HKACE ICT Mock Examination 2026 Report

Detailed marking and performance analysis for Paper 1A, Paper 1B, Paper 2A and Paper 2B

Candidate Name

To Chun Hei

Candidate Number

L2260008E

Papers Marked

Paper 1A, Paper 1B, Paper 2A, Paper 2B

Marking Basis

Official keys/schemes uploaded + HKDSE ICT best-fit professional judgement

Overall Result Snapshot

Raw total across attempted papers

136 / 160

Overall raw percentage: 85.0%

This is a strong overall performance. Paper 1 is particularly strong, while the main mark loss comes from precision in databases, a few web-programming details, and several concept-specific MC items.

Optional weighted estimate: if the same ICT weighting of Paper 1 = 65% and Paper 2 = 35% is used, the estimated final weighted score is 85.40 / 100.

Paper 1A

35 / 40

87.5%

Paper 1B

56 / 60

93.3%

Paper 2A

22 / 30

73.3%

Paper 2B

23 / 30

76.7%

Standardised paper totals

Paper 1 total: 35 + 56 = 91 / 100
Paper 2 total: 22 + 23 = 45 / 60 = 75 / 100

Important marker note

P1B Q8(c)(ii) appears inconsistent in the official scheme; it was marked by actual SQL meaning.
Some coding items were marked using HKDSE ICT knowledge first where the scheme was incomplete or overly narrow.
Because of this, another strict marker may differ by about 1 mark on a few border cases, but the overall profile will remain the same.

Part 1 — Marking

All questions are marked using the uploaded official keys/marking schemes, with best-fit HKDSE ICT judgement where the scheme is incomplete, inaccurate, or does not clearly cover the answer given.

Paper 1A — Multiple Choice: 35 / 40

Q	Area	Your Ans	Key	Mark	Reason
1	Input / data capture	C	C	1/1	Correct. Having another student enter the same data is double data entry.
2	File format / data type	B	B	1/1	Correct. SVG is an image format.
3	Data types	C	C	1/1	Correct. Product ID and supplier are strings; stock quantity is integer; unit price is float.
4	Parity	A	A	1/1	Correct according to the parity pattern in the options.
5	Image representation	D	D	1/1	Correct. With 5 bits each for R, G and B, the palette supports 32768 colours.
6	Hex colour pattern	A	D	0/1	The pure-colour values are 03E0 and 7C00, so the correct answer is (2) and (3) only.
7	Analog / digital conversion	C	C	1/1	Correct. Scanning a photo and playing audio through headphones both involve analog-digital conversion.
8	Character set	A	A	1/1	Correct. UTF-8 is needed to store English, Arabic and Korean correctly.
9	Privacy / primary key	C	C	1/1	Correct. HKID is not a suitable primary key for a company’s customer purchase history due to privacy concerns.
10	Spreadsheet	C	C	1/1	Correct. The mixed references in option C are suitable for copying the `SUMIF` formula down.
11	SQL	A	A	1/1	Correct. `WHERE height > 160 ORDER BY height ASC` matches the requirement.
12	Memory concepts	A	C	0/1	Statement (1) is true, statement (2) is also true, and statement (3) is false. So the correct answer is (1) and (2) only.
13	RAM performance	C	B	0/1	The best explanation is that software may not utilize additional RAM.
14	Hardware specifications	D	D	1/1	Correct. Wireless standard, battery capacity and storage size can all be listed as VR headset specifications.
15	CPU architecture	A	A	1/1	Correct. Word length affects the range and complexity of operations a CPU can handle.
16	Memory hierarchy	D	D	1/1	Correct. CPU cache has the highest transfer rate and the smallest capacity.
17	IP addressing	C	C	1/1	Correct. IPv6 solves IPv4 exhaustion, and both IPv4 and IPv6 identify devices on a network.
18	Transmission media	B	B	1/1	Correct. Fibre optic cables have low signal loss and high bandwidth.
19	HTTPS	A	A	1/1	Correct. HTTPS is used to encrypt transmitted data.
20	Networking devices	B	B	1/1	Correct. A switch forwards data packets within a LAN.
21	Network performance	C	D	0/1	According to the official key, the best answer is (2) and (3) only: local Wi-Fi congestion and interference from nearby networks.
22	Certificate error	B	B	1/1	Correct. Expired/invalid certificates and unverifiable identity both cause browser warnings.
23	Address bit length	D	D	1/1	Correct. Bit lengths are IPv6 > MAC > IPv4.
24	DNS cache	A	A	1/1	Correct. DNS cache stores domain-to-IP translations locally.
25	Security	C	C	1/1	Correct. Phishing tricks users into revealing sensitive information.
26	Algorithm tracing	C	C	1/1	Correct. Repeated doubling from 1 over 5 iterations gives 32.
27	Algorithm tracing	A	A	1/1	Correct. The outputs are 1 2 1 4.
28	Flowchart tracing	B	B	1/1	Correct. The values less than or equal to 7 are added, giving 18.
29	Flowchart purpose	C	C	1/1	Correct. The program sums array values that are less than or equal to 7.
30	REPEAT...UNTIL	B	B	1/1	Correct. `OUTPUT B` executes 6 times before `B > 15`.
31	While condition	A	A	1/1	Correct. The condition should be `N > 0`.
32	Loop logic	D	D	1/1	Correct. Since `i` is never updated, the result cannot be found.
33	Array update	D	D	1/1	Correct. The final value is 3600.
34	Boundary testing	A	A	1/1	Correct. The boundary cases are 8 and 9 around the condition `A > 8`.
35	Boolean logic	D	B	0/1	Expression (3) is equivalent to (1) by De Morgan’s law, but expression (2) is not. So the correct answer is (1) and (3) only.
36	Array / accumulation	D	D	1/1	Correct. The missing statement is `Total ← Total + A[i]`.
37	Testing values	B	B	1/1	Correct. -1, 0, 5 cover negative, zero and positive cases.
38	Boolean data type	B	B	1/1	Correct. Inventory availability status is naturally Boolean.
39	Digital divide	D	D	1/1	Correct. Free IT lessons with equipment promote equitable access.
40	Ethics / copyright	A	A	1/1	Correct. Generated images may still create copyright risks if their rights are not checked properly.

