International Journal of Statistical Sciences https://www.banglajol.info/index.php/ijss <p>Official journal of the Department of Statistics, University of Rajshahi, Bangladesh. Full text articles available.</p> <p><a href="http://creativecommons.org/licenses/by-nc-sa/4.0/" rel="license"><img style="border-width: 0;" src="https://i.creativecommons.org/l/by-nc-sa/4.0/88x31.png" alt="Creative Commons Licence" /></a></p> <p>International Journal of Statistical Sciences (IJSS) retains the copyright of the contents of this journal but grants the readers the right to use the contents with terms and conditions under a creative common attribution licenses 4 of Attribution, Share Alike and Non commercial type(CC-BY-NC-SA) that allows copy, distribute, display, and perform the work and make derivative works based on it only for noncommercial purposes. <br />International Journal of Statistical Sciences is licensed under a <a href="http://creativecommons.org/licenses/by-nc-sa/4.0/" rel="license">Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License</a>.</p> Department of Statistics, University of Rajshahi, Rajshahi-6205 en-US International Journal of Statistical Sciences 1683-5603 Bayesian Analysis of Singly Imputed Synthetic Data under the Multivariate Normal Model https://www.banglajol.info/index.php/ijss/article/view/70112 <p>We develop appropriate Bayesian procedures to draw inference about the parameters under a multivariate normal model based on synthetic data. We consider two standard forms of synthetic data, generated under plug in sampling method and posterior predictive sampling method. In addition to point estimates of the mean vector and dispersion matrix, Bayesian credible sets for the mean vector and the generalized variance are also provided under both the scenarios. The analysis in the case when some (partial) features are sensitive and need to be hidden is also briey indicated.</p> <p>Vol. 23(2), November, 2023, pp 1-18</p> Abhishek Guin Anindya Roy Bimal Sinha Copyright (c) 2023 Department of Statistics, University of Rajshahi, Rajshahi https://creativecommons.org/licenses/by-nc-sa/4.0 2023-11-30 2023-11-30 23 2 1 18 10.3329/ijss.v23i2.70112 Analysis of Mixed Effects Model for Randomized Block Designs with Both-sided Neighbor-Effects https://www.banglajol.info/index.php/ijss/article/view/70128 <p>In this study we confine our attention to Randomized Block Designs (RBDs) with random block effects in the presence of both-sided neighbor-effects and focus our attention to (i) estimation of variance components, and (ii) estimation of treatment contrasts as well as both left-sided and right sided neighbor-effects contrasts. We start with a simple set-up of 4 blocks and 4 treatments in an RBD. This we do by concurrently studying three distinct RBDs with the same parameters. However, the allocations of Left Neighbor (LN)- and Right Neighbor (RN)-Effects are assumed to be different across the RBDs.</p> <p>International Journal of Statistical Sciences, Vol. 23(2), November, 2023, pp 19-30</p> Sobita Sapam Bikas Kumar Sinha Copyright (c) 2023 Department of Statistics, University of Rajshahi, Rajshahi https://creativecommons.org/licenses/by-nc-sa/4.0 2023-11-30 2023-11-30 23 2 19 30 10.3329/ijss.v23i2.70128 Outliers as a Source of Overdispersion in Poisson Regression Modelling: Evidence from Simulation and Real Data https://www.banglajol.info/index.php/ijss/article/view/70105 <p>The Poisson regression model is a well-known technique for modelling count data. However, it is necessary to satisfy the overdispersion assumption in order to fit the Poisson regression model. Due to the overdispersion problem in the Poisson regression model, standard errors might be underestimated, which may lead to a highly misleading inference. There are several tests in the literature to check the presence of overdispersion in the Poisson model. In this study, we apply a regression-based t test to identify the overdispersion. The simulation study and real data example clearly show that the overdispersion in the Poisson model is caused by the existence of outliers.</p> <p>International Journal of Statistical Sciences, Vol. 23(2), November, 2023, pp 31-37</p> Sohel Rana Abu Sayed Md Al Mamun FM Arifur Rahman Hanaa Elgohari Copyright (c) 2023 Department of Statistics, University of Rajshahi, Rajshahi https://creativecommons.org/licenses/by-nc-sa/4.0 2023-11-30 2023-11-30 23 2 31 37 10.3329/ijss.v23i2.70105 An Empirical Study of Middle-income Farmers on Millet Production in Kano South Nigeria: A Stochastic Frontier Analysis Approach https://www.banglajol.info/index.php/ijss/article/view/70108 <p>This paper aimed at assessing the technical efficiency of millet farmers in Kano south Nigeria using the stochastic frontier production function from the parametric perspective. The determinants of technical efficiency, such as, education, extension access, access to loans, farm size, labour etc, which played a great role in technical efficiency level, are assessed to measure the level of efficiency of the middle-income farmers. 