The purpose of this study is designing a model based on Tobitregression, DEA, Artificial Neural Network, Genetic Algorithm and Particle Swarm Optimization to evaluate the efficiency and also benchmarking the efficient and inefficient units. This model has three stages, and it uses the data envelopment analysis combined model with neural network, optimized by genetic algorithm, to evaluate the relative efficiency of 16 regional electric companies of Tavanir. A two-staged approach of data envelopment analysis and Tobitregression has been used to measure the effects of environmental variables on the mean efficiency of companies. Finally we use a hybrid model of particle swarm algorithm and genetic algorithm to benchmark the efficient and inefficient units. The mean efficiency of regional electric companies have increased from 0.8934 to 0.9147, during 2012 to 2017, and regional electric companies of Azarbayjan, Isfahan, Tehran, Khorasan, Semnan, Kerman, Gilan and Yazd, had the highest mean efficiency of 1, and west regional electric companies and Fars had the lowest efficiency of 0.7047 and 0.6025, respectively.
Data Envelopment Analysis is a powerful technique for measuring the relative efficiency of organizational units with multiple inputs and outputs. This approach was introduced by Charnes, Cooper and Rhodes in 1978, and is gradually becoming a useful management tool. In addition to the efficiency score, Data Envelopment Analysis (DEA) indicates targets for inefficient units. The purpo se of this paper is to investigate the performance of fifteen Kingdom of Saudi Arabia universities for the academic year 2013. The study evaluates the technical efficiency of individual Saudi Arabia universities using the nonparametric frontier methodology, the Data Envelopment Analysis (DEA). To investigate the determinants of efficiency, the study use the Tobitregression. This analysis aims to explain the variation in calculated efficiencies to a set of explanatory variables, i.e. total number of students enrolled in undergraduate, graduates, number of academic staff, number of non-academic staff, and number of campuses. In this paper two stage efficiency analyses are applied and used to compare the efficiency of fifteen Saudi Arabia universities using DEA. Two suggested DEA models were used. The study used CRS, and VRS models to measure the efficiency of fifteen Kingdom of Saudi Arabia universities. A second stage the study used the Tobitregression to determine the most environmental factors that affecting the efficiency of this university. The analysis shows that the most influential factors affecting efficiency are the new graduate, graduate last year, previous year graduates PhD, and number of teaching staff.
In this paper, we employ both the DEA framework and tobitregression methodology to evaluate the efficiency of Kenya Insurance industry. The analysis is based on the DEA-BCC model, which allows for the use of multiple inputs and outputs in determining relative efficiencies. Additionally, the study uses tobitregression model to assess the endogenous drivers of efficiency scores obtained from DEA. From 2011-2014, we estimate that in 2011, 55percent attained the efficiency frontier, 33 percent in 2012, 19percent in 2013 and 36percent in 2014. Based on these findings, it can be concluded that a significant number of insurers especially for 2012-2014 were not employing their inputs effectively and judiciously in producing the existing level of outputs. These results are consistent with the other finding of the study which indicates that in 2011, majority of the firms lie on or very close to the efficient frontier compared to the case in the later years.
Egzistuoja keletas metodų bankų efektyvumams įvertinti. Norėdami gauti daug tikslesnį įvertinimą, mes pritaikome super efektyvumo metodą. Pagal pagrindinę CCR ‐ DGA koncepciją, sprendimų priėmimo grupės (SPGs), turėdamos geriausią veiklą, gali gauti grupės rezultatą. Tai rodo jos gamybos ribotumą. Todėl efektyvumo palyginimas neįmanomas, jei yra per daug SPGs-ų, kurių rezultatas lygus vienetui. Norint išspręsti šią problemą, Andersen ir Petersen (1993) pasiūlė super efektyvumo Duomenų gaubiamosios analizės (SE-DGA) koncepciją. Neseniai, super efektyvumas buvo finansinių institucijų priemonė įvairiuose tyrimuose. Super efektyvumas turi didelį privalumą lyginant pavyzdžiais pasirinktus bankus. Iš kitos pusės, mes dažniau naudojome Tobit regresijos modelį, negu paprasto, mažiausio kvadrato (PMK) metodą, nes priklausomas kintamasis (super efektyvumas) yra teigiamas. Tobit modelį pirmasis pasiūlė Tobin (1958). Šiame modelyje slaptas kintamasis atitinką nustatytą priklausomą kintamąjį, kai slaptas kintamasis yra teigiamas. Jei slaptas kintamasis yra neigiamas, priklausomas kintamasis bus prilyginamas nuliui. Nustatytas kintamasis bus lygus 0, kai slaptas kintamasis yra neigiamas.
