Data Analysis

.........................................................................................................................................................................
RM Data Solutions’ work in using data analyses to support their customers and stakeholder groups in making informed decisions falls into two categories.

• Providing tools and data portals which enable customers and their stakeholders to interact effectively with their data
• Employing statistical techniques to extract knowledge from data.

The facilities we provide to stakeholders to interact with data are covered under the section on dissemination. This section focuses on the use of statistical techniques to inform decisions.

Demonstrated Data Analysis Success

.........................................................................................................................................................................

Use of multiple regression models to measure school performance and inform improvement strategies (DENI)

As part of the eSchools project, RM Data Solutions has developed a robust statistical model for Northern Ireland measuring school performance in terms of how much value the school adds to their children’s attainment between key stage 4 and key stage 5. This is used to underpin further analyses to identify particular areas of weakness for management focus.

Using data to inform teaching in the classroom (OCR)

RM Data Solutions proposed and created Active Results; an application allowing teachers delivering OCR qualifications to identify trends and cater their teaching style to specific needs within the classroom. Users of the system can now drill down into a particular subject of the curriculum, such as science, to look at student performance in composite units, or even individual questions within a unit.

Underpinning the application are a number of statistical techniques such as benchmarking against national averages with tests of statistical significance and value added measures at the level of individual tests.

These simple measures provide a new type of insight for schools, enabling teachers to identify where their pupils have particular strengths and where they are not performing as well as might be expected. This knowledge is used to review teaching methods and to inform decisions for individual pupils.

Improving Estimates of Adult Attainment (BIS)

On behalf of BIS (Department for Business, Innovation and Skills) RM Data Solutions have proposed a number of options for improving estimates of attainment levels among the adult population of England derived from the Labour Force Survey (LFS). These include the imputation of attainment levels among those who respond by proxy using Multiple Imputation (MI) methods which are becoming increasingly common in social research. Levels of attainment reported by proxy tend to be low, and our imputation method, which controls for the background characteristics of respondents, attempts to correct for this bias.

Automating the Calculation of Confidence Intervals from Complex Survey Data (HSE)

On behalf of HSE, we have designed an online flexible reporting solution, using OLAP, to present a range of tables and charts to stakeholders. This includes analysis of work related injury and illness derived from the Labour Force Survey (LFS) that includes confidence intervals associated with central estimates that take account of the sample survey design of LFS. Our OLAP solution includes an automated method of calculating a linearised Taylor-Series variance estimate, normally only available in statistical software packages.