Interpreting and Presenting Data

Methods to Interpret and Present Data

Why Data Interpretation Matters

Collecting data is only the first step. Interpretation transforms raw data into actionable insights — conclusions that inform design decisions, product improvements, and evaluation judgments.

Poor interpretation (misreading trends, ignoring contradictions, over-generalising from small samples) leads to poor design decisions even when the underlying data is sound.

Interpreting Quantitative Data

Descriptive statistics:
- Mean (average): Sum of all values divided by count. Useful for overall trends but sensitive to outliers.
- Median: Middle value when ranked. More robust than mean when data is skewed.
- Mode: Most frequent value. Useful for rating data.
- Range: Spread between maximum and minimum values. Indicates variability.

Identifying trends:
- Look for patterns over time or across categories
- Identify outliers (values far from the average) — investigate why they differ
- Compare to benchmarks (evaluation criteria, competitor data, industry standards)

Statistical significance caution:
With small sample sizes typical in design research (5–20 end users), quantitative results should be interpreted cautiously. Do not over-claim certainty from small samples.

Interpreting Qualitative Data

Thematic analysis:
- Read all responses; identify recurring themes or patterns
- Code responses: assign a label to each recurring idea
- Tally how many participants mentioned each theme
- Prioritise themes mentioned most frequently or most emphatically

Triangulation:
- Cross-check qualitative findings with quantitative data
- If 70% of survey respondents rated ergonomics 3/5 or lower, and interview participants frequently mentioned discomfort, both sources confirm the same problem

Contradiction:
- Note when data sources disagree — this reveals complexity that simplistic interpretation would miss
- Report contradictions honestly; do not suppress inconvenient data

Presenting Data

Tables:
- Present raw measurement data or criterion-by-criterion evaluation results
- Easy to compare values; does not visualise trends well

Bar graphs:
- Compare values across categories (e.g. average rating per criterion across multiple end users)
- Clear visual comparison; effective for nominal and ordinal data

Line graphs:
- Show trends over time or sequences (e.g. load vs deflection in structural testing)
- Reveals patterns that are invisible in tables

Pie charts:
- Show proportions of a whole (e.g. percentage of users who preferred each design option)
- Use sparingly; can be misleading with many small segments

Rating matrices:
- Display criterion evaluations for multiple products or concepts side by side
- Useful for concept selection in design development

Annotated photographs:
- Present visual evidence of test results or product features
- Annotations connect visual evidence to evaluation criteria

Quotes and verbatim extracts:
- Present qualitative data authentically
- Select representative quotes; attribute to participant (anonymously if required)

Principles of Good Data Presentation

• Label all axes, tables, and graphs clearly
• Cite sources for secondary data
• Distinguish between data (what was observed) and interpretation (what it means)
• Use appropriate precision: don’t report 4 decimal places when 1 is sufficient
• Present a balanced picture: include negative findings, not just supportive data

KEY TAKEAWAY: Data interpretation requires both analytical skill (reading the numbers correctly) and critical judgment (understanding what they mean for the design). Presentation must be clear, accurate, and honest.

EXAM TIP: If asked to present data from a scenario, choose an appropriate format (table, graph, matrix) and explain why it is appropriate for that type of data. This demonstrates data literacy.