Paper 1B — Conventional Questions: 56 / 60

Question	Mark	Reason
Q1(a)	1 / 1	85 is within the accepted range 75–100.
Q1(b)	3 / 3	Your formula is essentially correct: `=IF(AND(C2>=75,OR(D2="A",E2="Y")),"Y","N")`.
Q2(a)	1 / 1	Accepted. You stated that the GPU improves rendering / graphics processing speed.
Q2(b) HDD	1 / 1	Correct. Moving parts make HDDs more vulnerable to shock / physical damage.
Q2(b) SSD	1 / 1	Correct. SSD has a higher cost per unit storage.
Q2(c)	1 / 1	Accepted. External storage is reasonable because it avoids Internet upload delay.
Q3(a)	2 / 2	Both examples are acceptable: larger font size and voice navigation.
Q3(b)(i)	1 / 1	Correct. You described an impersonation / phishing-style government scam by SMS.
Q3(b)(ii)	1 / 1	Accepted. Community education / talks are a valid way to reduce scam victims.
Q4(a)	2 / 2	Correct. You clearly stated that HTTPS uses encryption (SSL/TLS) and is more secure than HTTP.
Q4(b)	1 / 2	You mentioned Peter’s public/private key pair, but the explanation mixed it up with the user’s own keys. The essential model is: user encrypts with Peter’s public key; Peter’s site decrypts with Peter’s private key.
Q4(c)	1 / 1	Accepted. Open-source software can be modified and redistributed subject to licence conditions.
Q5(a)	2 / 2	Correct. Only 1 bit is required because there are only two states: occupied / unoccupied.
Q5(b)	2 / 2	Correct. Real-time processing is needed because reservation availability must be updated immediately.
Q5(c)	1 / 1	Resource monitoring / system monitoring is an acceptable utility program answer.
Q6 Image A	0 / 2	BMP is not the most suitable choice for a high-resolution web photograph. The expected better answers are JPG or PNG, depending on whether compression or lossless quality is emphasised.
Q6 Image B	2 / 2	Correct. SVG is ideal for a geometric logo because it scales without quality loss.
Q7(a)	1 / 1	Correct. Scanning printed books into e-books is valid digitisation.
Q7(b)(i)	3 / 3	Good packet explanation: segmentation, addressing/routing, and reassembly were all present.
Q7(b)(ii)	2 / 2	Accepted. Authentication / access control and firewall are reasonable security measures.
Q7(c)(i)	1 / 1	Correct. Access points provide wireless connection to the network.
Q7(c)(ii)	2 / 2	Accepted. You referred to interference and physical placement / wall effects, which are both relevant deployment factors.
Q7(d)(i)	1 / 1	Correct. Facial recognition is a valid biometric authentication technology.
Q7(d)(ii)	1 / 1	Accepted. Your answer points to privacy / data leakage risks involving biometric data.
Q8(a)	1 / 1	Correct. A range check on `CQ` is valid.
Q8(b)	2 / 2	Correct. `VM` should be character/text because it needs to preserve leading zeros.
Q8(c)(i)	1 / 1	Your listed records are correct and in acceptable order.
Q8(c)(ii)	1 / 1	Marked by actual SQL logic rather than the inconsistent scheme text. Your grouped result is semantically correct for `AVG(CQ)`.
