HKDSE ICT common mistakes
HKDSE ICT · Understand the traps and how to avoid them.
When converting between binary, denary, and hexadecimal, candidates misalign place values or forget to group bits correctly, producing answers like '1101₂ = 11₁₀' or '1010 1111₂ = AF₁₆' written as 'A F' with a space.
Candidates memorise conversion steps but do not practise enough with mixed bases; they lose track of the 8-4-2-1 weights or the 4-bit grouping for hexadecimal, especially under time pressure.
What to do instead
Always write the place-value table above the digits before converting. For binary to hex, pad the binary number to a multiple of four bits from the right and group with spaces; for hex to binary, expand each hex digit to exactly four bits. Check by converting back.
When calculating the storage size of an image or sound file, candidates use the wrong formula, e.g., forgetting to divide by 8 for bytes, or omitting the number of channels or sampling resolution.
They memorise 'width × height × colour depth' but do not internalise that colour depth is in bits and file size is usually in bytes; for sound they forget that sample rate is per second and that stereo doubles the data.
What to do instead
Write the formula with units: image size in bits = width × height × colour depth; sound size in bits = sampling rate × sampling resolution × channels × duration. Then divide by 8 for bytes and by 1024 for KiB. Do a sanity check: a 100×100 24-bit image should be about 30 KB, not 240 KB.
When describing the fetch-decode-execute cycle, candidates say the control unit 'executes' instructions or omit the role of the program counter and instruction register.
They learn the cycle as a slogan without mapping each step to the specific registers and buses involved; they confuse the control unit's decoding role with the ALU's execution role.
What to do instead
Draw the cycle with registers: PC holds address → MAR → address bus → memory → data bus → MDR → CIR → control unit decodes → ALU executes if needed → PC incremented. Practise labelling each component's function in one sentence.
When explaining why a computer uses RAM and ROM, candidates state that RAM is 'permanent' or that ROM can be written to easily, or they confuse virtual memory with RAM.
Everyday language ('memory') blurs the distinction; candidates do not link volatility to purpose (RAM for running programs, ROM for bootstrap) and they mix up virtual memory as an extension of RAM.
What to do instead
Create a comparison table: volatility, writability, typical contents, and role in booting. Emphasise that virtual memory is disk space used as an extension of RAM, not actual RAM. Test yourself by explaining what happens to each when power is lost.
When answering questions on internet protocols, candidates mix up the roles of HTTP, HTTPS, FTP, SMTP, and POP3/IMAP, or claim that HTTPS encrypts the entire URL and content.
They memorise protocol names without associating them with specific tasks (web browsing, file transfer, sending mail, retrieving mail) and do not understand that HTTPS encrypts the HTTP message but the domain may be visible via SNI.
What to do instead
Make a protocol-to-task matching list and practise with scenarios: 'Which protocol is used when an email client downloads mail from a server?' (POP3 or IMAP) versus 'sends mail' (SMTP). For HTTPS, note it encrypts content and headers but not the destination IP or SNI.
When describing how a search engine ranks results, candidates give a generic answer like 'it finds the most relevant pages' without mentioning crawling, indexing, or ranking factors such as keywords, links, and freshness.
They use search engines daily but have never considered the underlying process; they confuse searching with the whole pipeline and cannot distinguish crawling from indexing.
What to do instead
Learn the three stages: crawling (discovering pages via links), indexing (parsing and storing terms), and ranking (scoring by relevance signals). For ranking, list at least three concrete factors (keyword match, page authority via links, freshness) and explain each briefly.
When writing pseudocode or a program, candidates use '=' for comparison in conditions or forget to initialise a variable before a loop, leading to incorrect output or infinite loops.
They transfer mathematical notation ('=' means equals) into programming and do not mentally trace the code with sample data; they also overlook that loops need a starting value and an update.
What to do instead
Adopt a convention: use '==' for comparison and '=' for assignment, and always write initialisation before a loop. Before submitting, trace the code with a small table of variable values for each iteration. If a loop has no obvious termination, add a counter and test.
When asked to identify the output of a nested loop, candidates trace only the outer loop or assume the inner loop runs the same number of times every outer iteration.
Nested loops require tracking two or more variables simultaneously; candidates lose track of the inner loop's reset behaviour and the effect of changing bounds.
What to do instead
Use a trace table with columns for each loop variable and any output. Write the inner loop's full cycle for the first two outer iterations, then look for the pattern. For example, for i from 1 to 3 and j from 1 to i, list (1,1), (2,1), (2,2), (3,1), (3,2), (3,3) before stating the output.
When discussing social implications, candidates give one-sided answers, e.g., listing only advantages of social media or only privacy risks of biometric data, without addressing counterarguments or mitigation measures.
Exam questions often ask 'Discuss' and candidates treat them as 'List'; they do not structure answers to show balanced evaluation and specific examples.
What to do instead
Use a three-part structure: point, evidence/example, evaluation. For a 'discuss' question, give at least one advantage and one disadvantage, each with a concrete example (e.g., social media enables global connection but also cyberbullying), and end with a mitigation or condition (e.g., privacy settings, regulation).
When explaining the impact of an emerging technology such as AI or IoT, candidates describe only consumer benefits (e.g., smart home convenience) and ignore issues like data security, job displacement, or digital divide.
They rely on marketing narratives and do not connect technological change to broader social structures; they lack a framework for analysing impacts across stakeholders.
What to do instead
For any technology, analyse impacts on at least three stakeholders (individuals, businesses, government/society) and across three dimensions: economic, social, and ethical/legal. For IoT, for instance, note convenience for users, efficiency for businesses, but also privacy risks and the need for security standards.