On exploratory factor analysis: a review of recent evidence, an assessment of current practice, and recommendations for future use

Exploratory factor analysis (hereafter, factor analysis) is a complex statistical method that is integral to many fields of research. Using factor analysis requires researchers to make several decisions, each of which affects the solutions generated. In this paper, we focus on five major decisions that are made in conducting factor analysis: (i) establishing how large the sample needs to be, (ii) choosing between factor analysis and principal components analysis, (iii) determining the number of factors to retain, (iv) selecting a method of data extraction, and (v) deciding upon the methods of factor rotation. The purpose of this paper is threefold: (i) to review the literature with respect to these five decisions, (ii) to assess current practices in nursing research, and (iii) to offer recommendations for future use. The literature reviews illustrate that factor analysis remains a dynamic field of study, with recent research having practical implications for those who use this statistical method. The assessment was conducted on 54 factor analysis (and principal components analysis) solutions presented in the results sections of 28 papers published in the 2012 volumes of the 10 highest ranked nursing journals, based on their 5-year impact factors. The main findings from the assessment were that researchers commonly used (a) participants-to-items ratios for determining sample sizes (used for 43% of solutions), (b) principal components analysis (61%) rather than factor analysis (39%), (c) the eigenvalues greater than one rule and screen tests to decide upon the numbers of factors/components to retain (61% and 46%, respectively), (d) principal components analysis and unweighted least squares as methods of data extraction (61% and 19%, respectively), and (e) the Varimax method of rotation (44%). In general, well-established, but out-dated, heuristics and practices informed decision making with respect to the performance of factor analysis in nursing studies. Based on the findings from factor analysis research, it seems likely that the use of such methods may have had a material, adverse effect on the solutions generated. We offer recommendations for future practice with respect to each of the five decisions discussed in this paper.

A novel parallel algorithm for frequent itemsets mining in massive small files datasets

In big data analysis, frequent itemsets mining plays a key role in mining associations, correlations and causality. Since some traditional frequent itemsets mining algorithms are unable to handle massive small files datasets effectively, such as high memory cost, high I/O overhead, and low computing performance, we propose a novel parallel frequent itemsets mining algorithm based on the FP-Growth algorithm and discuss its applications in this paper. First, we introduce a small files processing strategy for massive small files datasets to compensate defects of low read-write speed and low processing efficiency in Hadoop. Moreover, we use MapReduce to redesign the FP-Growth algorithm for implementing parallel computing, thereby improving the overall performance of frequent itemsets mining. Finally, we apply the proposed algorithm to the association analysis of the data from the national college entrance examination and admission of China. The experimental results show that the proposed algorithm is feasible and valid for a good speedup and a higher mining efficiency, and can meet the actual requirements of frequent itemsets mining for massive small files datasets. © 2014 ISSN 2185-2766.

The trickle-down effect of the mining boom in Australia: fact or myth?

This paper examines the relationship between the output levels in the mining sector and various non-mining sectors in an attempt to understand the role of the mining sector in Australia. The unobserved components time series model is used to estimate the effects of the output gap and the growth regime in the mining sector on the output level of each of several non-mining sectors. Overall, the estimates obtained do not suggest an overwhelmingly positive effect running from the mining sector to other production and services sectors, implying that the trickle-down effect of the mining boom may be a myth.

Measuring nursing competencies in the operating theatre: instrument development and psychometric analysis using Item Response Theory

