Search

Highlights • School rankings based on value-added estimates are subject to prediction errors. • We use the random forest algorithm to improve predictions. • Monte Carlo simulations demonstrate the advantages of this approach. • We discuss policy implications using Italian middle school data. Abstract School rankings based on value-added (VA) estimates are subject to prediction errors, since VA is defined as the difference between predicted and actual performance. We introduce the use of random forest (RF), rooted in the machine learning literature, as a more flexible approach to minimize prediction errors and to improve school rankings. Monte Carlo simulations demonstrate the advantages of this approach. Applying the proposed method to Italian middle school data indicates that school rankings are sensitive to prediction errors, even when extensive controls are added. RF estimates provide a low-cost way to increase the accuracy of predictions, resulting in more informative rankings, and more impact of policy decisions. [ABSTRACT FROM AUTHOR]

The aim of this study was to evaluate the influence of students' reading literacy, measured by the PIRLS (Progress in International Reading Literacy Study) test, on their performance in the TIMSS (Trends in International Mathematics and Science Study) mathematics and science tests. The data on 4,125 Italian students from 199 schools were analyzed: the sample was representative of students in Grade 4, who participated both in TIMSS and PIRLS 2011. Italian descriptive results were compared with those of other countries that participated in both projects. The results showed a high correlation among all three subjects. Repeated measure variance analyses were conducted to determine the effect of reading literacy on mathematics and science performance on the Italian sample. As far as Italian students were concerned, results confirmed the influence of reading literacy on mathematics achievement, whereas the influence of reading literacy on science achievement was not clear. [ABSTRACT FROM PUBLISHER]

In this paper we investigate the relationship between social capital and cheating behaviour in standardized tests. Given the low-stakes nature of these tests, we interpret the widespread presence of cheating as a signal of low trust towards central education authorities and as lack of respect for the rule of law. We find that cheating is negatively correlated to several social capital proxies in the local environment where a school is located. We also distinguish between different dimensions of social capital: contrasting universalistic and particularistic social values, cheating appears to be negatively correlated only to measures of universalistic social values. [ABSTRACT FROM AUTHOR]

Social scientists, in particular sociologists, claim that the distinction between universalistic and particularistic values is relevant to explaining the social behaviour of individuals (and societies). This paper provides preliminary empirical evidence that supports the claim. It first defines a number of proxies for the degree of particularism embedded into long-celebrated dimensions of social behaviour (trust, political awareness, and associational activities). Then, it shows that the particularistic measures are positively correlated to each other and negatively correlated to some established generalist measures for all dimensions of social behaviour considered, both across and within countries and regions. Moreover, the paper relates that the various proxies for particularism share the same set of covariates (such as low education and income), which are neatly distinguishable from the determinants of the generalist measures. [ABSTRACT FROM AUTHOR]

The paper uses questions included in the 2010 wave of the Bank of Italy's Survey on Household Income and Wealth to investigate the role of family transmission of values. It presents three main empirical findings. First, the paper shows that a number of attitudes (generalized and personalized trusting behaviour, risk and time preferences) and outcomes (female labour force participation, fertility, entrepreneurship, productivity) are associated with the values received. Second, it documents that values received from parents are correlated with the values transmitted to descendants. Third, by using respondent moving patterns, the paper highlights that values received are slowly changing even after a discontinuity in the reference environment. Comparisons between first- and second-generation movers suggest that what matters for breaking the family chains are the formative years, when young people somehow strike a balance between the values transmitted by their parents and what they experience in the (possibly different) environment where they grow up. [ABSTRACT FROM AUTHOR]

Abstract: Of particular importance for education policy-makers is the possibility that teacher mobility adversely affects the quality of teaching in schools serving mainly disadvantaged and minority children. This paper examines the main drivers of the mobility of Italian teachers by using applications-to-transfer data. We find that teachers systematically try to move away from schools where teaching is likely to be more difficult because of the student mix or the social context of the school. Given the absence of any criteria other than seniority in regulating teachers’ allocation across schools, disadvantaged students frequently end up with less experienced teachers who are often just waiting to move elsewhere. [Copyright &y& Elsevier]

• Morbidity and mortality in children due to COVID-19 remained low • Adolescents had higher reported cases than younger children • Reported cases were lower in pediatric population than adult population • Testing protocols vary between countries; reflects bias in reported cases Few data on the COVID-19 epidemiological characteristics among the pediatric population in Africa exists. This paper examines the age and sex distribution of the morbidity and mortality rate in children with COVID-19 and compares it to the adult population in 15 Sub-Saharan African countries. A merge line listing dataset shared by countries within the Regional Office for Africa was analyzed. Patients diagnosed within 1 March and 1 September 2020 with a confirmed positive RT-PCR test for SARS-CoV-2 were analyzed. Children's data were stratified into three age groups: 0-4 years, 5-11 years, and 12-17 years, while adults were combined. The cumulative incidence of cases, its medians, and 95% confidence intervals were calculated. 9% of the total confirmed cases and 2.4% of the reported deaths were pediatric cases. The 12-17 age group in all 15 countries showed the highest cumulative incidence proportion in children. Adults had a higher case incidence per 100,000 people than children. The cases and deaths within the children's population were smaller than the adult population. These differences may reflect biases in COVID-19 testing protocols and reporting implemented by countries, highlighting the need for more extensive investigation and focus on the effects of COVID-19 in children. [ABSTRACT FROM AUTHOR]