Collecting internal school data
External student assessments such as NAPLAN and the HSC can be useful as a snapshot of learning at a point in time, but they are also limited, because they do not cover every scholastic year or the full range of subjects taught in schools. Because of this, internal qualitative and quantitative data that is collected within schools is a useful addition to the information from external assessments. This page describes some principles of data collection that may help to get the most out of the internal data your school is already collecting as evidence of student achievement. The principles are also relevant for other data you may be collecting, for example focus groups or parent surveys.
Download the Evidence Guide for School Excellence - Principles for collecting internal school data (DOCX 75.9 KB) or refer to the online version below.
Uses of internal data
Teachers can use internal data:
- As formative assessment (assessment for learning) – using evidence about students’ knowledge and skills to inform teaching. This usually occurs throughout the teaching and learning process to clarify student learning and understanding. For more information about formative assessment, see CESE's report on assessment.
- As summative assessment (assessment of learning) – using evidence of students’ learning to assess achievement against outcomes and standards. This usually occurs at defined points during a unit of work, term or semester and may be used to grade students. For more information about summative assessment, see CESE's report on assessment.
- To compare across cohorts – has this year’s group of students arrived in my class more or less prepared than last year’s? How can I change my lesson plans to reflect this?
- To experiment with teaching practices – if I teach a certain topic in a new way this year, do my students perform better on assessments than last year’s group, or students in other classes?
- To collaborate with other teachers – does the data indicate that students in other classes seem to understand particular topics more than in my class? What are those teachers doing, and can I replicate that for my students?
- To report to parents about individual students – what learning progress has been made, and what potential future areas of focus might there be?
School leaders can use internal data:
- To supplement external assessments like NAPLAN and HSC to get a picture of the performance of the whole school – are we supporting learning growth equally well in each Key Learning Area?
- To compare faculties or classes – are there particular classes that require additional support this year?
- To report to the community – what do our progress and achievements look like for subjects where assessment data is not typically made public?
Principles for collecting internal school data
If you are comparing student performance or growth data across years or between classes, you can be more confident that any changes you observe are due to differences in practice (instead of quirks in the data) if all of the information was collected in the same way each time. If you are conducting interviews, surveys or focus groups, keeping your method the same each time will similarly allow you to aggregate responses and compare trends over time.
Try to use the same assessment methods and same marking rubrics, and conduct the assessments at the same time during the year. One way to keep performance data somewhat consistent even when assessment methods differ slightly (for example, between classes) is by aligning tasks and assessments against curriculum outcomes.
Some assessments (such as NAPLAN) are undertaken by almost all students in your class or school. However, some data may be for only a selection of individuals. When some people are captured by the data collection and some are not, there is a risk that generalising to the entire school based on only those for whom you have data will be misleading. A common circumstance where this may occur is with student, staff or parent surveys. However, it can also occur when assessments are tied to participation in non-compulsory programs or subjects (such as MultiLit, or the HSC). Where possible, attempt to gather data from as many people from your target audience as possible (for example a whole class or stage). Where this is not possible (for example with parent surveys), be aware of the potential limitations of the data, think about how people not in your sample might be different from the people who are and consider whether there are other ways to engage with them.
Captured data will not be beneficial if it is not used. This means that the easier the data is to analyse, the more valuable it will be. When analysing data across the whole school, it is especially important to be able to perform calculations, create graphs and combine different datasets together. Many schools already use specialised software to capture student report data (which contain qualitative and quantitative components). However, often an excel spreadsheet will do just as well for this purpose. If you are currently recording assessment information with pen and paper, then capturing this data electronically instead is unlikely to take you additional time – and may save you time in the long run.
Much of the data you might be collecting on students, staff or parents may be personal or sensitive information. This could include student performance data, demographic characteristics, or responses to surveys. It is important to let people know why you want to collect their information and what you will use it for, handle the data carefully, and only provide access to people who are authorised to use it. Ethical use of data requires the process to be open and transparent. For more information on privacy obligations, see the Information and Privacy Commission’s information protection principles.
Sometimes when collecting data there is a temptation to collect or record a lot of extra information ‘just in case’. This can mean it takes a lot of extra time to record, manage, or analyse data, or can lead to surveys becoming very long, which may annoy the people taking them and lead to a smaller response rate. In order to avoid this problem, start out thinking about how you want to use the information. Once this is clear, collect the minimum amount of data that will allow you to answer those questions.