distribution with 1 degree of freedom. The hypothesis that α = 0 is rejected. When compared to the initial Tobitregression assuming homoskedasticity, the coefficients for β of both regressions are very similar. Also the signs of the coefficients are equal for all of the statistically significant coefficients. This hints to the conclusion that the coefficients of the Tobitregression assuming homoskedastic and normally distributed errors are close to the true values. Additionally, one has to be careful when comparing the estimates of β of both models. Although, the β coefficients may differ, the marginal effects ∂E(y|y > 0)/∂x and ∂E (y)/∂x can be very similar. As a second measure to test the robustness of the results given the found het- eroskedasticity and non-normality we employ the censored least absolute deviations estimator (CLAD). The CLAD estimator was introduced by Powell (1984). Given a data set censored at zero, the CLAD estimator is based on the minimization of the sum of absolute residuals. Unlike the Tobit model, Powell’s CLAD estimator is consistent and asymptotically normal for a wide range of error distributions. The estimator considers alternative conditional moments that are less altered by censor- ing. It assumes an error term with a conditional median equal to zero. The CLAD model is described by the equation
This research has been performed for dealing with some of the important working capital management policies and firm efficiency regarding to firm specification. For this purpose, a detailed analysis has been performed on manufacturing firms. DEA and Tobitregression analysis has been performed to achieve the objective of this study. Analysis revealed some important areas from which the firm efficiency can be improved and an optimum level can be achieved. Moreover, with the help of above stated tools, analysis has made to measure the impact of working capital management on firm efficiency. DEA results indicate that only fifteen companies require increase in inputs to attain better output whereas six companies require decrease in the input. However sixteen companies have to consistent with their existing proportionate of inputs to sustain the output maximization. Further results indicate that the input slack requires to rectify either it exists in input elements or in output elements. However inputs have greater importance for the slack values. It is because we have to redesign the policies for such firms regarding to the said input parameters. Output slack
Many studies have been conducted to analyze efficiency of railways for different countries. However, these studies have mainly focused on quantitative aspects of railway transportation and quality has been neglected. In this paper three new data envelopment analysis (DEA) models are presented. The first model is solved for assessing quality of passenger railway services in 71 countries of the world by including perceived quality of railways among outputs of DEA models for the first time in the literature. For the second model which is applied to 27 railways in Europe, a safety index is defined based on number of fatalities and serious injuries and is added as another output. Both models are solved for constant return to scale (CRS) as well as variable return to scale (VRS) setting with output orientation. The follow-up Tobitregression for the first model shows that efficiency results are positively correlated with quality of road and for the second model negatively correlated with the number of level crossings. In the third model which is applied to 19 train operating companies in the UK, passenger-km is the input and stated passenger satisfaction derived from questionnaires together with punctuality level are outputs which proved to be helpful for ranking companies based on quality of their services.
The paper also considers the capital structure of the firm. We measure the ratio of foreign currency bonds outstanding to total liabilities, and making use of a Tobitregression technique to control for left censored observations (zeros in the dependent variable) we demonstrate that firm-specific and market development variables influence capital structure. The results provide support for market depth, agency, static trade off and risk management theories, consistent with our earlier work (see Mizen et al. (2012)). When we allow for an ABF2/ABMI effects on capital structure we find this is supportive of the market depth hypothesis as the mechanism by which these reforms have influenced corporate capital structure.
