FIT5152 Chap.9 Heuristic Evaluation of Interfaces
Heuristic Evaluation of Interfaces
Week 9 turns from designing to evaluating. The seminar explains why evaluation runs throughout user-centred design, validating ideas early and implementations later, and separates methods by who does the evaluating: experts run cognitive walkthroughs and heuristic evaluations, while users take part in usability testing, field studies and thinking-aloud sessions.
Expert review is quick and cheap, but experts may misunderstand the domain, disagree with each other and bring biases, which is why the unit pairs it with testing on real users.
Heuristic evaluation is the centrepiece.
Jakob Nielsen’s ten heuristics, drawn from a factor analysis of 249 usability problems, give evaluators a shared checklist, and Nielsen’s figures show that one evaluator finds about 35% of problems while five find about 75%. The chapter follows the process from two individual passes to severity rating by frequency, impact and persistence, then to collating findings into one list and debriefing with the design team.
It closes by comparing Nielsen’s heuristics with Ben Shneiderman’s eight golden rules and by separating mistakes from slips. Submission 3 asks every member to evaluate the group’s prototype against all ten heuristics, with at least three violations rated for severity.
What this chapter covers
- 01
Why and when to evaluate
- 02
Expert reviews and cognitive walkthroughs
- 03
Nielsen’s ten usability heuristics
- 04
How many evaluators to use
- 05
Severity factors and the rating scale
- 06
Collating findings and debriefing
- 07
Shneiderman’s eight golden rules
Worked example · free
Rating a violation on a food-delivery checkout
- 1Heuristic 9, help users recognise, diagnose and recover from errors: the message is a code, gives no cause and suggests no fix.
- 1Frequency: it occurs whenever a code fails, which is common during promotions.
- 1Impact and persistence: the order can still be placed, but customers lose the discount and the cleared field forces retyping, so the problem keeps biting; severity 3.
- 1Recommendation: keep the typed code, explain the reason in plain language, such as an expired offer, and suggest what to do next.
Key terms
- Heuristic
- A rule of thumb: general design guidance that holds for most interfaces without being guaranteed in every case.
- Heuristic Evaluation
- An expert method in which a few evaluators inspect an interface against an agreed list of usability heuristics.
- Cognitive Walkthrough
- An expert method that steps through a scenario’s ideal path, asking four questions about goals, actions and progress at each step.
- Severity Rating
- A score for a usability problem based on its frequency, impact and persistence.
- Golden Rules
- Ben Shneiderman’s eight interface design principles, from striving for consistency to reducing short-term memory load.
Heuristic Evaluation of Interfaces FAQ
How many evaluators does a heuristic evaluation need?
The seminar reports Nielsen’s finding that one evaluator uncovers about 35% of problems and five uncover about 75%, and it treats the best number as a cost-benefit decision for each project.
What severity range should Submission 3 use?
The brief asks for a rating from 1 to 4, justified by frequency, impact and persistence. The seminar’s full scale also includes 0, which an evaluator uses when they do not accept that the issue is a problem.
How do Nielsen’s heuristics relate to Shneiderman’s golden rules?
They overlap on consistency, error prevention, feedback, user control and memory load, but the golden rules add ideas such as universal usability and dialogs that yield closure. Week 9 asks you to compare the two sets.
What is the difference between a mistake and a slip?
Following Norman, a mistake is choosing the wrong action, while a slip is intending the right action and executing it wrongly by accident, such as pressing Reply All instead of Reply.
Assessment move
Memorise Nielsen’s ten heuristics in order, then attach one example of your own to each, half compliances and half violations, because Submission 3 needs one unique instance per heuristic. Practise writing a full violation row on a real app every day for a week: the problem, the heuristic, a cropped screenshot, a 1 to 4 severity justified by frequency, impact and persistence, and a recommendation.
Then compare your list with a classmate’s on the same app and merge them into one collated list, averaging severity where you overlap; that rehearses the team process the seminar describes. Finally, map each of Shneiderman’s golden rules to its closest heuristic and note the two that have no clean match.
Working through Heuristic Evaluation of Interfaces in FIT5152? Sia is AskSia’s AI Information Technology tutor — ask any FIT5152 Heuristic Evaluation of Interfaces question and get a clear, step-by-step explanation grounded in how FIT5152 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