BACKGROUND: Concern about the process of identifying underlying competencies that contribute to effective nursing performance has been debated with a lack of consensus surrounding an approved measurement instrument for assessing clinical performance. Although a number of methodologies are noted in the development of competency-based assessment measures, these studies are not without criticism. RESEARCH AIM: The primary aim of the study was to develop and validate a Performance Based Scoring Rubric, which included both analytical and holistic scales. The aim included examining the validity and reliability of the rubric, which was designed to measure clinical competencies in the operating theatre. RESEARCH METHOD: The fieldwork observations of 32 nurse educators and preceptors assessing the performance of 95 instrument nurses in the operating theatre were used in the calibration of the rubric. The Rasch model, a particular model among Item Response Models, was used in the calibration of each item in the rubric in an attempt at improving the measurement properties of the scale. This is done by establishing the 'fit' of the data to the conditions demanded by the Rasch model. RESULTS: Acceptable reliability estimates, specifically a high Cronbach's alpha reliability coefficient (0.940), as well as empirical support for construct and criterion validity for the rubric were achieved. Calibration of the Performance Based Scoring Rubric using Rasch model revealed that the fit statistics for most items were acceptable. CONCLUSION: The use of the Rasch model offers a number of features in developing and refining healthcare competency-based assessments, improving confidence in measuring clinical performance. The Rasch model was shown to be useful in developing and validating a competency-based assessment for measuring the competence of the instrument nurse in the operating theatre with implications for use in other areas of nursing practice.

Rule-based extraction of goal-use case models from text

Goal and use case modeling has been recognized as a key approach for understanding and analyzing requirements. However, in practice, goals and use cases are often buried among other content in requirements specifications documents and written in unstructured styles. It is thus a time-consuming and error-prone process to identify such goals and use cases. In addition, having them embedded in natural language documents greatly limits the possibility of formally analyzing the requirements for problems. To address these issues, we have developed a novel rule-based approach to automatically extract goal and use case models from natural language requirements documents. Our approach is able to automatically categorize goals and ensure they are properly specified. We also provide automated semantic parameterization of artifact textual specifications to promote further analysis on the extracted goal-use case models. Our approach achieves 85% precision and 82% recall rates on average for model extraction and 88% accuracy for the automated parameterization.

Recruitment of a keystone tree species must concurrently manage flooding and browsing

Multiple pressures (land-use change, water extraction and climate change) interact to influence biodiversity and ecosystem processes, but direct evidence for interactions among multiple pressures is limited. Floodplain forests are an acute example of how interacting pressures (river regulation, water extraction, decreasing rainfall and mammal browsing) interact to degrade native ecosystems. We conducted a 2-year field experiment to determine how flooding, browsing and sediment salinity interacted to determine in situ seedling survival and growth of the keystone floodplain tree species (Eucalyptus camaldulensis Dehnh.). On semi-arid floodplains of southern Australia, 1-year-old seedlings were planted on the banks of six ephemeral creeks, three of which were flooded with management flows before planting while the others remained dry. Four plots were established at each creek, two open to browsing and two fenced to exclude mammal herbivores. Flooding had a strong positive effect on seedling survival and height, but browsing had strong negative effects. Sediment salinity (a covariate rather than a designed effect) had a weak negative effect on seedling survival and height. The positive effects of flooding were largely offset by the negative interaction with browsing and, to a lesser extent, sediment salinity. Although flooding has been restored to some degraded floodplain forests subjected to river regulation and a drying climate, the long-term success of such actions is likely to be undermined by persistent browsing. Synthesis and applications. Management actions that focus on single pressures (e.g. infrequent flooding) and processes (e.g. mature tree survival) while ignoring other pressures are unlikely to sustain populations of keystone species, suggesting that complementary strategies (managed flooding with herbivore control) are necessary to sustain recruitment and, therefore, ensure the future health of these essential ecosystems.

Experimental comparison of approaches for feature extraction of facial attributes

In this paper, we compare the effectiveness of widely used approaches for representation of facial features in face images. Feature extraction is performed on face images for representation of four facial attributes, namely gender, age, race, and expression, by using discrete wavelet transform (DWT), Gabor wavelet, scale-invariant feature transform, local binary pattern (LBP), and Eigenfaces. After feature extraction and dimension reduction, demographic and expression classification is performed to identify the most discriminating techniques for representation of facial features. Extensive experiments are performed using publicly available face databases, namely Yale, Face95 Essex, and Cohn-Kanade (CK+) databases. Experimental results show that DWT, LBP, and Gabor wavelet methods are robust to variations of illumination, facial expression, and geometric transformations. Experimental results also show that race and expression are more difficult to predict than gender and age.