Evaluating visualisation tools and interrogating a visualisation: HSC Enterprise Computing Data Visualisation
“Evaluate the effectiveness of software tools used to develop data visualisations, including spreadsheets used to develop dashboards, presentation software used to present data analysis, business analytics services including 'as a service' products, and custom software solutions; interrogate data from a data visualisation, including interpreting what you see, aggregation, filtering, the effect of outliers and reasoning”
Evaluate visualisation tools by cost, skills, data volume, interactivity, sharing and security: spreadsheets are cheap and flexible, presentation software tells a static story, business analytics services offer live interactive dashboards by subscription, and custom software fits exactly but costs most. Interrogate any visualisation by interpreting it carefully, questioning aggregation and filtering, checking outliers and reasoning about causes.
Jump to a section
What this dot point is asking
There are two halves. First, evaluate the tools that build visualisations. Second, interrogate a finished visualisation: read it critically rather than accepting it at face value. Exam items often give a chart as stimulus and ask what can and cannot be concluded.
The answer
Evaluating software tools
Useful criteria: cost, ease of use and skills needed, data volume and connection to data sources, interactivity, real-time updating, collaboration and sharing, security, and customisation.
| Tool | Strengths | Limitations |
|---|---|---|
| Spreadsheets used to develop dashboards | Cheap, widely known, flexible formulas, pivot tables, charts and slicers | Limited with very large or live data, manual errors, version control and sharing problems |
| Presentation software used to present data analysis | Builds a clear narrative for an audience, combines charts, text and images | Static snapshots, charts must be updated by hand, easy to cherry-pick |
| Business analytics services, including 'as a service' products | Connect to many data sources, automatic refresh, interactive drill-down and maps, browser and mobile access, provider maintains the software | Subscription costs, data held in the provider's cloud, dependence on internet and vendor |
| Custom software solutions | Built exactly for the enterprise's needs and integrated with its systems | Expensive, slow to develop, requires programmers to maintain |
"As a service" means the software is hosted by the provider and rented by subscription (software as a service), rather than installed and maintained by the enterprise.
Interrogating data from a visualisation
- Interpreting what you see: read the title, axes, units, scale, legend and source before the shapes. State what the chart actually shows.
- Aggregation: ask what has been summed or averaged, and what variation that hides.
- Filtering: ask what data has been included or excluded (time range, groups, products).
- The effect of outliers: check whether extreme values distort the scale or the average, and whether they are errors or real.
- Reasoning: separate evidence from interpretation, avoid confusing correlation with causation, and consider alternative explanations.
A dashboard shows "Average customer spend rose from 42 to 55 dollars this month".
- Interpret: it is a mean, for online orders only (the filter is shown in the corner).
- Aggregation: the mean hides the distribution; a few very large orders could raise it.
- Outliers: drilling down shows one corporate order of 12,000 dollars. Without it, average spend is 43 dollars, almost unchanged.
- Filtering: store sales are excluded, so the claim does not describe all customers.
- Reasoning: the rise is caused by one unusual order, not a change in typical customer behaviour. Report the median as well.
- Evaluating tools without a scenario
- Always judge fitness for the given enterprise and data.
- Reading only the shape of a chart
- Check axes, scales, filters and sources first.
- Deleting outliers automatically
- Investigate them; they may be errors or the most important finding.
Practice questions
Original practice questions graded from foundation to exam level, each with a full worked solution. Try them before revealing the solution.
foundation3 marksA dashboard shows average weekly sales per store as a single bar per state. Explain how aggregation and filtering could hide important information.Show worked solution →
Aggregation: averaging all stores in a state hides differences between them; one very successful city store could mask several struggling regional stores.
Filtering: if the dashboard is filtered to the current quarter or to certain product lines by default, trends across the year or in excluded products are hidden.
Marking guide: 1 mark for aggregation, 1 mark for filtering, 1 mark for a specific consequence.
core4 marksA chart shows average house sale prices by suburb, and one suburb is far higher than the rest. Explain how you would interrogate this outlier.Show worked solution →
- Check the data: is it a data entry error (an extra zero) or a real sale? Look at the individual sales behind the average.
- Check the count: if only two houses sold, one mansion could create the high average; the median would be more representative.
- Check filtering and time range: was the period unusual?
- Reason about causes: if the value is real, is there an explanation (waterfront homes, a new development) before drawing conclusions about the suburb.
Marking guide: 1 mark per interrogation step.
exam6 marksA medium-sized logistics company must choose between (a) spreadsheet dashboards, (b) a business analytics service on a subscription and (c) custom software to visualise its delivery data from 200 trucks. Evaluate the options and recommend one.Show worked solution →
- (a) Spreadsheet dashboards
- Cheap and familiar, with pivot tables, charts and slicers. However, spreadsheets struggle with large, continuously updated data from 200 trucks, are prone to manual errors and version conflicts, and are hard to share securely in real time.
- (b) Business analytics service
- A cloud subscription service connects to the company's databases and GPS feeds, refreshes automatically, supports interactive maps and drill-down, and can be viewed on any device with role-based access. Costs are predictable per user, and the provider handles updates. Disadvantages: ongoing fees, dependence on the provider and internet, and data stored in the provider's cloud.
- (c) Custom software
- Could be tailored exactly (live truck maps inside the dispatch system) but is expensive, slow to build and needs ongoing developer support, which a medium-sized company may not have.
- Recommendation
- Option (b) best balances capability and cost: it handles the data volume and real-time needs without the cost of custom development. Spreadsheets can still be used for ad hoc analysis, and custom software could be reconsidered if needs become very specialised.
Marking guide: 1 mark each for evaluating (a), (b) and (c) with strengths and weaknesses (3 marks), 2 marks for linking to the company's data volume and needs, 1 mark for a justified recommendation.