DATA 5003 Module 3 Descriptive Statistics and Visualization Analysis Example

One in Five Fail: Descriptive Statistics and Charts for 11,426 Chicago Restaurant Inspections in 2025

Student Name

American College of Education

DATA5003: Data Analytics

Module 3 Assignment

Instructor Name

March 20, 2028

Introduction

This paper takes the first analytical step toward answering the question I set in Module 1 for our restaurant group: what predicts a failed city inspection? Before looking for relationships, the data must be described. From the city's open inspection records (City of Chicago, 2026), I selected every inspection of a facility coded as a restaurant during calendar year 2025. Of 13,864 such inspections, 11,426 ended in one of the three outcomes that matter for this question: pass, pass with conditions or fail. This paper describes those 11,426 inspections and explains how each finding should be shown visually.

Approach

The approach follows the spirit of exploratory data analysis that Tukey (1977) championed: look at the data from several angles, with simple summaries and pictures, before testing any hypothesis. Each inspection was treated as one observation. For each, I recorded the result, the inspection type, the city's risk category, the month, the zip code and the set of distinct violation numbers cited, extracted from the text field as described in Module 2. Percentages are calculated within each group, so that a group's failure rate is the number of failed inspections in that group divided by all inspections in that group that reached a result. No inspections were removed for unusual values; the only exclusions were the outcomes without a result.

Overall Results

Of the 11,426 inspections, 7,198, or 63.0 percent, passed; 1,805, or 15.8 percent, passed with conditions, meaning violations were found that had to be corrected; and 2,423, or 21.2 percent, failed. Roughly one restaurant inspection in five therefore ended in failure. The remaining 2,438 inspections excluded from the analysis were mostly cases in which inspectors could not enter, the business had closed or a new kitchen was not ready, outcomes that say little about food safety practice. The inspections covered 7,548 distinct licensed establishments, so most restaurants were inspected once or twice during the year. Of those establishments, 2,079 failed at least one inspection during the year and 280 failed two or more times.

Results by Inspection Type

Failure rates differed sharply by why the inspector came. Routine canvass inspections, the largest group at 5,961, failed 23.2 percent of the time. Inspections triggered by a complaint, 1,367 in all, failed 35.3 percent of the time, the highest rate of any common type. License inspections of new or changed businesses failed 22.8 percent of the time. Re-inspections, which follow an earlier failure or conditional pass, had much lower failure rates: 8.5 percent after a canvass, 12.7 percent after a complaint and 4.2 percent after a license inspection. Most kitchens fix the problems found on the first visit.

Results by Risk Category and Month

The city assigns each establishment a risk category based on the kind of food handling it performs. Most restaurants, 9,699 inspections, were in the high-risk category, and they failed 21.3 percent of the time; medium-risk restaurants failed 20.9 percent of the time and the 226 low-risk inspections 18.1 percent. The differences are small, suggesting the category reflects menu complexity more than kitchen practice. By month, the failure rate ranged from 18.6 percent in April to 24.7 percent in August, with higher rates in late summer, when heat strains refrigeration and pests are more active. A line chart of the twelve monthly rates makes the late-summer rise easy to see, and a reference line at the annual average of 21.2 percent shows which months stand out.

Results by Zip Code

Geography shows wider variation. Among zip codes with at least 150 inspections, the lowest failure rate was 7.5 percent, in the zip code covering O'Hare International Airport, where concessions operate under airport management. The highest were 29.3 percent in the River North area, 30.7 percent in the Wicker Park and Ukrainian Village area and 32.3 percent in a zip code on the Northwest Side. These differences may reflect restaurant density, building age, the mix of independent and chain operators or inspection patterns, and they cannot be explained by description alone. They do show that location deserves attention in the relationship analysis.

Violations per Inspection

The number of distinct violations cited at each inspection rises steadily with the severity of the result. Passing inspections averaged 1.44 violations, with a median of one; conditional passes averaged 3.23, with a median of three; and failed inspections averaged 4.10, with a median of four. Grouping inspections by violation count shows the same pattern: inspections with no violations failed 4.2 percent of the time, those with three failed 24.9 percent, and those with six or more failed 67.3 percent. Failures are rarely the result of a single problem; they tend to reflect several at once.

The Most Frequent Violations

Across all 11,426 inspections, the most frequently cited violation was number 55, physical facilities installed, maintained and clean, which covers floors, walls and ceilings, with 7,464 citations. It was followed by number 47, surfaces cleanable and properly designed, at 4,044; number 49, surfaces clean, at 3,164; number 10, handwashing sinks properly supplied and accessible, at 2,257; number 38, insects, rodents and animals not present, at 2,216; and number 58, allergen training as required, at 1,929. Frequency is not the same as importance, however: a common violation may appear at passing and failing inspections alike, a question Module 4 will test.

Choosing the Right Charts

How these findings are shown affects how managers read them. Cleveland and McGill (1984), in experiments on graphical perception, found that people judge values most accurately when they are shown as positions along a common scale, as in bar and dot charts, and less accurately when shown as angles or areas, as in pie charts. The dashboard will therefore use horizontal bar charts for failure rates by inspection type and zip code, a line chart for monthly failure rates, a bar chart of violation counts for each outcome and a sorted bar chart of the most frequent violations. Pie charts will be avoided, even for the overall pass, conditional and fail shares.

What the Description Suggests

Description cannot establish cause, but it points the analysis in useful directions. Complaint inspections fail far more often, so customer complaints are an early warning worth tracking at our kitchens. Late summer brings higher failure rates, which suggests extra attention to refrigeration and pest control in July and August. Failed inspections involve several violations at once, so the next step is to find which violations most often accompany failure rather than which are most common overall. Our own two failed inspections in 2025 both occurred in August, both followed complaints and both cited pests and handwashing sinks, which fits the citywide pattern.

Conclusion

In 2025, 21.2 percent of Chicago restaurant inspections that reached a result ended in failure, with much higher rates for complaint inspections, in late summer and in certain zip codes. Failed inspections averaged four violations, and the most frequent violations concerned facility cleanliness, surfaces, handwashing sinks, pests and allergen training. Charts built on position along a common scale will present these findings clearly. Module 4 will test which violations are most strongly associated with failure.

References

City of Chicago. (2026). Food inspections [Data set]. Chicago Data Portal. https://data.cityofchicago.org/Health-Human-Services/Food-Inspections/4ijn-s7e5

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

Tukey, J. W. (1977). Exploratory data analysis. Addison-Wesley.