Диссертация (1136614), страница 23

Просмтор этого файла доступен только зарегистрированным пользователям. Но у нас супер быстрая регистрация: достаточно только электронной почты!

Текст из файла (страница 23)

This result holds true for all ​k​ = 1,..,100.Finally, we didn’t find any significant correlation between educational outcomes of a school andthe price of ​k​ closest apartments (​r​ = 0.053, ​p​-value = 0.20 for ​k​ = 20) The correlation is notsignificant for all ​k​.DiscussionWe demonstrate that schools are segregated in the digital space: schools with similar academicperformance are closer to each other in terms of number of online friendship ties.

We show thatthis result cannot be explained by geographical position of schools or the affluence of schoolneighborhoods. The single digital metric (eigenvector centrality) explains as much as 30% of thevariation in educational outcomes. That is comparable with the total effect of socio-economiccharacteristics of school students as reported in the past research. These results indicate that,despite the apparent democratizing nature of the internet, information technologies may amplifyinequality rather than mitigate it.

The results may have important implications for educationalpolicy and for understanding of inequality reproduction in the digital age.Figure 1. School network. Each circle represents a school. Different colors correspond todifferent administrative districts. A tie between two schools exists if there is at least onefriendship tie between students from these schools. For visual clarity only strong connections (atleast 4 students involved) are shown.Figure 2. Centrality in the digital space is a good predictor of educational outcomes (a) whilethere is no significant correlation between geographical centrality and educational outcomes (b).Red color corresponds to schools in the city suburbs that are formally part of the city.

If they arecombined with other schools the resulting correlation is -0.071 and still not significant (​p​-value =0.08)Figure 3. The probability of a friendship tie between two schools as a function of the distancebetween these schools. For close schools the probability is 0.75 and then it declines with distancefollowing power law (inset).Figure 4. Schools are segregated in the digital space. Academic performance of a school issimilar to academic performance of its closest digital neighbors (a). This segregation cannot beexplained by the geographical segregation. (b)..

Характеристики

Список файлов диссертации