An improved robust variance inflation factor: Reducing the negative effects of good leverage points

In multiple linear regression analysis, the variance inflation factor is a well-known collinearity measure. It is defined as the function of the coefficient of determination between the explanatory variables, and it is based on the maximum likelihood estimator of the regression coefficients. Nevertheless, in addition to outliers, leverage observations can have significant impact on the coefficient of determination, and thereby the variance inflation factor. This study presents an improved robust variance inflation factor estimator that is not affected by these observations. Simulation studies and a real data analysis indicate that the modified robust variance inflation factor estimator performs better than the traditional one.

Robust Multicollinearity Diagnostic Measure For Fixed Effect Panel Data Model

It is now evident that high leverage points (HLPs) can induce the multicollinearity pattern of a data in fixed effect panel data model. Those observations that are responsible for this phenomenon are called high leverage collinearity-enhancing observations (HLCEO). The commonly used within group ordinary least squares (WOLS) estimator for estimating the parameters of fixed effect panel data model is easily affected by HLCEOs. In their presence, the WOLS estimates may produce large variances and this would lead to erroneous interpretation. Therefore, it is imperative to detect the multicollinearity which is caused by HLCEOs. The classical Variance Inflation Factor (CVIF) is the commonly used diagnostic method for detecting multicollinearity in panel data. However, it is not correctly diagnosed multicollinearity in the presence of HLCEOs. Hence, in this paper three new robust diagnostic methods of diagnosing multicollinearity in panel data are proposed, namely the RVIF (WGM-FIMGT), RVIF (WGM-DRGP) and RVIF (WMM) and compared their performances with the CVIF. The numerical evidences show that the CVIF incorrectly diagnosed multicollinearity but our proposed methods correctly diagnosed no multicollinearity in the presence of HLCEOs where RVIF (WGM-FIMGT) being the best method as it has the least computational running time.

Remote sensing techniques in mapping spatial variability of salinity in Kano River Irrigation Project (KRIP), Nigeria

Salinity has become a major issue in most large scale irrigation schemes, assessing the extent of the spread has become daunting and laborious. Remote sensing techniques were used to map salinity and develop models for extracting and identifying salinity in soils. Sentinel-2B optical imaging satellite with 13 spectral bands and 10 m spatial resolution was used. SNAP Desktop, ERDAS Imagine, and ArcGIS 10.6 software were used as the main GIS packages for building models and running functions such as input, output, analysis, and processing. Stepwise Multiple Linear Regression (MLR) techniques were carried out for the assessment of the spatial distribution of ECe and to predict salinity level at different locations of the Kano River Irrigation Project (KRIP). Four models were developed, however, due to the lower Variance Inflation Factor (VIF), model 2 which is a combination of salinity Index and band 3 (Green band) was used in delineating the spatial extent of the salinity. Close monitoring of the salt development and application of reversal measures were recommended.

Pengaruh Elektronikal Marketing terhadap Minat Wisatawan Berkunjung ke Pulau Saronde

The purpose of the authors conducted research to determine the effectiveness of E-Marketing Interests Against Tourist Visits in Gorontalo Saronde Island and can be beneficial for the institutions, the object of research, author and upcoming research. The location of the research is Saronde Island in Gorontalo. E-Marketing is one of the factors that play an important role in running a marketing business. E-marketing or Electronic Marketing is one of the breakthroughs that are quite reliable in marketing a tourism product. Saronde Island as one of the tourist attractions managed by GAB (Natural Gorontalo Maritime) uses the concept of E-Marketing as one way powerful enough to market their products. This research uses quantitative, measurement data in the form of numbers. Variables examined include Facebook marketing, email marketing, web marketing and blackberry messenger marketing. The number of samples from the total sample is 60 respondents to the normality test, autocorrelation test results and test multilinear data formula that relies on the analysis of the results of Durbin Watson and VIF (Variance Inflation Factor). With the technique of multiple regression analysis. The analysis result showed Blackberry Messenger is the most effective changer variable, which conducted to influent visiting of tourists, than others variable. The results showed a significant effect on the interest in visiting tourists at Saronde island of Gorontalo. The implications of this study are to be helpful to the reader, or for anyone who wants to continue the same thing in future research. Hypothesis Testing depends variable to the independent variable, the probability value is < 0.01.

The Effects of Omitted Variable on Multicollinearity in Hierarchical Linear Modelling

This study investigates the impact of violation of the assumption of the hierarchical linear model where covariate of level – 1 collinear with the correct functional and omitted variable model. This was carried out via Monte Carlo simulation. In an attempt to achieve this omitted variable bias was introduced. The study considers the multicollinearity effects when the models are in the correct form and when they are not in the correct form. Also, multicollinearity test was carried out on the data set to find out whether there is presence of multicollinearity among the data set using Variance Inflation Factor (VIF). At the end of the study, the result shows that, omitted variable has tremendous impact on hierarchical linear model.