This table presents multivariate regression results of …rm-wide recovery rates on lagged macroeconomic conditions, controlling for other determinants of recovery rates. The dependent variable is the …rm-wide ultimate recovery rates. Ln(Total Book Assets) is the natural logarithm of Total Book Assets, a proxy for …rm size. Debt-to-Assets is the ratio of Total Debt to Total Book Assets. Bank-Debt Share is the portion of bank debt in Total Debt. HH Index of Bank Debt is the Her…ndahl-Hirschman index of a …rm’s bank debt, measuring a …rm’s bank-debt concentration. Distressed Exchange is a dummy variable that is set to be 1 if the default leads to a distressed exchange, and 0 otherwise. The macroeconomic condition variables include GDP Growth (trailing four quarter U.S. GDP growth rates), Default Rate (Moody’s trailing twelve month issuer-weighted global speculative grade corporate default rates), Bond Spread (yield spreads between Moody’s BAA-rated and AAA-rated corporate bonds), and S&P 500 Return (trailing twelve month return of Standard & Poor’s 500 index). The lagged values are taken at the origination of the credit facility, while the non-lagged values are taken at the time of default. Panel A reports the OLS regression results, and Panel B reports the two-sided Tobitregression results. The t-statistics are reported in parentheses. The superscripts a, b, and c represent signi…cance at 1%, 5%, and 10% levels, respectively.
characteristics beyond managerial control, which reflected differences in the population demographics and regional health expenditure. Predicted efficiency scores (Stage II DEA scores) reflected the amount of efficiency that was predicted by organisational and contextual characteristics. Finally, the two-limit random effects Tobitregression, with hospitals hierarchically nested within regions was used. As explanatory variables, the average length of stay (ALS) and Case-mix Entropy (CME) were considered as organisational variables at the hospital level, which are correlates of inefficiency related to patients and payer mix ; the male youth unemployment rate (MYU), the elderly dependency rate (EDR) and the average annual per capita health expenditure in the period 1997-2007 (HE) were included , respectively, as indicators of social, demographical and economical dimensions of the regional context [44, 45]. The log-likelihood criterion has been used to assess the goodness-of-fit of models. A p value <0.05 has been considered statistically significant. Efficiency scores have been calculated for each hospital using the DEA frontier Analyst software  while two-level Tobitregression analysis was performed by using Stata/MP 11.2.
In the above regression model, the dependent variable, net amount of reinsurance, takes on the value zero with positive probability but is roughly continuous and bounded on [0,1]. For instance, of 2419 observations in the sample of private passenger auto liability rein- surance, 262 observations take a value of 0; of 2211 observations in the sample of home- owners reinsurance, 144 observations take on the value zero. Figures 8, 9, and 10 provide the histograms of net amount of reinsurance for private passenger auto liability reinsur- ance, homeowners reinsurance, and product liability reinsurance, respectively. Therefore, the dependent variable, net amount of reinsurance, is a typical corner solution outcome, which renders the coefficient estimates of generalized linear regression model inconsistent (Wooldridge, 2001). Hence, a corner solution model, or Tobit model, should be employed to deal with this kind of “censored” data. Moreover, in order to take advantage of our panel data feature, we estimate a random-effects Tobitregression of the amount of reinsurance purchased by primary insurers on a measure of individual firm risk and other exogenous economic factors. 9 If the coefficient on the measure of risk, loss reserve error, is positive and significant, we find evidence for the existence of adverse selection.
