Search

Popular word embedding methods such as GloVe and Word2Vec are related to the factorization of the pointwise mutual information (PMI) matrix. In this paper, we link correspondence analysis (CA) to the factorization of the PMI matrix. CA is a dimensionality reduction method that uses singular value decomposition (SVD), and we show that CA is mathematically close to the weighted factorization of the PMI matrix. In addition, we present variants of CA that turn out to be successful in the factorization of the word-context matrix, i.e. CA applied to a matrix where the entries undergo a square-root transformation (ROOT-CA) and a root-root transformation (ROOTROOT-CA). An empirical comparison among CA- and PMI-based methods shows that overall results of ROOT-CA and ROOTROOT-CA are slightly better than those of the PMI-based methods.

Technology-Assisted Review (TAR) aims to reduce the human effort required for screening processes such as abstract screening for systematic literature reviews. Human reviewers label documents as relevant or irrelevant during this process, while the system incrementally updates a prediction model based on the reviewers' previous decisions. After each model update, the system proposes new documents it deems relevant, to prioritize relevant documentsover irrelevant ones. A stopping criterion is necessary to guide users in stopping the review process to minimize the number of missed relevant documents and the number of read irrelevant documents. In this paper, we propose and evaluate a new ensemble-based Active Learning strategy and a stopping criterion based on Chao's Population Size Estimator that estimates the prevalence of relevant documents in the dataset. Our simulation study demonstrates that this criterion performs well on several datasets and is compared to other methods presented in the literature.

The initial dimensions extracted by latent semantic analysis (LSA) of a document-term matrix have been shown to mainly display marginal effects, which are irrelevant for information retrieval. To improve the performance of LSA, usually the elements of the raw document-term matrix are weighted and the weighting exponent of singular values can be adjusted. An alternative information retrieval technique that ignores the marginal effects is correspondence analysis (CA). In this paper, the information retrieval performance of LSA and CA is empirically compared. Moreover, it is explored whether the two weightings also improve the performance of CA. The results for four empirical datasets show that CA always performs better than LSA. Weighting the elements of the raw data matrix can improve CA; however, it is data dependent and the improvement is small. Adjusting the singular value weighting exponent often improves the performance of CA; however, the extent of the improvement depends on the dataset and the number of dimensions.

We present a method to generate contingency tables that follow loglinear models with prescribed marginal probabilities and dependence structures. We make use of (loglinear) Poisson regression, where the dependence structures, described using odds ratios, are implemented using an offset term. Other statistical models related to loglinear models that fall into the scope of this paper, such as the logistic regression model, the latent class model and the extended Rasch are discussed as well. We apply this methodology to carry out simulation studies in the context of population size estimation using dual system and triple system estimators, popular in official statistics. These estimators use contingency tables that summarize the counts of elements enumerated or captured within lists that are linked. The simulation is used to investigate these estimators in the situation that the model assumptions are fulfilled, and the situation that the model assumptions are violated.

If part of a population is hidden but two or more sources are available that each cover parts of this population, dual- or multiple-system(s) estimation can be applied to estimate this population. For this it is common to use the log-linear model, estimated with maximum likelihood. These maximum likelihood estimates are based on a non-linear model and therefore suffer from finite-sample bias, which can be substantial in case of small samples or a small population size. This problem was recognised by Chapman, who derived an estimator with good small sample properties in case of two available sources. However, he did not derive an estimator for more than two sources. We propose an estimator that is an extension of Chapman’s estimator to three or more sources and compare this estimator with other bias-reduced estimators in a simulation study. The proposed estimator performs well, and much better than the other estimators. A real data example on homelessness in the Netherlands shows that our proposed model can make a substantial difference.

Many multivariate statistical models have dimensional structures. Such models typically require judicious choice of dimensionality. Likelihood ratio tests are often used for dimensionality selection. However, to this day there is still a great deal of confusion about the asymptotic distributional properties of the log-likelihood ratio (LR) statistics in some areas of psychometrics. Although in many cases the asymptotic distribution of the LR statistic representing the difference between the correct model (of specific dimensionality) and the saturated model is guaranteed to be chi-square, that of the LR statistic representing the difference between the correct model and the one with one dimension higher than the correct model is not likely to be chi-square due to a violation of one of regularity conditions. In this paper, we attempt to clarify the misunderstanding that the latter is also assured to be asymptotically chi-square. This common misunderstanding has occurred repeatedly in various fields, although in some areas it has been corrected.

The Extended Crosswise Model (ECWM) is a randomized response model with neutral response categories, relatively simple instructions, and the availability of a goodness-of-fit test. This paper refines this model with a number sequence randomizer that virtually precludes the possibility to give evasive responses. The motivation for developing this model stems from a strategic priority of WADA (World Anti-Doping Agency) to monitor the prevalence of doping use by elite athletes. For this model we derived a maximum likelihood estimator that allows for binary logistic regression analysis. Three studies were conducted on online platforms with a total of over 6, 000 respondents; two on controlled substance use and one on compliance with COVID-19 regulations in the UK during the first lockdown. The results of these studies are promising. The goodness-of-fit tests showed little to no evidence for response biases, and the ECWM yielded higher prevalence estimates than direct questions for sensitive questions, and similar ones for non-sensitive questions. Furthermore, the randomizer with the shortest number sequences yielded the smallest response error rates on a control question with known prevalence.