Q8(d)	4 / 4	Good flowchart. The refill condition, loop structure, output and refill amount `A = MC[i] - CQ[i]` are correct.
Q8(e)	3 / 3	All three test data choices are suitable: normal refill, boundary value and invalid input.
Q9(a)	2 / 2	Correct. Arrays are easier to iterate through, and a single variable stores only one value at a time.
Q9(b)(i)	1 / 1	Correct. This is a logic error.
Q9(b)(ii)	1 / 1	Correct. The fix is `average = total / count`.
Q9(c)	3 / 4	Your algorithmic idea is correct, but the Python notation / identifier use is not fully clean. Logic is strong; syntax precision is the only issue.
Q9(d)	3 / 3	Correct logic: initialize, loop through the list, test `> 30`, and accumulate the count.
Q9(e)	1 / 1	Accepted. Easier maintenance / reuse is a valid modularization advantage.

Paper 2A — Databases: 22 / 30

Question	Mark	Reason
Q1(a)	3 / 4	Your ER diagram captured the two relationships and the basic 1-to-many structure, but the optional / mandatory participation constraints were not fully shown.
Q1(b)	2 / 2	Correct. You explained rollback and restoring the database to a previous normal state.
Q2(a)	1 / 1	Correct. `UNION ALL` lists 8 records here.
Q2(b)(i)	1 / 2	`LEFT JOIN` is correct, but `IS NULL` was needed rather than `= 'C03'`.
Q2(b)(ii)	1 / 1	You gave the correct intended output: C03, F01.
Q3(a)	1 / 1	Correct. Scores should be stored as integer because they are numeric and used for calculation such as averages.
Q3(b)(i)	2 / 2	Correct. You identified that the query finds the S2 class with the highest average mark in MATH101.
Q3(b)(ii)	0 / 1	Your “faster retrieval speed” point is not a reliable advantage of a normal SQL view. Better answers are simplification of complex SQL or access restriction.
Q4 FLIGHT relation	1 / 1	`FLIGHT(FID, FROM, TO, FTIME)` is correct.
Q4 FLIGHT foreign keys	1 / 1	`FROM` and `TO` both correctly reference `LID` of `LOC`.
Q4 ROSTER fields	0 / 1	The `ROSTER` relation should include `FID`; your filled fields do not match the normalized design fully.
Q4 primary key	0 / 1	The expected primary key is the composite key RID + FID, not RID alone.
Q4 foreign key references	1 / 1	Your foreign key references for `FID` and `PID` are correct.
Q5(a)	2 / 2	Your SQL for junior members is correct.
Q5(b)	1 / 2	You selected the right field and conditions, but you missed the join condition `BOOK.ISBN = LOAN.ISBN`.
Q5(c)	1 / 3	You showed the correct direction with filtering by member and using a subquery, but the joins, subquery condition, and grouping were incomplete / inaccurate.
Q5(d)	1 / 1	Correct. The updated date violates domain integrity because the date value is invalid.
Q5(e)(i)	1 / 1	Correct. Combining tables / adding redundancy is a valid denormalization description.
Q5(e)(ii)	2 / 2	Both points are correct: faster / simpler querying as an advantage, and redundancy / inconsistency / anomalies as a disadvantage.

