer, larger hidden size usually means more diverse semantic features, which enables the LM to perform better. However, we have observed that when a certain LSTM layer reaches a sufficiently large scale, the promotion of overall effect will slow down, as its hidden size increases. In this article, we analyze that an important factor leading to this phenomenon is the high correlation between the newly extended hidden states and the original hidden states, which hinders diverse feature expression of the LSTM. As a result, when the scale is large enough, simply lengthening the LSTM hidden states will cost tremendous extra parameters but has little effect. We propose a simple yet effective improvement on each LSTM layer consisting of a large-scale Major LSTM and a small-scale Minor LSTM to break the high correlation between the two parts of hidden states, which we call Major-Minor LSTMs (MMLSTMs). In experiments, we demonstrate the LM with MMLSTMs surpasses the existing state-of-the-art model on Penn Treebank (PTB) and WikiText-2 (WT2) data sets and outperforms the baseline by 3.3 points in perplexity on WikiText-103 data set without increasing model parameter counts.
In this article we consider a multilevel first-order autoregressive [AR(1)] model with random intercepts, random autoregression, and random innovation variance (i.e., the level 1 residual variance). Including random innovation variance is an important extension of the multilevel AR(1) model for two reasons. First, between-person differences in innovation variance are important from a substantive point of view, in that they capture differences in sensitivity and/or exposure to unmeasured internal and external factors that influence the process. Second, using simulation methods we show that modeling the innovation variance as fixed across individuals, when it should be modeled as a random effect, leads to biased parameter estimates. Additionally, we use simulation methods to compare maximum likelihood estimation to Bayesian estimation of the multilevel AR(1) model and investigate the trade-off between the number of individuals and the number of time points. We provide an empirical illustration by applying the extended multilevel AR(1) model to daily positive affect ratings from 89 married women over the course of 42 consecutive days.
ilecent inv~sti~ations into the attitudinal effects of mere exposure (2ajonc, 1968) report the surprising finding that exploratory behavior and favorable attitudes appear to be negatively related (Harrison, 1968).This finding was supportive of Zajonc's general suggestion that a response-conflict phenomenon may well be mediating his well-docurn.entedmere exposure effects.This research examines the relationship between exploratory behavior and affective rating in light of some theoretical and intuitive considerations which lead to predictions opposite those of Harrison.The empirical question which is asked is whether looking time (the generally accepted measure of exploratory behavior) reflects a positive or negative disposition towards slides of paintings in a relatively demand-free situation.A significant positive correlation was found to exist between the measures of exploratory behavior and affective rating.A critique is made of the Harriflon (1968) study which reported.conflictingresults and upon which this study is based.Some evidence supporting specific criticisms is advanced and methodological considerations relevant to this area of research are explored. I, I
Trial-and-error and errorless training procedures were used to train two killer whales on a simple visual discrimination problem. Both whales demonstrated difficulties in solving the problem under trial-and-error training. The relative effectiveness of the two training procedures is evaluated.
This study is an exploration of the attitude of low \nsocio-economic whites and non-whites toward themselves and \nstandardized testing and an attempt to determine if an individual's \nself-image is related to his attitude toward testing. \nA Self-Image Example-Anchored Rating Scale was developed \nand administered to determine an individual's self-concept. \nA Semantic Differential Rating Scale was adapted to determine \nattitude toward standardized testing. The Semantic Differential \nalso investigated attitudes toward authority, peer \ngroups and surroundings. Two hundred East-San Francisco Bay \nArea continuation school students, homogeneously grouped, \ncomprised the population. The Semantic Differential and Self-Image \nExample-Anchored Rating Scale were administered in four \ncontinuation schools, results were tabulated and subjected \nto computer analysis. Self-Image ratings divided the population \ninto four groups. These four groups were analyzed for \ncorrelations and variance. A factor analysis was performed \nto determine relationship of variables. \nResults of the investigation demonstrated significant \nrelationship between self-image and attitude toward standardized \ntesting.
A maintenance technique was developed in which neonatal monkeys obtain all liquid food by placing their heads in a face mask mounted on their cage wall. Complete self-feeding required only 3-6 days for animals started at birth. Once under a self-feeding Schedule, operant responses were shaped to study visual perception, visually guided motor performance, and discrimination learning at ages much younger than those allowed by most alternative methods. Dark rearing, with the only source of visual input being through the face mask eyeholes, allowed the E to control completely the neonate’s visual experiences and its opportunities for visual-motor responding. The method has proven useful in rhesus monkey newborns for studying adaptation to prismatic displacement at 30 days of age, and to performance on CRF, FI, and FR reinforcement schedules.
An evaluation model was developed to aid the classroom instructor in teaching his course more effectively by providing him with the basic data upon which to make judgments about student abilities, course attitudes, grade distributions, test appropriateness, etc. The MERMAC system was designed so that each instructor could obtain the data in a usable format and also be able to provide relatively immediate feedback to the students taking an examination or filling out a course evaluation questionnaire. A description of the software system and its interface to the evaluation model is presented.
This article examines only published work directly related to the study of French literature; excluded are automatic translation, production of bibliographies or critical editions, and linguistic studies.
(1970). The Effects of Frequency and Duration of Exposure on Response Competition and Affective Ratings. The Journal of Psychology: Vol. 75, No. 2, pp. 163-169.
Research into the literary Polish language of the 17th century was undertaken relatively late. No significant relevant attempts had been made before the mid-20th century. The first detailed studies were made in the 1950s while the 1970s marked significantly advanced studies on the subject. The first monograph of the 17th century Polish language was published as late as in 2002. However, the monograph, presenting the entire grammar system together with the spelling rules and phonetics, does not conclude research into this period as it is rather general and fails to discuss many major issues related to the language’s evolution, for example the regional conditionings of the course of the normalization processes. This is a shortage plaguing many works on the 17th century Polish language. Therefore, it is very difficult to reconstruct the emergence and geographic distribution of linguistic innovations, over time transferring into a language norm. As for the 17th century, we can only refer to the text norm (or a practical norm/usus). In the century in question, a codified (theoretical) norm had not yet been developed although the grammar instructions from that time include some guidelines on linguistic correctness. As printing had already been well developed, we have enough sources to reconstruct text norms and to show its evolution. By referring to studies and text overviews one may conclude that the 17th century inherited from the 16th century a system largely normalized on various levels: the spelling, phonetics and morphology. On the other hand, it also either inherited or produced co-functional variants which only aspired to becoming a part of the norm. On the basis of selected issues from the realm of phonetics, declension and conjugation, the article shows examples of norms inherited from the previous century and, more importantly, processes of new linguistic norms emerging in the 17th century. These processes have been presented not only in reference to the averaged status in the general Polish language but also, whenever possible, their course is presented in the regional varieties of the language. By resorting to the methodology and terminology developed by Irena Bajerowa, the authors have touched upon the issue of the regionally diversified rate of the normalization processes. However, with reference to the 17th century, we are not able to present many linguistic facts in this specific way. What we need is detailed statistical research that shows the share of the specific variants in the regional variations of the Polish language, in several synchronic sectional views (vertical time segments). This research procedure will allow to show the source of new linguistic norms, the rate at which they disseminate and which regions are the most innovative ones.
An automated “intelligence” test for rats is described. Supporting data indicate that this maze-problem series is sensitive to hypothyroidism-induced learning deficit and sex differences. Comparisons between this new problem set and the Rabinovitch-Rosvold (1951) closed-field tests are presented with respect to difficulty, internal consistency, and test-retest reliability.