# Details of school banding and ranking

The school banding and ranking shown in this website are estimated by integration, processing and analysis of the HKDSE statistical information announced by more than 400 schools through their official channels (e.g. school websites, open day brochures, etc).

BigExam.hk has been collecting the data that can reflect the overall HKDSE performance of the schools to build a database of key features for estimating academic levels of the schools. The features include:

- percentage of students attaining minimum entrance requirement of local bachelor degree programmes, 3322+2 or above
- percentage of students attaining minimum entrance requirement of local sub-degree programmes, 22+222 or above
- percentages of subject entries at Level 5 or above, Level 4 or above, Level 3 or above, Level 2 or above respectively
- percentage of Level 5 or above, Level 4 or above, Level 3 or above, Level 2 or above in the 4 core subjects (Chinese, English, compulsory part of Mathematics, Liberal Studies)
- Number of Category A subjects with percentages of Level 5 or above, Level 4 or above, Level 3 or above, Level 2 or above higher than Hong Kong averages
- Number of Category A subjects with 100% pass rate
- percentage of students obtained JUPAS degree offers
- percentage of students obtained any JUPAS offers
- percentage of students obtained offers from the top 3 universities (HKU, CUHK, HKUST)
- percentage of students obtained degree offers from the eight UGC-funded universities
- percentage of students admitted to local degree programmes
- percentage of students admitted to local and non-local degree programmes
- percentage of students admitted to degree or sub-degree programmes
- percentage of S. 6 graduates pursuing further studies after sitting HKDSE (including universities and other tertiary institutions, but excluding S. 6 repeaters and employed school leavers

The following criteria are used when counting the number of subjects in above-mentioned items 5 & 6:

- The compulsory and extended parts of Mathematics are counted as two separate subjects, but Module 1 and Module 2 of the extended part are considered as one subject.
- ”Accounting Module” and “Business Management Module” of “Business, Accounting and Financial Studies” are counted as one subject.
- The 3 combinations of elective parts in “Combined Science” are counted as one subject.
- The 2 strands in “Technology and Living”: “Food Science and Technology” and “Fashion, Clothing and Textiles” are counted as one subject.

Some features that only reflect the HKDSE performance of a few individuals are discarded, e.g. percentage of subject entries at Level 5* or above, or number of students attaining Level 5**, etc.

So-called Banding is originally the 3 bands into which the primary 6 pupils are divided by the EDB according to the academic performance for Secondary School Places Allocation. Therefore, a band is assigned to a student, but not to a school. However, the places allocation system prioritizes the students from Band 1 to Band 3. This leads to the reality that most of the students of a school come from the same band. That is why the public usually assign banding to the secondary schools.

Under the school places allocation system, the parents should know the bands to which their children belong, and also the “bands” of the schools (i.e. the main band of which the school is assigned the students). The students will have higher chance of being sent to the favourite schools by choosing the schools that matches the banding of the students. However, all the schools basically do not disclose the bands to which their students belong. The parents are unable to know the school banding from official channels. As a result, the parents usually nose into the school banding information from different public channels, which may include school ethos, teacher qualifications, school facilities, parents’ comments in addition to academic performance. Getting to the heart of the matter, the bands defined by EDB are determined purely based on the results of the 3 internal examinations in P .5 and P .6, with regardless of school ethos, teacher qualifications, etc. Therefore, BigExam.hk believes that estimation of the school banding by their academic performance only would be more accurate.

The schools disclose their HKDSE statistical information in different manners. It is difficult to compare the information directly. BigExam.hk therefore uses technology of big data analysis in which sophisticated mathematics is involved. Ignoring mathematics, it is more important to understand how the boundaries between the bands are determined from the HKDSE statistical information. BigExam.hk adopts the HKDSE statistics announced by the HKEAA as a reference standard. The amber curve in the figure below shows cumulative curve of the average exam results of all HK day school candidates. This curve deviated very little in the past cohorts. It can be assumed as fixed. If all the students were randomly assigned to the schools, you would imagine that all the schools would share the same curve. In reality, the students are divided into 3 bands in places allocation. This results 3 kinds of curves for the HKDSE statistics of the HK schools.

According to data analysis by BigExam.hk, the 3 kinds of curves are shown in the figure as red, green and blue respectively. A Band 1 school has percentages above the HK average at all Levels (as the red curve). A Band 2 school has percentages below the HK average at the higher Levels, but above the HK average at the lower Levels (as the green curve). A Band 3 school has percentages below the HK average at all Levels (as the blue curve). From the characteristics of these 3 kinds of curves, the boundaries between the bands can be defined. The boundary between Bands 1 and 2 closes to the HK average at the higher Levels, whereas, the boundary between bands 2 and 3 closes to the HK average at the lower Levels. Similarly, BigExam.hk further divides each Band into top, middle and bottom, indicated by A, B and C respectively. On the other hand, each school only discloses part of the HKDSE statistical information. There must be a tolerance in the banding evaluated by data analysis. This leads to the situation that some of the schools are evaluated as in between two bands. For example, 2A ‒ 1C indicates that the band is between top of Band 2 and end of Band 1.

The school ranking is obtained from the same results of data analysis for the school banding. Similarly, tolerance exists in the school ranking. Therefore, some of the school ranks are expressed as a range. If a school discloses more details about the HKDSE statistical information, less tolerance will be in the analysis results, i.e. more accurate the school banding and ranking. BigExam.hk has used the wordings: “High accuracy”, “Median accuracy” and “Rough estimate” to indicate the accuracy of data analysis. School ranks are only given to the schools that have disclosed sufficient HKDSE statistical information in the past two years. If no information is available in the recent three or more years, the schools will not be ranked.

The academic performance of the students may vary from cohort to cohort. The school banding and ranking provided by BigExam.hk will be adjusted every year. For a few of the schools that do not disclose HKDSE statistical information, their estimated banding may involve subjective factors. The school banding and ranking provided by BigExam.hk are for reference only. BigExam.hk has no responsibility for the accuracy of the information and any loss or damages due to the information.