Sometimes people assume that data science and related areas are all about consumer facing businesses trying to learn ways to sell more product. I’m not sure whether this viewpoint is more naive or more cynical. Data science is about translating raw data into knowledge and insight to enable better decision making. Therefore it has wide and varied application. Data science and related fields like data mining, machine learning and big data have huge potential to drive innovation in social enterprise and business for social good.
Education too is being impacted by the data revolution. This isn’t something that may happen in the future. It has already begun. Schools in America are already using systems that combine data points like attendance and grades to predict which students are at risk of school dropout years in advance.
Two research areas have grown out of the application to education of methodologies like analytics and data mining. Learning analytics and educational data mining share considerable overlap but educational data mining places more emphasis on automated knowledge discovery and intervention.
There are many possible ways that the application of data science techniques can benefit educational institutions.
- Improve student retention, minimise students exiting courses early and increase course completion rates.
- Analyse and predict student performance,educator performance and institutional performance.
- Improve student engagement with educational software.
- Help understand how social learning occurs across various learning spaces.
- Help improve teaching practice.
- Drive efficiency in administration procedures for example streamline financial functions.
Organisations in one sector of the education system may be more likely to prioritize certain organisational goals associated with data analysis than those in another. Large third level institutions for example may be keen to streamline administration or look at ways of increasing donations from alumni. However all institutions are likely to be interested in a goal such as improving student retention.
Consider the case of early school leaving, a phenomenon that generates significant societal costs and to which much resources have been devoted to tackling. There is a lot of research to show that school dropout, rather than a single event, is the culmination of a process of disengagement from school that may be evident from early childhood. Early delivery of resources at those who are at risk of later school dropout is likely the most effective way to reduce early school leaving rates. This is the idea behind DEIS schools in Ireland and although successful to some extent, can this scheme be considered a somewhat blunt instrument in its current form? Possibly so when one considers the fine grained capability of data analytics to identify those at risk of later school drop out at an early stage so that resources can be targeted to those who need them most.
The Department of Education have announced changes to the way DEIS will operate from September 2017 in an effort to make the system more flexible and to deliver resources more closely matched to identified needs. While there may be some initial difficulties in implementation, decisions about education policy, be they at national regional or local level, will be increasingly informed by insights gained from data science.