Paper 2B — Web Application Development: 23 / 30

Question	Mark	Reason
Q6(a)	3 / 3	Your sketch included the key input elements: date selection, exhibition type input, ticket category choice and ticket quantity control.
Q6(b)	2 / 2	Accepted. Technical support is a valid advantage; dependency on provider availability is a reasonable disadvantage.
Q7(a)	1 / 1	Correct. Broad write access may allow staff to alter or delete sensitive files accidentally.
Q7(b)(i)	1 / 1	Correct. `ReportCard.pdf` becomes read-only to those users.
Q7(b)(ii)	1 / 1	Correct. Class-based group permissions are the right approach.
Q7(c)(i)	1 / 1	Accepted. Better central management / tighter local control is a valid intranet advantage.
Q7(c)(ii)	1 / 1	Correct. Cloud storage is accessible anywhere with Internet access.
Q8(a)(i)	1 / 1	Correct. DHCP corresponds to function 5.
Q8(a)(ii)	1 / 1	Correct. Domain controller corresponds to function 1.
Q8(b)(i)	0 / 1	B1 should be the proxy server (Function 2), not router (3).
Q8(b)(ii)	1 / 1	Correct. B2 is the firewall (Function 4).
Q8(b)(iii)	0 / 1	B3 should be the wireless router / access point (Function 3), not proxy server (2).
Q9(a)(i)	1 / 1	Correct. The background colour is #f4f4f4.
Q9(a)(ii)	1 / 1	Correct. The font size of “You are here” becomes 150% on hover.
Q9(a)(iii)	1 / 1	Correct. `onkeypress` detects a key press.
Q9(b)	1 / 2	You correctly identified that the image path is wrong, but the corrected path you wrote is still not the best-fit expected answer.
Q9(c)(i)	0 / 1	`$_COOKIE[...]` reads a cookie; it does not create one. A browser-side cookie should be set by `document.cookie` or a proper `Set-Cookie` header.
Q9(c)(ii)	1 / 1	Correct. `$_COOKIE['currentFloor']` is the expected replacement.
Q9(d)	0 / 2	The correct validation is `!in_array($shop_type, $accepted)`. Your condition does not test membership in the accepted list properly.
Q9(e)	2 / 2	Your count query is correct in structure and intent.
Q9(f)	2 / 3	Your decision logic is correct, but you used `alert()` instead of updating the page with `document.getElementById("guide").innerText`.
Q9(g)	1 / 1	Correct. Freeing the result set releases memory and helps performance.

Marking judgement: This script is strong overall. Your best papers are Paper 1B and Paper 1A. Your weakest area is not general understanding, but precision in database structure / SQL and a few exact web-programming details.

Part 2 — Summary

Performance analysis for MC, each Paper 1 long question, and each Paper 2 question.

MC performance

Paper 1A: 35 / 40

This is a strong MC result.
You were solid in data representation, networking, spreadsheet, security, and algorithm tracing.
Only 5 questions were wrong, so your core knowledge base is clearly good.
The losses were mainly in exact concept discrimination, not in basic understanding.

Strong theory base Good tracing skill A few concept traps

Overall paper profile

Raw profile

Paper 1 total: 91 / 100
Paper 2 total: 45 / 60 = 75%
Your Paper 1 is clearly stronger than your Paper 2.
Paper 2 is still good, but database modelling and some web coding precision stopped it from becoming outstanding.

Very strong Paper 1 Solid Paper 2 Precision issues

Paper 1B — Question-by-question summary

Spreadsheet logic

Excellent. You understood both the threshold condition and the nested logical formula accurately.

Multimedia hardware / storage

Excellent. Very practical and safe answers on GPU use, HDD/SSD comparison and storage choice.

