KASNEB DICT common mistakes
KASNEB DICT · Understand the traps and how to avoid them.
When asked how to compute something in a spreadsheet, the candidate writes a sentence such as "use VLOOKUP to get the price from the price list" instead of writing the actual entry, e.g. =VLOOKUP(A2,$D$2:$F$50,3,FALSE), and does not state the cell references, absolute/relative referencing or argument order.
Candidates revise office-application topics by reading notes and recognising functions, but almost never type the syntax under exam conditions, so they recall the name of the function and not its structure.
What to do instead
Drill each of the top functions (VLOOKUP, HLOOKUP, SUMIF, COUNTIF, IF, PMT, AVERAGE) by writing the full formula with sample references, then labelling each argument. Practise stating where absolute ($) references are required and why.
Gives the function of the wrong or a neighbouring component - describing cache as "the main memory", calling ROM "the working memory", or explaining a register as permanent storage - in questions that ask for the function of a named piece of hardware.
Hardware is memorised as a list of names and sizes rather than as a hierarchy of speed, cost, volatility and location, so lookalike components collapse into one vague idea.
What to do instead
Build a comparison table for every component group (register / cache / RAM / ROM / secondary storage; HDMI / VGA / USB; hub / switch / router) with columns for function, volatility, relative speed, typical capacity and one everyday analogy, and answer each definition question by giving function first, then the distinguishing property.
In a troubleshooting scenario (e.g. the machine powers on but shows no display, or emits repeated beeps), the candidate dumps every possible cause in a single list - "faulty RAM, faulty motherboard, virus, bad power supply, corrupted Windows" - instead of working through a diagnostic order and saying what the first check would prove.
Revision guides present fault lists as flat bullet points, and candidates reproduce the list rather than the logic that connects symptom to test.
What to do instead
Practise fault-finding as a sequence: note the symptom, state the most likely and cheapest test first, state the result that would confirm or eliminate each cause, and only then move to the next. Anchor to standard routines such as POST/beep codes, power and cabling checks, then component substitution, and always justify why the order is sensible (cost, risk, likelihood).
Mislabels the OSI layer or address type of a device or protocol - placing a switch at the network layer, saying a hub uses IP addresses, or describing a router as a device that "connects computers in a LAN" - and gives subnetting answers as a bare address with no binary working.
The layer model is memorised as seven names with no link to the addressing each device actually reads, and subnetting is copied from worked examples without learning to convert to binary.
What to do instead
For every device and protocol answer in two parts: the address or identifier it uses (MAC, IP, port, none) and therefore the layer. For subnetting and masking, always write the conversion to binary and back, show the network address, broadcast address and usable host range, and state the number of usable hosts.
Answers "write an algorithm / pseudocode" questions with a paragraph of connected English prose ("first of all the program will ask the user for the marks and then it will calculate...") with no START/END, no variable declarations, no indentation for selection or loops, and no explicit input, process and output sections.
Candidates can explain the logic in speech but have never been marked on structure, so they write the explanation they would give verbally.
What to do instead
Adopt one fixed pseudocode convention and use it all the time: labelled INPUT / PROCESS / OUTPUT or a numbered algorithm with START and END, named variables declared up front, indented IF-THEN-ELSE and WHILE/REPEAT blocks, and explicit counters and totals initialised to zero. Convert the same problem to pseudocode and to a flowchart until the two shapes match.
States normalisation rules as slogans - "2NF removes partial dependency, 3NF removes transitive dependency" - but then cannot locate the partial or transitive dependency in a relation given in the question, and draws a schema or table without marking primary and foreign keys.
The definitions are short enough to memorise verbatim, so candidates skip the practice of writing out functional dependencies and rebuilding the tables.
What to do instead
For every normalisation question, first list all functional dependencies from the sample data, then apply them in order to one table at a time, showing the before and after tables, and underline the primary key and label each foreign key with an arrow to the table it references. Practise with realistic un-normalised records containing repeating groups.
Name-drops a methodology or lists the SDLC phases as a memorised sequence that is not connected to the case study given (a hospital, a school, a saccos), producing a generic answer that would fit any organisation and ignoring the specific problem, actors and constraints in the scenario.
Candidates treat the scenario as decoration and the phase list as the answer, because past revision has rewarded recall of the list rather than application.
What to do instead
Before writing, underline the problem statement, the users and the constraints in the scenario, then tie each phase or model you mention to a concrete artefact from that scenario ("requirements gathered from the records clerk via interviews → output: a requirements specification listing the fee-balance query"). If a method is named, state why it suits this scenario's risk and requirement stability.
Answers "state two controls" with two controls from the same category (two password rules, or "lock the door" and "lock the server room"), and uses threat, vulnerability, risk and control interchangeably in definitions.
Control lists are memorised as an undifferentiated set of security words, and the four terms are learned as synonyms rather than as links in one chain.
What to do instead
Force breadth by answering in three labelled groups - physical, technical/logical, administrative - and pick one from each where the question allows. For definitions, always use the chain: a threat exploits a vulnerability, creating a risk, which is reduced by a control with a stated residual risk. Cite a specific example for each term.
Writes HTML fragments without the document skeleton (no <!DOCTYPE html>, <html>, <head>, <title> or <body>), uses deprecated presentational tags such as <font> and bgcolor instead of CSS, and answers "state the tag/attribute" questions with a description rather than the actual tag or attribute name.
Candidates practise isolated tags in a browser that tolerates broken markup, so the required structure and the separation of content from presentation are never learned.
What to do instead
Memorise and always type the five-line skeleton first, then insert content. Learn the current replacement for each deprecated attribute (style/CSS properties, for semantically, for ), and rehearse tag-recognition by writing the exact tag, its attributes and a one-line purpose for a fixed set of about thirty elements.
Answers ethics, entrepreneurship and project-management questions with personal moral opinion or everyday business vocabulary - "it is wrong to steal data", "a business plan is the same as a feasibility study" - instead of citing the specific professional code principle, legal provision, or the formal document's defined sections.
These topics feel like general knowledge, so candidates answer from intuition and skip the named codes, statutes and standard document structures that carry the marks.
What to do instead
Build a one-page reference of the named instruments and structures you must quote: the relevant data protection and computer-misuse provisions, the ICT professional body's code principles (public interest, competence, confidentiality, integrity), the distinction between a feasibility study (viability) and a business plan (implementation and financing), and the standard contents of a project charter, work breakdown structure and risk register. Answer by naming the instrument, then applying it to the facts given.