Abdul Quyyum and Khalid Riaz examined the Technical Efficiency of 28 Pakistani Banks including 6 Islamic Banks using DEA technique and Tobitregression analysis during the period 2003-2010. The model used interest income and non-interest income as outputs and labour cost, compensation to directors and executives, user cost of fixed assets, interest cost and non interest cost as inputs. The study found that public conventional banks were the most efficient banks followed by private conventional and private Islamic banks with an average bias of 10%. The results of the study revealed that conventional banks were more efficient and the study also indicated that the market share in terms of deposits, the ratio of financing to deposits (FDR) and the public ownership of the bank had statistically significant influence on the efficiency. Anil K. Sharma et al. conducted a study to examine the relationship of commercial banks efficiency with the banks specific factors like bank size, age, ownership, profitability, deposits etc. They applied Data Envelopment Analysis (DEA) technique to assess the performance of Indian banking sector and further they evaluated the association of bank specific factors with the efficiency of banks using TobitRegression Model. They found from the study that age, ownership and profitability were positively and significantly associated with the banks performance and efficiency whereas bank diversification practices were negatively and significantly affecting the banks efficiency scores. Ashish Kumar and Sunil Kumar investigated the efficiency of Indian public sector banks with the help of Data Envelopment Analysis (DEA). The data for the study related to a sample of 27 public sector banks operating in India during the period 2008 - 2009. It was found that the overall level of technical efficiency in these banks has been found to be 95.7 percent. The study identified that only 6 banks were efficient on the criteria of technical
The current study offers important new insights for work zone speed data analysis. The successful application to analyze speed data at three separate work zones with different traffic characteristics in the current study validates the use of the Tobitregression model. All three models produced similar results in demonstrating the influence of surrounding traffic on vehicle speeds. The model showed that a vehicle is more likely to speed in higher traffic volumes, where there are high proportions of other vehicles speeding, and where other vehicles are speeding by a large margin. Vehicles not in a platoon and the leaders of platoon were more likely to speed than those in the middle of a platoon. Independent of platoons, speeding was more likely where larger gaps exist. The estimated regression coefficients and their marginal effects on both the amount of excess speed and the probability of a driver being non-compliant were consistent and of plausible signs across the three work zones studied. Importantly, the modeling technique used is transferrable and may be applied in a wide range of studies to examine vehicle speeds and to evaluate effectiveness of speed- reduction countermeasures, both within work zones and elsewhere in the general road network. In doing before-after studies of speed reduction countermeasures, special considerations need to be given to any potential site-selection effects, as demonstrated by Kuo and Lord (2013).
Table 2 shows results of Tobitregression analysis. Results show that variables RATIO, EXP, RECORD, KNOW and STATUS are statistically significant at 1% level of significance. Meanwhile, variables AGE, EDU, DIST and INVOLVE are not statistically significant. These results show that farmers’ experience and socio-economic factors such as level of knowledge, record keeping and farmer’s status could affect productivity. For KNOW, it reveals that farmers who possess basic skills and knowledge on cocoa farming are more efficient and productive. This finding supports the studies by Gotland et al. (2003), Rasula et al. (2012) and Abang et al. (2014). In Malaysia, cocoa harvesting are not yet mechanized. The cocoa pods are still hand-harvested throughout Malaysia, making this industry more labor-intensive than other agricultural sectors. Thus causing productivity of labor strongly correlated with farm size and age of farmers. For record keeping, the study finds a statistically significant relationship with efficiency index of cocoa farmers. It proves that cocoa farmers with proper record keeping tend to be more efficient than farmers who do not. Finally, variables that explain education, age and involvement of spouse or partner in farming are not statistically significant in determining the technical efficiency of cocoa farmers in Malaysia. This is not surprising because most of the farmers interviewed are those age 55 and above with only primary or secondary level
First, Data Envelopment Analysis (DEA) was employed in estimating the efficiency of Bangladesh banking industry during 2008-2012. Second, Tobitregression was, first, run on each year’s technical efficiency and then it was run on the pooled data of technical efficiency for determining significant factors. There are two approaches for obtaining the efficiencies of any decision making unit (DMU) such: (i) Stochastic frontier function/method (SFM) developed byAigner, Lovell, and Schmidt (1977) and later refined by Pitt and Lee (1981) and Batties and Colie (1992). The SFM is a parametric approach. (ii) Data Envelope Analysis (DEA) method.