UX for elderly / scams

Excellent. You gave sensible accessibility features and a realistic anti-scam response.

Web security / PKI / copyright

Good overall, but your PKI explanation was not clean enough. You know the security idea, but key ownership and encryption flow must be stated more precisely.

Parking system / processing

Strong. Good basic systems thinking. You correctly linked the reservation problem to real-time processing.

Image formats

This was the main weak spot in Paper 1B. You chose BMP for a web photograph, which is technically valid as an image file but not the most suitable format for website use.

Library system / networking / security

Excellent. Your packet transmission explanation was good, and your understanding of access points and deployment factors was strong.

Database / SQL / flowchart

Excellent overall. You handled the SQL output, data type choice, flowchart, and testing very well. This is one of your strongest long questions.

Programming

Good. Your algorithmic thinking was sound, but code notation / syntax precision in Python was not completely clean.

Paper 2A — Databases summary by question

ER diagram + transaction handling

You understood the main entity relationships and rollback, but you did not fully express the participation constraints in the ER diagram.

UNION / JOIN / NULL logic

The counting part was fine. The main weakness was the exact SQL condition for records not existing in another table.

Data type + view purpose

Good. You understood why SCORE should be numeric and the overall aim of the view, but your advantage of using a view was not conceptually strong.

Normalization / keys

Mixed. You got the FLIGHT relation right and understood foreign-key references, but the ROSTER structure and composite primary key were not handled accurately.

SQL writing + denormalization

Your simpler SQL and denormalization explanations were good, but your multi-table SQL with filtering and grouping was weaker.

Paper 2B — Web Application Development summary by question

UI design + hosting

Strong. Your layout was practical and you understood the trade-off of commercial hosting.

Permissions / intranet / cloud

Strong. You showed a good applied understanding of file permissions and sharing methods.

Network devices

Mixed. You understood DHCP, domain controller and firewall, but confused the proxy server and wireless router / access point roles.

HTML/CSS/PHP/JavaScript/cookies

Moderate to good. You understood most of the front-end basics, but exact coding details such as cookie creation, validation logic, path correction and DOM updating lost marks.

Overall pattern: You are strongest when the task is practical, structured and familiar. You lose more marks when the question depends on exact formalism: composite keys, participation constraints, SQL join details, PKI flow, or exact code statements.

Part 3 — Mistakes

All wrong or partially wrong questions across all papers, with analysis of why marks were lost.

Paper 1A mistakes

Q6 — Pure colour hex pattern: This is a careful bit-pattern interpretation question. You likely identified one pure colour correctly but missed that both 03E0 and 7C00 represent single-channel pure colours.
Q12 — Memory components and performance: You underestimated statement (2). A wider address bus does allow access to a larger physical memory space.
Q13 — Why more RAM does not always help: You focused on CPU limitation, but the better answer is that software may not use the additional RAM.
Q21 — Slow Internet factors: This was a “best answer” judgement item. According to the official key, the issue is most directly tied to local Wi-Fi congestion and wireless interference.
Q35 — Boolean equivalence: This is a logic-manipulation mistake. Expression (3) is actually equivalent to (1), while (2) is not.

Paper 1B mistakes / partial losses

Q4(b) — PKI explanation: You know that public/private keys are involved, but your wording mixed up whose key is used. This is a precision of security terminology problem.
Q6 Image A: The main issue is suitability for web use, not just image quality. BMP preserves data, but it is not the best practical web choice for a campus photograph.
Q9(c): Your max-temperature logic is good, but the code is not fully “exam-safe” in notation / syntax. This is not a logic failure; it is a coding-accuracy issue.

Paper 2A mistakes / partial losses

Q1(a) — ER participation constraints: Your entities and cardinalities were mostly fine, but database modelling requires not only “how many”, but also whether participation is optional or mandatory.
Q2(b)(i) — SQL for non-matching records: You correctly chose LEFT JOIN, but the final condition must test the missing record with IS NULL. This is a classic SQL anti-join pattern.
Q3(b)(ii) — Advantage of a view: “Faster retrieval” is not a safe general answer for ordinary SQL views. You should prefer points like simplifying complex queries or restricting user access.
Q4 — 3NF schema design: The biggest problem was the ROSTER design. You missed the need for FID in the relation and the composite primary key RID + FID.
Q5(b) — SQL join: You omitted the join between BOOK and LOAN, so the query would not work correctly.
Q5(c) — Multi-table SQL: This was the hardest database item for you. The intended joins, subquery, exclusion logic and grouping were not all brought together correctly.

