You are here:
- Home »
- Using evidence for learning »
- Gathering evidence »
- Types of analysis
Consider the evidence: Types of analysis
- How to use this resource
- Introduction – Evidence-driven decision making
- What is meant by "data and other evidence"
- What evidence does a school have?
- What can we do with evidence?
- Evidence-driven strategic planning
- The evidence-driven decision making cycle
- Evaluation and reflection
- Types of analysis
- Terminology
- Resources
Before we move on to the various stages of the decision-making cycle, let’s look at how to analyse data. You can compare achievement data by subject or across subjects for an individual student, groups of students, or whole cohorts.
Again, questions are important. The type of analysis you use depends on the question you want to answer.
Three of the most common types of analysis used in schools:
- Inter-subject analysis
- Intra-subject analysis
- Longitudinal analysis
The examples below use only student achievement data in order to highlight the different approaches.
Inter-subject analysis
Have my students not achieved a particular history standard because they have poor formal writing skills, rather than poor history knowledge?
We can explore this question this by cross referencing the history results of individual students with their formal writing results in English.
If the trend is for your students to do less well in formal writing than in other aspects of English and/or other aspects of history, the answer could be 'Yes'.
Intra-subject analysis
What are the areas of strength and weakness in my own teaching of this class?
You could compare the externally assessed NCEA results of your students with results from appropriately matched schools using published national data. Where any differences are greater or less than average, areas of strength/weakness may exist.
For example, you might find that, on average, your students have gained credits at a rate five percentage points better than the comparison schools. But in one standard the difference is 15 points, indicating a possible area of strength. In another standard, there is zero difference, indicating a possible area of weakness.
Longitudinal analysis
Are we producing better results over time in year 11 biology?
You can compare NCEA biology results for successive year 11 cohorts at this school with the national cohorts for successive years. But the results for the national cohorts might be improving too. So you need to work out how the school’s cohort differs from the national cohort in each year.
If the school’s rate of change is better than the change for the national cohort, the answer could be 'Yes'. To be sure that your teaching is influencing this improvement, you should look at the overall levels for achievement of each cohort. Otherwise, any improvement could be a result of having more able students in a cohort, rather than better teaching.
Longitudinal analysis of student performance is more demanding. That would require comparable assessments over successive years – NCEA results at successive levels probably would not suffice for this.
This is a good time to think about NCEA data. They have been described as "both rich and subtle". Your school’s NCEA data give you access to a huge amount of fine-grained information. You can also aggregate NCEA data to show trends, etc – but you need to be careful that in aggregating the data you don’t lose the subtlety and even produce misleading information.
Education Counts has for information relating educational statistics and research in New Zealand.