Compare Trip Speeds Across Time

# Compare Trip Speeds Across Time Derive average trip speed from distance and duration. Test whether average speeds are materially the same across all weeks of September, and develop a hypothesis for how average speed varies by time of day. ### Constraints & Assumptions - State units and timezone. - Exclude or separately report nonpositive durations, negative distances, impossible timestamps, and implausible speeds. - “Materially the same” requires a practical threshold, not only a p-value. ### Clarifying Questions to Ask - How are partial weeks at the month's boundaries defined? - Is the estimand trip-weighted mean speed, distance divided by total time, or another measure? - What speed difference would matter operationally? ```hint Define the estimand before the test The mean of per-trip speeds is not the same as total distance divided by total duration. ``` ### What a Strong Answer Covers - A defensible speed calculation and quality filters. - Week definitions, exploratory summaries, and a suitable inferential method. - Effect sizes, uncertainty, multiple comparisons, and practical significance. - A time-of-day pattern evaluated against plausible confounders. ### Follow-up Questions - How would repeated traffic patterns by weekday affect the test? - How would you distinguish congestion from a changing trip-location mix?