9.3 Interpreting NAPLAN Displays, Proficiency Levels, Percentiles, and Standardised Scores

Australian teachers regularly interpret national and statewide assessment reports—most notably the National Assessment Program — Literacy and Numeracy (NAPLAN)—alongside internal school diagnostics. Reading these displays accurately requires distinguishing criterion-referenced raw marks from norm-referenced percentiles, cohort growth charts, and standardised scale scores.

---

Percentiles, Deciles, and Quartiles vs. Percentage Correct

A frequent misconception in parent-teacher interviews and LANTITE questions is confusing percentage correct with a percentile rank:

• Percentage Correct (Raw / Criterion Measure): Calculated as $\frac{\text{Marks Achieved}}{\text{Total Possible Marks}} \times 100\%$. It measures what proportion of the test items a student answered correctly, regardless of how other students performed.
• Percentile Rank (Norm-Referenced Position): Indicates the percentage of students in the reference cohort whose scores are at or below a given student's score. A student at the $80\text{th}$ percentile performed as well as or better than $80\%$ of the cohort and sits in the top $20\%$ ($100\% - 80\% = 20\%$). On a challenging competition paper, a raw mark of $54\%$ correct might place a student at the $85\text{th}$ percentile; on a foundational mastery check, $80\%$ correct might sit at the $30\text{th}$ percentile.
• Deciles and Quartiles: Deciles divide an ordered cohort into $10$ equal bands of $10\%$ each (the $1\text{st}$ decile cutoff is the $10\text{th}$ percentile; the $9\text{th}$ decile cutoff is the $90\text{th}$ percentile). Quartiles correspond to the $25\text{th}$ percentile ($Q_1$), $50\text{th}$ percentile ($Q_2$, median), and $75\text{th}$ percentile ($Q_3$).

---

Reading Australian NAPLAN Displays and Proficiency Levels

1. The Four NAPLAN Proficiency Levels

Since 2023, NAPLAN individual and school summary reports classify student scale scores into four proficiency levels set against challenging national expectations for each Year level (Years 3, 5, 7, and 9):

1. Exceeding: The student's result exceeds expectations at the time of testing.
2. Strong: The student's result meets challenging but reasonable expectations at the time of testing.
3. Developing: The student's result indicates that they are working towards expectations at the time of testing.
4. Needs additional support: The student's result indicates that they did not achieve the learning outcomes expected at the time of testing and are likely to need additional support to make satisfactory progress.

When analysing stacked bar charts of proficiency levels, watch the exact wording of the question: "meeting or exceeding expectations" combines Strong + Exceeding, whereas "below the Strong benchmark" combines Developing + Needs additional support.

2. The 20th–80th Percentile School Summary Display and ICSEA Benchmarks

On NAPLAN and My School graphical summaries (as well as legacy 10-band growth charts), a school's score distribution for a Year level is commonly represented by a vertical or horizontal bar extending from the $20\text{th}$ percentile at the bottom/left to the $80\text{th}$ percentile at the top/right:

• Below the $20\text{th}$ percentile mark: The lowest $20\%$ of students in that cohort.
• Inside the bar ($20\text{th}$ to $80\text{th}$ percentile): The middle $60\%$ of students ($80\% - 20\% = 60\%$). Note how this differs from a box plot, where the box spans $Q_1$ ($25\text{th}$) to $Q_3$ ($75\text{th}$) and covers the middle $50\%$.
• Above the $80\text{th}$ percentile mark: The highest $20\%$ of students in that cohort.
• Benchmarking markers: A central dot or line shows the school mean (or median) scale score, compared against two external benchmarks: the Australian national average and the average for schools with similar students (matched via the Index of Community Socio-Educational Advantage [ICSEA], which is scaled to a median of $1{,}000$ and a standard deviation of $100$).

3. Longitudinal Cohort Growth vs. Cross-Sectional Comparisons

NAPLAN assesses students in Years 3, 5, 7, and 9 (every two years). Educators must distinguish two different types of multi-year comparisons:

• Longitudinal Cohort Growth (Same Students): Tracks the same cohort as they progress two calendar years and two Year levels—for example, comparing Year 3 in 2023 with Year 5 in 2025, or Year 7 in 2023 with Year 9 in 2025.
• Cross-Sectional Comparison (Different Students): Compares the same Year level across consecutive calendar years—for example, Year 5 in 2024 vs. Year 5 in 2025. Because these are two completely different groups of children, the difference does not measure individual two-year learning gain.

---

Standardised Scale Scores and z-Scores in Educational Reporting

Raw test marks cannot be compared directly across different subjects when the assessments have different means and spreads. To place any score $x$ on a common relative scale, calculate its standardised score ($z$-score):

$$z = \frac{x - \text{Mean}}{\text{SD}}$$

Here, $\text{SD}$ is the standard deviation, which measures the typical distance of scores from the cohort mean. A positive $z$-score means the student scored above the mean, a $z$-score of $0$ equals the mean, and a negative $z$-score lies below the mean.

Worked Example: Comparing Relative Standing Across Subjects

Suppose a Year 11 student achieves:

• $78$ in Science, where the cohort mean is $66$ and $\text{SD} = 6$: $z_{\text{Science}} = \frac{78 - 66}{6} = \frac{12}{6} = +2.0$.
• $82$ in History, where the cohort mean is $72$ and $\text{SD} = 10$: $z_{\text{History}} = \frac{82 - 72}{10} = \frac{10}{10} = +1.0$.

Although the raw score in History ($82$) is higher than in Science ($78$), the student performed relatively better in Science because $+2.0$ standard deviations above the cohort mean represents a much stronger relative standing than $+1.0$ standard deviation above the mean. Similarly, on a standardised educational scale with $\text{Mean} = 500$ and $\text{SD} = 100$, a $z$-score of $+1.5$ corresponds to a scale score of $x = 500 + (1.5 \times 100) = 650$.