Abstract: This study examined the effect of adoption of improved technologies in maize-based cropping systems on income of farmers in Ondo state, Nigeria. It specifically identified the socio-economic characteristics, determine the intensity of adoption of improved technologies introduced to maize-based farmers and examine the effect of adoption of improved technologies introduced to maize-based farmers on their income in the study area. Multi-stage sampling technique was used to obtain data from 160 maize-based farmers that were selected from 3 Local Government Areas in Ondo State. Descriptive Statistical Analysis, Budgetary Technique Analysis and TobitRegression Analysis were used to analyze the data. Results on socio-economic characteristics disclosed that most (59.82 percent) of the respondents were relatively old (more than 50 years) with the mean age of 54.76 years. Also, 50.63 percent of the respondents had between 5 and 8 members in their households. It was revealed that only 29.62 percent of the respondents had secondary education and above. The farmers were well experienced with 65 percent of them having more than 20 years of farming experience and 73.75 percent of them had less than 2 hectares of land as farm size. The result of profitability analysis shows that the total revenue, gross margin and net farm income for the improved technology adopters are N750,450, N573,130 and N52,1940 respectively. On the other hand, the total revenue, gross margin and net farm income for the non-adopters of improved technology are N320,140, N244,180and N 227,830 respectively, implying that improved technology adopters performed way better than non-adopters in terms of total revenue, gross margin and total cost. The determinants of intensity of adoption of improved technologies in the study area as shown by Tobitregression estimate revealed that age, extension access, farming experience, marital status, household size, farm size and educational level were statistically significant, implying that they are the important variables found to greatly influence the intensity of adoption of improved technologies by maize-based farmers in Ondo state. It is therefore recommended that policy option requires the traditional technology users to embrace the improved agricultural technologies in order to increase their earning per unit of land cultivated should be introduced.
This study applied the Tobitregression model to estimate the determinants of intensity of adoption of the NERICA rice seeds by the households. Factors hypothesized to be influenced by policy and development partners to improve adoption and use of improved agricultural technologies were identified. Likewise, these factors will guide rice scientists, agricultural extension agents and other stakeholders in refining their research and development procedures.
In this paper we first intend to examine the probability of falling into the realm of child labor by using the conditional probability theorem. Furthermore, we will compare the extent of each factor’s effect on boys and girls using a TOBITregression model. Finally, we will analyze different aspects of Iran’s labor market to assess the future ahead of the children who are working at the moment. As the results will show, the probability of participating in the labor market conditional on belonging to the age group of 15 to 17 is 20.34 % for boys in comparison with girls that is 4.42 %. More probability scores are estimated in the paper. Moreover, the results from the TOBITregression show that in general boys are more affected than the girls by the same factors. Also, based on the macro statistics published by Iran’s Statistical Center, the graduation of numerous people from graduate schools combined with the low and slow rate of economic growth makes it quite difficult to find a decent work in the country. As a result, the skilled labor force will be content with accepting low wage jobs which are more suitable for the unskilled workers. Therefore, those who left school earlier in their lives will face several problems in the future.
Nigerian agriculture is dominated by smallholder farmers and commercialisation of smallholder agriculture by bringing the farmers to markets is taking place all over the country. This study therefore assessed the current level of crop commercialisation, analyzed variation in the level of commercialisation among households and examined the determinants of crop commercialisation among smallholder farming households. Primary data were collected from 400 selected smallholder farm households in the study area with the aid of structured questionnaire using multistage sampling procedure. Analysis was done using descriptive statistics, Household Commercialization Index (HCI) and tobitregression model. The assessment of the current level of commercialisation among the smallholder farming households showed that average HCI was 0.83. Farmers with low, medium and high HCI were 6.44%, 9.65% and 83.91%. The tobitregression analysis further showed that, age, gender, level of education, household size, membership of an association, farm size, access to credit, market distance, farm and off farm income, were associated with increase in the extent of crop commercialisation. It was recommended that farmers should increase farm size while government should provide support such as credit facilities and input subsidy to enable smallholder farmers increase the level of agricultural production.
In this dissertation a review of the literature as it applies to the modelling of educational performance data is undertaken. Statistical linear models, including the novel Beta, Tweedie and Tobitregression models, are then ap- plied to the performance data of students who have undertaken a preparatory mathematics course. These models are then critically reviewed and compared with the commonly used standard linear regression model.