227 middle income farmers were selected for the study and provide an empirical analysis of the efficiency determinants which aim to find a way of improving the millet production and productivity. The results show that middle income farmers’ productivity is very weak ranging from 0.023t/ha to 5.62t/ha, averaging 1.07t/ha. While the technical efficiencies of middle-income farmers range from 0.013 to 0.930 with mean of 0.621. However, the results also indicate that most of the above determinants have negative effect on technical efficiency for the farmers. So, there is need by the government at all levels to review the existing agricultural policy that will favour the reality on the ground which will result in improving the output level of the middle-income farmers with their minimum available input.</p> <p>International Journal of Statistical Sciences , Vol. 23(2), November, 2023, pp 39-46</p> M E A Khondaker M U Baba Copyright (c) 2023 Department of Statistics, University of Rajshahi, Rajshahi https://creativecommons.org/licenses/by-nc-sa/4.0 2023-11-30 2023-11-30 23 2 39 46 10.3329/ijss.v23i2.70108 Prediction Ranifall Using Machine Learning Algorithms: Empirical Evidence from Bogura and Rangpur District of Bangladesh https://www.banglajol.info/index.php/ijss/article/view/70110 <p>Around the world, forecasting rainfall has been regarded as one of the most difficult task. Exact and timely rainfall forecasting may be extremely helpful. By uncovering novel links between the readily available elements of historical data, data mining algorithms may accurately anticipate the amount of rainfall. Therefore, it remains intriguing to forecast rainfall data with both the highest degree of accuracy by combining and improving various data mining approaches in case of different weather stations. In this study we compare the forecasting performance of different data mining techniques such as Classification and Regression Trees (CART), Support Vector Machine (SVM), K-Nearest Neighbors (K-NN), Random Forest (RF), and Linear Discriminant Analysis (LDA) in case of Bogura and Rangpur district of Bangladesh. For this analysis, the monthly time series data from January 1964 to December 2017 are taken into account. For empirical investigations, the data mining process, including data collection, data pre-processing, modeling, and assessment, is closely adhered to. The empirical study shows that SVM approach is the best option for predicting rainfall in the case of both Bogura and Rangpur district, Bangladesh, for the next time period. The above study will be useful in providing information to support crop, water, and flood control, which will protect people's lives and property and promote economic in its growth.</p> <p>International Journal of Statistical Sciences , Vol. 23(2), November, 2023, pp 47-62</p> Md Rafiqul Islam Md Mostafizur Rahman Afsana Afruz Md Abdul Khalek M Sayedur Rahman Copyright (c) 2023 Department of Statistics, University of Rajshahi, Rajshahi https://creativecommons.org/licenses/by-nc-sa/4.0 2023-11-30 2023-11-30 23 2 47 62 10.3329/ijss.v23i2.70110 Multilevel Approach of Factors Influencing Child Marriage among Bangladeshi Women: Data from the 2017-18 Bangladesh Demographic and Health Survey https://www.banglajol.info/index.php/ijss/article/view/70129 <p>Child marriage is defined as the marriage of individuals below the age of 18 and is considered a violation of children's rights. The study aimed to determine the prevalence of child marriage among women in Bangladesh and identify the associated factors. This study utilized data from the Bangladesh Demographic and Health Survey (BDHS-2017-18) to analyze a sample of 18,851 ever-married women in Bangladesh. Chi-square test and multilevel logistic regression model was used in this study. By employing this statistical approach, the study aimed to uncover significant factors that contribute to the occurrence of child marriage in the country. The prevalence of child marriage among Bangladeshi women was 73.6%. The multilevel logistic regression model revealed several significant risk factors for child marriage among women in Bangladesh were; division (p&lt;0.01), place of residence (p&lt;0.01), women's and their husbands' education level (p&lt;0.01), number of family members (p&lt;0.01) and religion (p&lt;0.01). The median age at first marriage among Bangladeshi women has been decreasing but still it is a great problem in the nation. This study emphasize the need for targeted interventions and policy measures to address the root causes of child marriage, such as improving access to education, promoting gender equality, and raising awareness about the negative consequences of early marriage on the well-being and development of children. Efforts should be made to empower women and girls, enhance their agency, and create an enabling environment that protects their rights and promotes their overall well-being.