Issues in Level 3 NCEA Statistics Internal Assessment

Education Committee of NZ Statistical Association (NZSA), November 2014.

The education committee of NZSA are aware of teacher concern about several related issues that arise when they assess using the revised internal achievement standards in Statistics. The committee offers this discussion paper as a source of possible solutions for these issues. The paper has had input from the whole committee, and particularly from members Pip Arnold, Derek Smith, Michelle Dalrymple, and Ruth Kaniuk. The issues include:

  • amount of research into background, length of student scripts, and referencing
  • possible assessment structures
  • ensuring authenticity
  • assessment opportunities within a year and across years
  • assessment re-use across years.

The current statistics achievement standards provide the opportunity to use different assessment conditions from what may have traditionally been used within schools. This is especially true at level 3. The plan therefore is to give students every opportunity to show what they are capable of achieving, whilst ensuring the assessment conditions meet the requirements of NZQA, and individual school policies.

Background information and student research

The level 3 statistics internal standards require students to research the given context to gain an understanding of the given context. The choice of context is crucial to allow students to be successful by being able to relate to their prior knowledge and/or their interests. For example, taking a context such school attendance rate and GPA allows students to bring quite a lot of contextual understanding to the situation compared to polar ice for which they may have to complete research to gain a basic understanding of the key issues.

The purpose of having students research the context is to enable them to make sense of the data and its stories so they can write sensible and meaningful contextual statements. To this end the students need an overview of the key areas of the topic, but they do not need to become experts in the topic or the process of research itself.

The context provided to students should be one that is accessible and able to be understood by students. What this means in a particular school will depend on the students in that school. Students’ literacy level needs to considered, as well as their real-world experiences. Allowing some students to use Wikipedia as their basis for research might be sufficient, whereas for other students providing them with links to original academic research abstracts may be just as appropriate. Time needs to be allowed within the assessment planning for students to complete this research (see suggestions below for different assessment styles).

Referencing

Students should reference their research sources in their report. It is not expected that this be a formal process, but sufficient for a “check” on their research if needed. Students might provide a list of websites, add footnotes to their report, or simply hand in a copy of the material they used when submitting their assessment.

Written length of assessments

From experiences to date, page or word length guidelines may be appropriate to help students focus on the statistical aspects of their reports and not provide unnecessary or irrelevant contextual background. A suggested page or word length helps students to become more discerning writers, a useful skill when they are heading off to university and the world beyond school. This needs to be balanced between teacher workloads and allowing students to provide sufficient evidence towards the achievement standard.

Possible assessment structures

Remember the conditions of assessment document states for all level 3 statistics standards that “sufficient time should be allowed for students to complete the investigation”.

There are many different styles of assessments that allow students the potential to access all grade levels. It is also good practice to make a (generic) marking rubric available to students well before the actual assessment. Some suggestions for varied assessment styles include:

  • a week long, open-book assignment
  • multiple period assessment with students given the context prior to the start of the assessment, and allowed to bring in their contextual research and a page of notes
  • student blogs, videos or other presentation method for example, orally
  • portfolio-type assessments (this could lead to a good discussion with your technology or art departments which routinely use this type of assessment)
  • carry out a survey with students at the school - the data set collected can be magnified into a much bigger set with some careful manipulation. This works well at large schools, but would have to be done carefully with smaller schools.

Ensuring authenticity of student work

Often one of the main concerns with moving to a more open style of assessment is how to ensure authenticity of students’ work. As well as students signing an authenticity declaration to hand in with their work, suggestions to alleviate this concern include:

  • reserving the right to interview “random” students after submission
  • having a one period closed assessment where students use their main assessment to write the media release for their report
  • completion of work towards the achievement standard in class, check (milestone), before allowing some time outside of class to finish off (to allow for research)
  • completion of a written task during class to use as benchmark/comparison for assessment work (or several of these).

Further assessment opportunities within the same calendar year

The conditions of assessment state that “a further assessment opportunity needs to involve a new data set”. Therefore, two different assessments are needed for each achievement standard that a re-sit is planned to be offered.

The issue of recycling assessments and data sets in future years

Security and integrity of assessments between years can be an issue as students will discuss these with older siblings and friends. Any action taken here needs to be in line with school policies, whilst ensuring teacher workload is manageable – it is always a challenge tracking down appropriate data sets for assessments. Some suggestions include:

  • rotating assessments between first and second opportunities – for example, use the re-sit task from the previous year as the main task the current year. When further assessments are available, these can easily be added into the rotation cycle
  • rotating data sets/contexts between 91581 (3.9) and 91582 (3.10) each year
  • giving a different “hook” into a data set to get a different direction in the analysis. For example, with the income SURF ( one year could use “gender disparity” as a hook and another year could use “education”
  • taking a different (large) sample from a very large data set each year (for example, income SURF The super-SURF ( contains 100 different versions of a dataset with about 11,000 records each time
  • swapping assessments/data sets with colleagues from other schools, while making sure individual school policies are okay with this.

Other things to remember:

  • The same context can be used for 91035 (1.10), 91264 (2.9) and 91582 (3.10) within one year. This means that only one context has to be found and then the data can be adapted as needed for each level. Marking would have to be done carefully!
  • Level 2 conditions of assessment document for 91264 (2.9), 91265 (2.10) and 91268 (2.13) also state “Care must be taken to allow students opportunities to present their best evidence against the standard that is free from unnecessary constraints. Collection of evidence for these standards could include, but is not restricted to, an extended task, an investigation, or a more formal activity.” so there is a need to consider some of the above points at level 2 as well.