DSC 330 Week 2 Matching Chart Types to Questions Example

Which Chart Answers Which Question? Matching Visuals to a Hotel Company's Business Questions

[Student Name]

University of Phoenix

DSC/330: Data Communication and Visualization for Business

Week 2 Assignment

[Instructor Name]

[Date]

Music Row Hospitality, its hotels, data and figures are composites written for a model paper.

Week 1 gathered requirements for a dashboard at Music Row, the invented Nashville firm that runs nine hotels for six owners. Stakeholders produced a set of business questions, from the chief operating officer's portfolio comparisons to general managers' daily checks. This paper chooses a chart for each question before anything is built.

Questions Have Types

Shneiderman (1996) proposed a task-by-data-type taxonomy for information visualization, summarized in his mantra "overview first, zoom and filter, then details on demand," and argued that visual designs should match both the type of data and the task the user is trying to perform. A business version of this idea classifies questions by what the user wants to see: a comparison among categories, a trend over time, the distribution of values, a relationship between measures or the composition of a whole. Each type points toward certain charts.

What Research Says About Reading Charts

Cleveland and McGill (1984) ranked the elementary perceptual tasks people use to read graphs by accuracy, finding that people judge positions on a common scale best, then lengths, then angles and slopes, then areas, with color saturation least accurate. Ware (2021) explains that visual features such as position, length, color and shape are processed differently by the human visual system, and that designers should use the most accurately perceived features for the most important comparisons. These findings favor bars, dots and lines over pies, bubbles and color scales whenever precise comparison matters.

Choosing a Chart for Each Question

Question 1, comparison: Which hotels have the highest revenue per available room this week? Chart: horizontal bars ordered from the best hotel down to the weakest, each with a small tick for its budget. Sorting lets readers find the best and worst at once; horizontal bars leave room for hotel names.

Question 2, trend: Is occupancy at my hotel above or below budget this month? Chart: a line of daily occupancy for 30 days with a shaded band showing the budget, so days below budget stand out.

Question 3, distribution: How are guest satisfaction scores spread out? Chart: a histogram of individual survey scores from 1 to 10. An average of 8.1 hides whether most guests are content or whether many are delighted and some furious; the histogram shows a cluster at 9 and 10 and a smaller cluster at 3 and 4.

Question 4, relationship: Do higher room rates reduce occupancy? Chart: a scatter plot of daily average rate against occupancy for each hotel, with one dot per day. The pattern shows little effect at most hotels except the roadside property, where occupancy falls sharply above $129.

Question 5, composition: Which departments generate most complaints? Chart: a sorted bar chart of complaint counts by department. A pie with eight departments would make the close categories, housekeeping and front desk, impossible to compare.

Question 6, comparison over time: How does each hotel's satisfaction score this quarter compare with last year? Chart: a dot plot with two dots per hotel connected by a line, showing direction and size of change.

Question 7, ranking with detail: Which hotels are below target on several measures at once? Chart: a simple table with conditional marks, since users need exact values for three measures.

Question 8, geographic: Where are competitor hotels and how do their rates compare? Chart: a map with competitor locations and a side table of rates; the map gives context, the table gives precision.

An average satisfaction score of 8.1 sounded fine until the histogram showed a second, unhappy crowd scoring 3 and 4.

Charts Rejected and Why

Several charts were considered and rejected. Gauges and speedometer dials, popular in dashboards, use angles and take much space to show one number; a bullet chart or simple bar with a target mark conveys the same information more compactly. Three-dimensional bars distort length. Stacked area charts of revenue by source make every layer except the bottom hard to read. Dual-axis charts invite readers to see relationships that depend on arbitrary axis scales.

Testing the Choices Quickly

Before building, paper sketches of the eight charts were shown to two general managers and the revenue manager, each asked what they would conclude from the chart and what they would do next. The sorted bar chart and the occupancy line were understood immediately. The histogram confused one general manager, who expected an average; adding a marked median and a one-line title, "Most guests score 9 or 10, but one in eight scores 4 or lower," solved it. The scatter plot needed a note explaining that each dot is one day. Testing sketches costs an hour and prevents building charts people misread.

Design Details

Charts on the dashboard will share conventions: sorted bars, labels on the data rather than in legends, one accent color for items needing attention and gray for everything else, axes that start at zero for bars and titles that state the takeaway. For general managers viewing on phones, each chart will fit the screen width with no more than one question per chart.

Audience Differences

The chief operating officer's portfolio view emphasizes comparisons across hotels; the general managers' view emphasizes trends and daily status; owners' monthly view emphasizes comparisons with market benchmarks. The same question can need a different chart for a different user: an owner wants a monthly bar against the market, while a general manager wants a daily line against budget.

What the Dashboard Will Not Show

Some requested items were left out. One general manager asked for a chart of every guest comment; the dashboard will instead show complaint counts by topic, with comments available on demand. Owners asked for staff names next to complaints; that raises privacy and fairness concerns and stays out.

Conclusion

Each question from Week 1 has a type, and each type has charts that answer it well. Research on graphical perception and visualization tasks favors sorted bars for comparisons, lines for trends, histograms for distributions, scatter plots for relationships and bars rather than pies for most compositions. Rejecting gauges, 3-D effects and dual axes and adopting shared conventions prepare the dashboard for building in Week 3.

References

Cleveland, W. S., & McGill, R. (1984). Graphical perception: Theory, experimentation, and application to the development of graphical methods. Journal of the American Statistical Association, 79(387), 531-554. https://doi.org/10.1080/01621459.1984.10478080

Shneiderman, B. (1996). The eyes have it: A task by data type taxonomy for information visualizations. In Proceedings of the IEEE Symposium on Visual Languages (pp. 336-343). IEEE. https://doi.org/10.1109/VL.1996.545307

Ware, C. (2021). Information visualization: Perception for design (4th ed.). Morgan Kaufmann.