Latent semantic analysis (LSA) and correspondence analysis (CA) are two techniques that use a singular value decomposition (SVD) for dimensionality reduction. LSA has been extensively used to obtain low-dimensional representations that capture relationships among documents and terms. In this article, we present a theoretical analysis and comparison of the two techniques in the context of document-term matrices. We show that CA has some attractive properties as compared to LSA, for instance that effects of margins arising from differing document-lengths and term-frequencies are effectively eliminated, so that the CA solution is optimally suited to focus on relationships among documents and terms. A unifying framework is proposed that includes both CA and LSA as special cases. We empirically compare CA to various LSA based methods on text categorization in English and authorship attribution on historical Dutch texts, and find that CA performs significantly better. We also apply CA to a long-standing question regarding the authorship of the Dutch national anthem Wilhelmus and provide further support that it can be attributed to the author Datheen, amongst several contenders.

The increasing need for energy resources and the threat of shortage of available fossil fuels in the future along with the adverse environmental issues arising from consumption of these fuels, especially climate change, have led countries to pursue sustainable and eco-friendly modes of producing energy. Meanwhile, in recent years, many researchers have considered renewable energies and their related technologies. However, the evidence indicates that social or public acceptance of Renewable Energy Technologies (RETs) is low and people, especially in rural areas of developing countries, are less likely to adopt and use such technologies. Accordingly, the current study was undertaken to investigate the factors influencing Iranian villagers’ intention to use RETs through two socio-psychological models with pro-self and pro-social motivations and to integrate them into one comprehensive theoretical framework. A questionnaire survey was conducted for 393 villagers in Zanjan County in northwestern Iran. The study results disclosed that personal norm, attitude, social norm, and perceived behavioral control had significant positive impacts on the intention to use RETs. Most importantly, the findings confirmed the effectiveness of the original models of Norm Activation Model (NAM) and Theory of Planned Behavior (TPB) in explaining the villagers' intention. However, the utility and applicability of an integrative model of TPB-NAM were superior to the original models. Additionally, the original NAM contributed considerably more to the integrative model than the original TPB. Collectively, the pro-social orientations outweighed the pro-self motives in the case of explaining the behavioral intention regarding the use of RETs among Iranian villagers. Overall, this study's findings contribute to the theory and practice around sustainable energy development in Iran and other developing countries.

We investigate the use of two or more linked lists, for both population size estimation and the relationship between variables appearing on all or only some lists. This relationship is usually not fully known because some individuals appear in only some lists, and some are not in any list. These two problems have been solved simultaneously using the EM algorithm. We extend this approach to estimate the size of the indigenous Māori population in New Zealand, leading to several innovations: (1) the approach is extended to four lists (including the population census), where the reporting of Māori status differs between registers; (2) some individuals in one or more lists have missing ethnicity, and we adapt the approach to handle this additional missingness; (3) some lists cover subsets of the population by design. We discuss under which assumptions such structural undercoverage can be ignored and provide a general result; (4) we treat the Māori indicator in each list as a variable measured with error, and embed a latent class model in the multiple system estimation to estimate the population size of a latent variable, interpreted as the true Māori status. Finally, we discuss estimating the Māori population size from administrative data only. Supplementary materials for our article are available online.

In dual system estimation and multiple system estimation problems there may be lists available that only cover subsets of the population by design. We discuss under which assumptions such structural undercoverage in the lists can be ignored in these estimation problems.

Compositional data are non-negative data collected in a rectangular matrix with a constant row sum. Due to the non-negativity the focus is on conditional proportions that add up to 1 for each row. A row of conditional proportions is called an observed budget. Latent budget analysis (LBA) assumes a mixture of latent budgets that explains the observed budgets. LBA is usually fitted to a contingency table, where the rows are levels of one or more explanatory variables and the columns the levels of a response variable. In prospective studies, there is only knowledge about the explanatory variables of individuals and interest goes out to predicting the response variable. Thus, a form of LBA is needed that has the functionality of prediction. Previous studies proposed a constrained neural network (NN) extension of LBA that was hampered by an unsatisfying prediction ability. Here we propose LBA-NN, a feed forward NN model that yields a similar interpretation to LBA but equips LBA with a better ability of prediction. A stable and plausible interpretation of LBA-NN is obtained through the use of importance plots and table, that show the relative importance of all explanatory variables on the response variable. An LBA-NN-K- means approach that applies K-means clustering on the importance table is used to produce K clusters that are comparable to K latent budgets in LBA. Here we provide different experiments where LBA-NN is implemented and compared with LBA. In our analysis, LBA-NN outperforms LBA in prediction in terms of accuracy, specificity, recall and mean square error. We provide open-source software at GitHub.

Tools for reliable assessment of socially sensitive or transgressive behavior warrant constant development. Among them, the Crosswise Model (CM) has gained considerable attention. We systematically reviewed and meta-analyzed empirical applications of CM and addressed a gap for quality assessment of indirect estimation models. Guided by the PRISMA protocol, we identified 45 empirical studies from electronic database and reference searches. Thirty of these were comparative validation studies (CVS) comparing CM and direct question (DQ) estimates. Six prevalence studies exclusively used CM. One was a qualitative study. Behavior investigated were substance use and misuse (k = 13), academic misconduct (k = 8), and corruption, tax evasion, and theft (k = 7) among others. Majority of studies (k = 39) applied the “more is better” hypothesis. Thirty-five studies relied on birthday distribution and 22 of these used P = 0.25 for the non-sensitive item. Overall, 11 studies were assessed as high-, 31 as moderate-, and two as low quality (excluding the qualitative study). The effect of non-compliance was assessed in eight studies. From mixed CVS results, the meta-analysis indicates that CM outperforms DQ on the “more is better” validation criterion, and increasingly so with higher behavior sensitivity. However, little difference was observed between DQ and CM estimates for items with DQ prevalence estimate around 50%. Based on empirical evidence available to date, our study provides support for the superiority of CM to DQ in assessing sensitive/transgressive behavior. Despite some limitations, CM is a valuable and promising tool for population level investigation.