Discriminant Canonical Tool for Differential Biometric Characterization of Multivariety Endangered Hen Breeds

This study aimed to develop a tool to perform the morphological characterization of Sureña and Utrerana breeds, two endangered autochthonous breeds ascribed to the Mediterranean trunk of Spanish autochthonous hens and their varieties (n = 608; 473 females and 135 males). Kruskal–Wallis H test reported sex dimorphism pieces of evidence (p < 0.05 at least). Multicollinearity analysis reported (variance inflation factor (VIF) >5 variables were discarded) white nails, ocular ratio, and back length (Wilks’ lambda values of 0.191, 0.357, and 0.429, respectively) to have the highest discriminant power in female morphological characterization. For males, ocular ratio and black/corneous and white beak colors (Wilks’ lambda values of 0.180, 0.210, and 0.349, respectively) displayed the greatest discriminant potential. The first two functions explained around 90% intergroup variability. A stepwise discriminant canonical analysis (DCA) was used to determine genotype clustering patterns. Interbreed and varieties proximity was evaluated through Mahalanobis distances. Despite the adaptability capacity to alternative production systems ascribed to both avian breeds, Sureña and Utrerana morphologically differ. Breed dimorphism may evidence differential adaptability mechanisms linked to their aptitude (dual purpose/egg production). The present tool may serve as a model for the first stages of breed protection to be applicable in other endangered avian breeds worldwide.

A Robust Approach for Identifying the Major Components of the Bribery Tolerance Index

The paper aims to emphasize the advantages of several advanced statistical and data mining techniques when applied to the dense literature on corruption measurements and determinants. For this purpose, we used all seven waves of the World Values Survey and we employed the Naive Bayes technique in SQL Server Analysis Services 2016, the LASSO package together with logit and melogit regressions with raw coefficients in Stata 16. We further conducted different types of tests and cross-validations on the wave, country, gender, and age categories. For eliminating multicollinearity, we used predictor correlation matrices. Moreover, we assessed the maximum computed variance inflation factor (VIF) against a maximum acceptable threshold, depending on the model’s R squared in Ordinary Least Square (OLS) regressions. Our main contribution consists of a methodology for exploring and validating the most important predictors of the risk associated with bribery tolerance. We found the significant role of three influences corresponding to questions about attitudes towards the property, authority, and public services, and other people in terms of anti-cheating, anti-evasion, and anti-violence. We used scobit, probit, and logit regressions with average marginal effects to build and test the index based on these attitudes. We successfully tested the index using also risk prediction nomograms and accuracy measurements (AUCROC > 0.9).

Validation of Maternal Recall of Iron Folic Supplementation During Antenatal Care in Rural Southern Nepal

Objectives This study's primary objective is to examine the validity of maternal recall of iron folate supplementation during antenatal care and factors associated with accuracy of maternal recall. Methods A longitudinal cohort design was employed for the validation study. The direct observation of all iron folate supplementation (IFA) received during all antenatal care visits at the five study health posts served as the “gold standard” to the maternal report of IFA received collected during a postpartum interview. Individual-level validity was assessed by calculating indicator sensitivity, specificity and area under the receiver operating curve (AUC). The inflation factor (IF) measured population-level bias, comparing the true coverage to the survey measure (maternal report) coverage of IFA. A multivariable log-binomial model was used to assess factors associated with accurate recall. Results The majority (95.8%) of women were observed receiving IFA during pregnancy. Women overreported the number IFA tablets received compared to what was observed during ANC visits. On average the reported number of tablets received was 45 tablets greater than the number observed. Individual-level accuracy of maternal report of any IFA receipt was moderate (AUC = 0.60) and population bias was low (IF = 1.01). However, the individual-level validity was poor across the seven IFA tablet count categories; the AUC for categories ranged from 0.47 to 0.58, indicating a performance that at best was slightly better than a random guess and at worst, misleading. Driven by the trend of maternal overreport, the inflation factor indicated that the survey measure drastically underestimated the prevalence of lower tablet categories and overestimated the prevalence of higher tablet counts. Accuracy of maternal report was not associated with months since last ANC observation nor any maternal characteristics. Conclusions Maternal report of the amount IFA supplementation received during pregnancy produced extremely biased population prevalence and performed comparably to or worse than a random guess for individual level validity. It's imperative to improve this indicator for future use, as it is included in global frameworks, initiatives and national program planning. Funding Sources This research was funded by the Bill and Melinda Gates Foundation.

Implementing kNearest Neighbor Methods to Predict Car Prices

The demand for automotive in Indonesia has never subsided, considering that the human need for transportation greatly affects people's daily lives. Various attempts are made by manufacturers to produce cars of a quality that is comparable to the costs incurred and following market demand. Prediction is a process that can be done to achieve this goal. One of the prediction methods that can be used in this case is the kNearest Neighbor. The prediction process consists of a preprocessing stage that cleans and filters unnecessary variables, followed by a variable multicollinearity test stage with Variance Inflation Factor (VIF). The multicollinearity test found 4 variables that had a specific influence in predicting the VIF value of these variables, respectively 2.22, 2.08, 1.53, 1.10 for Horse Power, Car Width, Highend, and, Hatchback respectively. The four variables of the VIF test results have a positive correlation with the price variable as the dependent variable. The prediction model is made using 4 variables selected based on the VIF test, to determine the accuracy of the method used, the Linear Regression model and, the kNearest Neighbor through the validation test with Mean Absolute Error (MAE) and R2. The kNearest Neighbor method produces an MAE test of 0.06 and R2 results are 0.843. This can be concluded if the overall kNearest Neighbor method has qualified performance in making predictions with continuous value variables or in other words using the concept of regression.