</p> <p>International Journal of Statistical Sciences, Vol. 23(2), November, 2023, pp 63-73</p> Md Ariful Islam Md Kamrul Islam Abu Sayed Md Al Mamun Md Sohel Rana Md Golam Hossain Copyright (c) 2023 Department of Statistics, University of Rajshahi, Rajshahi https://creativecommons.org/licenses/by-nc-sa/4.0 2023-11-30 2023-11-30 23 2 63 73 10.3329/ijss.v23i2.70129 Health Performance of the Districts in India: An Empirical Analysis of NFHS-5 Data https://www.banglajol.info/index.php/ijss/article/view/70130 <p>Development and health, specifically child health, can be multi-dimensional and need to be investigated more deeply specially in a developing nation like India. This paper tries to find out the best and the worst performing districts using NFHS-5 district level percentage figures. First, some groups of variables named ‘development’, ‘health’ and ‘health behavior’ were created by taking similar types of variables relevant for the group and then the variables within each group were combined through principal component analysis. The districts were ranked according to the performance of each group and then these ranks were combined to get the overall rank to choose the best and the worst performing districts. We see that among top ten best performing districts, five are from Tamil Nadu, four are from Kerala and one is from Punjab, whereas most of the ten worst performing districts are from Jharkhand and Bihar and the rest are from UP, Assam and Meghalaya. Since another objective of the paper is to find the variables related to the child health scores, this variable was regressed on scores of other groups along with state dummies. Surprisingly all the variables relating to mothers like mother’s health and health facilities, health behaviour, mother being not underweight and mother being not anemic turned out to have significant positive effect on child health. Thus, mothers are the key factor towards the good health of children, and it is the mothers who should be given maximum attention to.</p> <p>International Journal of Statistical Sciences, Vol. 23(2), November, 2023, pp 75-86</p> Manoranjan Pal Ravi Pandey Premananda Bharati Susmita Bharati Copyright (c) 2023 Department of Statistics, University of Rajshahi, Rajshahi https://creativecommons.org/licenses/by-nc-sa/4.0 2023-11-30 2023-11-30 23 2 75 86 10.3329/ijss.v23i2.70130 Fitting Statistical Distributions to Rainfall Data with Different Estimation Techniques: An Empirical Study from Pabna and Dinajpur Districts https://www.banglajol.info/index.php/ijss/article/view/70131 <p>Bangladesh is a country of diverse climatic conditions because of its rainfall and other geographical conditions which have a complex impact on economic and social aspects. The statistical distributions are used in many real life data for modeling and predicting. Knowing the real distribution of rainfall rather than depending on basic summary statistics would improve many uses of rainfall data. The aim of this paper is twofold: first, the performance of different statistical distributions such as Normal, Log-Normal, Gamma Weibull, and Gumbel are compared for modeling the monthly rainfall data from Pabna and Dinajpur districts from January 1971 to December 2015; second, the performance of the Maximum Likelihood Estimation (MLE), Quantile Matching Estimation (QME), and the Maximum Spacing Estimation are also compared for fitting these statistical distributions. The empirical study showed that Gamma distribution performs better for fitting the monthly rainfall data for both Pabna and Dinajpur districts of the three methods like Maximum Likelihood Estimation (MLE) method, the Quantile Matching Estimation (QME), and the Maximum Spacing Estimation (MSE) method. The Normal distribution performs worse of these study areas. By the comparison of these three methods we found that Maximum Likelihood Estimation (MLE) gives better results. This study provides the actual distribution of rainfall data of these study areas.</p> <p>International Journal of Statistical Sciences, Vol. 23(2), November, 2023, pp 87-106</p> Afsana Afruz Md Mostafizur Rahman Md Rafiqul Islam Md Abdul Khalek M Sayedur Rahman Copyright (c) 2023 Department of Statistics, University of Rajshahi, Rajshahi https://creativecommons.org/licenses/by-nc-sa/4.0 2023-11-30 2023-11-30 23 2 87 106 10.3329/ijss.v23i2.70131 Logistic Regression Analysis of Mode of Delivery Influencing Nutritional Status among Under-Five Children in Bangladesh: Survey in Rural Area of Rajshahi District https://www.banglajol.info/index.php/ijss/article/view/70132 <p>Caesarean delivery is a surgical procedure in which a fetus is delivered through an incision in the mother's abdomen and uterus. If there are no complications with women pregnancy or labor, a vaginal birth is safer than a caesarean birth. Worldwide the rates of C-section have gradually increased in the latest decades. The aim of the study was to determine the prevalence of caesarian delivery, and to investigate the impact of caesarian delivery on nutritional status of under-five children in the rural area of Rajshahi district, Bangladesh. This was a cross sectional study. Data was collected from Rajshahi district using multistage sampling. A total number of 540 mothers having at least one under-five children currently living together were considered. In order to data analysis, descriptive statistics, Z-proportional test and both univariable and multivariable logistic regression models were in this study according