Paper 2B mistakes / partial losses

Q8(b)(i) and Q8(b)(iii) — Device roles: You confused proxy server with wireless router / access point. This shows a role-mapping weakness in network architecture questions.
Q9(b) — CSS image path: You knew the path was wrong, but your correction was still not fully safe.
Q9(c)(i) — Cookie creation: You wrote a cookie-reading form instead of a cookie-setting statement. This is a server-side vs client-side distinction issue.
Q9(d) — Fixed-value validation: You did not use list-membership checking. The expected logic is !in_array($shop_type, $accepted).
Q9(f) — JavaScript output: Your decisions were right, but the question wanted the result displayed in the page rather than through alert().

Main mistake pattern: Most of your mark loss is from exact wording, exact SQL structure, exact code expression, or exact modelling detail. That is good news, because it means your underlying understanding is already quite solid.

Part 4 — Improvement

Weakness analysis, strengths, and personal feedback on how to improve further.

Your main weaknesses

Database precision: anti-join logic, composite keys, schema design, and the exact purpose / advantage of views.
Exact security explanation: especially PKI and certificate-related concepts.
Exact coding detail: cookies, validation functions, DOM output methods, and “what statement should be used here” type questions.
Device-role mapping: proxy server vs router vs access point in network diagrams.
Best-format judgement: choosing the most suitable file format for a specific real-world context.

Your main strengths

Very strong Paper 1 foundation: 35/40 in MC and excellent performance in most long questions.
Good practical sense: your answers are usually sensible and grounded in real usage.
Strong algorithmic thinking: tracing and flowchart work are generally reliable.
Strong applied networking / web understanding: especially in permissions, UI layout and general web ideas.
Good potential for rapid improvement: because most errors are fixable precision issues, not deep misunderstandings.

Personal feedback

This is a strong mock paper overall. You are not struggling with ICT as a subject; you already have a good working grasp of many areas.
Your profile suggests that you are a student who usually understands what the system is doing, but sometimes loses marks when the exam wants the formal textbook-safe expression.
That means your next step is not “learn everything again from zero”. Your next step is tightening precision.
If you improve database modelling / SQL exactness and clean up your web-code precision, your Paper 2 score can rise noticeably without needing a huge amount of new content learning.
In short: your knowledge base is already good. Your biggest improvement area is exam-safe expression.

Priority 1

Databases

Redo ER diagram questions with optional / mandatory participation.
Practise LEFT JOIN ... IS NULL, composite keys, and multi-table SQL.
Write why a view is used in one sentence, using textbook-safe wording.

Priority 2

Web coding precision

Practise cookies: create vs read vs send from server.
Revise PHP validation tools such as in_array().
Revise DOM output methods like innerText and innerHTML.

Priority 3

Exact terminology

Review PKI using one fixed model: server public key encrypts, server private key decrypts.
Memorise the role of proxy server, firewall, router and access point distinctly.
Practise “best format / best device / best answer” questions.

Suggested short revision plan

Focus	What to do next
Database	Redo 10 SQL questions and 5 ER / normalization questions. For each one, write not only the answer, but also the exact reason why the structure is correct.
Web coding	Practise 5 short PHP / JavaScript tasks involving cookies, form validation and DOM updates. Aim for exact syntax, not only approximate logic.
Security concepts	Write a one-page summary of PKI, HTTPS, certificates, phishing and authentication / authorization.
Device roles	Make a small comparison sheet for switch, router, firewall, proxy server, DHCP server, domain controller and access point.
MC refinement	Review your 5 wrong MC items and rewrite the correct reasoning in one sentence each.

Final comment: This mock suggests that your ceiling is high. You already have the ability to score very well. The next improvement will come from precision, not reinvention.