CSP-11 CSP-11 D2.14 · Domain 2 · Program Management · 25%
Descriptive Stats
Nine facilities each record 2 work-related injury or illness cases in one year; the tenth records 40. A director cites the median of 2 to defer inspection of the tenth facility. Which summaries describe these data, and what must be checked before ranking sites or choosing controls?
Updated . Independent CSP-11 practice. Not BCSP-endorsed. No pass guarantee.
Decision rule
Describe the data before you infer. Mean is not the data set.
Mean, median, mode, range. Outliers and mixed units lie.
In 20 seconds
- Describe the data before you infer.
- Mean is not the data set.
- Mean, median, and mode summarize differently — outliers pull the mean.
- Range and standard deviation describe spread.
How to apply Descriptive Stats
THEORY: Descriptive stats summarize a sample. Mean is pulled by extremes; median resists. Mode is the most frequent class. Do not treat a mean of mixed tasks as an exposure. CALC: 2, 3, 3, 12. Mean = 20/4 = 5. Median = 3. Mode = 3. Range = 10. TI-30XS: 2+3+3+12= then /4=. FIELD: Four noise samples, one 12-hour shutdown spike. Report median and the condition of the spike; do not hide it in the mean alone. ASSESS: Using the mean to declare ‘typical exposure’ when tasks differ is the trap.
Worked case
Wrong first move: quote only the mean of 5 after a 12 outlier. Correct move: show mean, median, and the operating condition of each point.
Why the rule holds
Related: 37, 38, 62, 92.
Same rule, different setting
Nine pads have 2 recordables a year. One pad had 40 after a blow-by. Leadership wants “the average is fine — fund nothing.” You are the CSP. The question is mean vs median vs the decision the numbers are allowed to make.
Nine jobs have 2 recordables a year. One job had 40 after a collapse. Leadership wants “the average is fine — fund nothing.” You are the CSP. The question is mean vs median vs the decision the numbers are allowed to make.
Picture / model
Compute location and spread for an explicit safety dataset, distinguish a typical facility from total burden and investigate unusual observations. Use matched case counts and hours for rate comparisons without confusing a descriptive measure with proof of cause or acceptable risk.
Exam traps
Original SafePath contrasts for Descriptive Stats. These are study traps, not recalled BCSP items.
| Trap | Correct read |
|---|---|
| Mean | Σx/n — sensitive to outliers |
| Median | Middle of sorted data — robust to skew |
| Mode | Most frequent value |
| Mean of TRIR | Still lagging, still not a control |
Field move
The mean is 58/10 = 5.8. It is pulled by the 40 and is not the typical value.
Scenarios
Scenario 1 · Fundamental · CSP-11 D2.14
You are the CSP reviewing a live decision at a plastics compounding turnaround. Operations proposes a convenience path. For the ten counts 2, 2, 2, 2, 2, 2, 2, 2, 2, and 40, what is the mean?
Best action: 5.8 because 58/10.
Why: The sum is 9×2 + 40 = 58. The mean is 58/10 = 5.8. That average is not the median or the mode of this set, and it should not be described as the typical count when one observation dominates the sum.
Scenario 2 · Applied Scenario · CSP-11 D2.14
You are the CSP reviewing a live decision at a Gulf Coast chemical terminal. Operations proposes a convenience path. For the ten counts 2, 2, 2, 2, 2, 2, 2, 2, 2, and 40, what is the median?
Best action: 2 (sorted; even n, two middles are 2 and 2).
Why: Sorted order is 2, 2, 2, 2, 2, 2, 2, 2, 2, 40. For ten observations the median is the average of the 5th and 6th values, both 2, so the median is 2. The mean 5.8 uses every observation. The midpoint of the minimum and maximum is not the median.
Scenario 3 · Applied Scenario · CSP-11 D2.14
Nine shops each report 2 events and one reports 40. The board asks for the typical shop despite the skew. What should the CSP report?
Best action: Median 2 for typical, plus the 40 as a separate site that still needs investigation/engineering.
Why: Evaluate the best response. Nine shops report 2 events and one reports 40. The board wants the typical shop despite the skew. The best report is the median of 2 for typical, plus the shop at 40 as a separate site that still needs investigation and engineering. Quoting the mean of 5.8 as typical and…
If the stem says this
| Stem signal | Best direction |
|---|---|
| outliers dominate the mean | consider median or investigate outliers |
| the stem asks for spread | use range or standard deviation — not only the mean |
| units differ across datasets | do not average until converted |
| two answers work | higher hierarchy / system / design |
| a chart truncates the axis to exaggerate change | reject the misleading display |
Check before you leave
- ✓ Mean, median, and mode summarize differently — outliers pull the mean.
- ✓ Range and standard deviation describe spread.
- ✓ Choose the statistic that matches the question and the distribution.
- ✓ Do not average incompatible units or categories.
- ✓ Graphs must not mislead with truncated axes or cherry-picked windows.
- ✓ Sample versus population language matters in explanations.
Blueprint mapping
Descriptive Stats is the SafePath study class for CSP-11 D2.14 inside CSP-11 Domain 2, Program Management (25%). A high-weight management domain focused on measuring performance, investigating incidents, managing change, building systems, leading people, controlling records and resources, and interpreting data.
SafePath uses the publicly posted CSP-11 outline as a study map. It does not reproduce official exam content, and it is not affiliated with or endorsed by BCSP or Pearson VUE.
Practice / flashcards / formulas
More classes in Program Management
Nearby lessons mapped to the same CSP-11 domain.
CSP-11 D2.14
Other classes on this task
Guides that use Descriptive Stats
- Oil and gas safety study guideProcess safety, change, and investigations
- CSP application and exam guideEligibility, scheduling, and exam method
- CSP-11 exam prep guideDomain weights, sample items, and a study sequence
- CSP-11 blueprint and weightsSeven published domains and the class map
- Seven-domain study directoryScope, tasks, and links into each lesson
Descriptive Stats study questions
- What decision does the Descriptive Stats class train?
- Describe the data before you infer. Mean is not the data set. This is independent SafePath CSP practice for CSP-11 D2.14. It is not an official BCSP item and it is not a pass guarantee.
- What exam trap shows up in Descriptive Stats?
- A common miss treats this as sufficient: Mean The stronger read is: Σx/n — sensitive to outliers
- How does Descriptive Stats map to the CSP-11 blueprint?
- SafePath maps this class to CSP-11 D2.14 in Domain 2, Program Management (25% of the published blueprint). Use the public outline for study planning. SafePath is not affiliated with or endorsed by BCSP.
Verify
Verify: Confirm current certification requirements with BCSP and technical requirements with the cited authoritative standard.