to our objectives. This study revealed that the prevalence of caesarian delivery among mothers living in the rural area of Rajshahi, Bangladesh was 34.4%. The prevalence of stunting, underweight and wasting among under-five children were 43.70%, 39.63% and 27.96% respectively. The stunting, underweight and wasting under-five children were 32.6%, 28.0% and 29.1% respectively who were born by caesarian delivery, these were significantly (p&lt;0.01) lower compared to children who were born by vaginal delivery. However, the logistic model demonstrated that children who were born by vaginal delivery were more likely to have underweight [COR= 1.617; 95%CI: 1.114-2.347; p&lt;0.05] compared to children who were born by caesarian delivery. After controlled the effect of other independent variables such as mothers’ age, education, occupation, nutritional status, fathers, education, occupation, number of family members, type of family and social class, found that children who were born by vaginal delivery had a 1.589-fold higher chance to have underweight compared to children who were born by caesarian delivery [AOR= 1.589; 95%CI: 1.063-2.377; p&lt;0.05]. This study found the high prevalence in stunting, underweight and wasting among under-five children in the rural area of Bangladesh. Bangladesh government and other non-government organizations should take special attention to overcome the problem.</p> <p>International Journal of Statistical Sciences, Vol. 23(2), November, 2023, pp 107-116</p> Md Aslam Hossain Md Reazul Karim Abu Sayed Md Al Mamun Md Saimul Islam Md Golam Hossain Copyright (c) 2023 Department of Statistics, University of Rajshahi, Rajshahi https://creativecommons.org/licenses/by-nc-sa/4.0 2023-11-30 2023-11-30 23 2 107 116 10.3329/ijss.v23i2.70132 Influence of Sex Preference on Demand for a Child in Bangladesh https://www.banglajol.info/index.php/ijss/article/view/70133 <p>If son preference were to affect the demand for additional children then, within any parity, those with one or more sons would be more likely not to want more children as compared to those who did not have any sons. Conversely, if the desire for a balance sex composition was to affect fertility within a given parity, respondent who had either all sons or all daughters would be more likely to want additional children as compared to those who had children of both sexes. Data from BDHS 2017 clearly indicates a significant evidence of son preference in Bangladesh. The findings indicate that whenever gender preferences would be eliminated, the percentage of respondents who did not want more children would be increased approximately by 10 percent. Finally, we conclude that further reduction of fertility to achieve replacement level is unlikely without considerable reduction in the desire of male children or sex preferences.</p> <p>International Journal of Statistical Sciences, Vol. 23(2), November, 2023, pp 117-127</p> Saroje Kumar Sarkar Copyright (c) 2023 Department of Statistics, University of Rajshahi, Rajshahi https://creativecommons.org/licenses/by-nc-sa/4.0 2023-11-30 2023-11-30 23 2 117 127 10.3329/ijss.v23i2.70133 Day Level Forecasting for COVID-19 Pandemic Spread in SAARC Countries https://www.banglajol.info/index.php/ijss/article/view/70135 <p>The COVID-19 is declared a global epidemic by the World Health Organization (WHO). Till now there is no effective medicine and vaccine is not available to treat COVID-19 and the genetic mutation capability of COVID-19 is very aggressive. Day-level knowledge of COVID-19 spread distribution is essential to measure the nature of the virus. The aim of this paper is two folds. Firstly, we compare the forecasting performance of three models namely Simple Exponential Smoothing (SES), Moving Average (MA) and Weighted Moving Average (WMA) for South Asian Association for Regional Cooperation (SAARC) countries. Secondly, the best forecasting model is used to predict the day level number of infected case for SAARC countries. In this study we consider the daily infected case from March 08, 2020 to May 18, 2020. The empirical analysis showed that the SES model give beter forcasting performance for all of the SAARC countries based on graphical comparison as well as Mean Absolute Percentage Eorror (MAPE), Mean Absolute Deviation (MAD), Mean Square Error (MSE). The one month aheah forecast result of infected case for Bangladesh, India, Pakistan, Bhutan, Nepal, Maldives and Afghanistan will be 71702, 246726, 90926, 26, 1339, 1490, 2152 and 17934 respectively. Besides these we also shows the growth rate of the infected case based on Exponential and logarithmic scale. Having access to skilful daily updated forecast value of infected case could help better informed decision on how to manage the spread of COVID-19.</p> <p>International Journal of Statistical Sciences, Vol. 23(2), November, 2023, pp 129-142</p> Md Mostafizur Rahman Md Giash Uddin Mohammad Rashedul Islam Md Kaderi Kibria Md Nurul Haque Mollah Copyright (c) 2023 Department of Statistics, University of Rajshahi, Rajshahi https://creativecommons.org/licenses/by-nc-sa/4.0 2023-11-30 2023-11-30 23 2 129 142 10.3329/ijss.v23i2.70135