Handle Freshness and Conflicting Evidence

Today's Goals

Today you will inspect a correction to the fictional Harborfield transit count without letting the newer page erase the older observation. You will record publication date, observation period, unit, denominator, and method separately. You will distinguish a real contradiction from two compatible numbers measuring different things. You will write a conditional finding that Jo can use while the remaining questions are unresolved. The artifact is a freshness and conflict matrix, not a choice of whichever number looks strongest for the proposal.

The new card, HF-05, is entirely fictional. It was published after HF-01 but refers to the same July period. It corrects a label in a preliminary presentation: 140 boarding events had been called “140 riders.” It also reports 102 distinct fare cards among those events, while cash riders could not be linked to cards. This is a teaching case about definitions. It does not describe real transport data and was not discovered by a live Perplexity search.

Plain-Language Walkthrough

Put several clocks on the table

When someone says a source is “recent,” ask which clock they mean. The publication date tells when a page appeared. The observation period tells when the underlying events occurred. The access date tells when Maya saw the page. The correction date tells when an interpretation changed. These clocks can point to different months. A page published in August about July activity is not evidence that conditions in September stayed the same. A correction published later can improve the interpretation of July data without measuring a new July or August population.

Think of a grocery receipt printed today for food bought yesterday and then corrected tomorrow because one item had the wrong category. Tomorrow's receipt is newer, but it does not mean the food was bought tomorrow. It may keep the same total while changing what that total represents. HF-01 and HF-05 work that way. The source register from D2 preserved the original wording; today's matrix makes the correction and its limits explicit. If D2 had recorded only “140 riders,” the error would already have entered the evidence base before the correction appeared.

HF-01 was published on August 1 and covered four sampled July weekday weeks after 21:00 at Stops A and B. It records 84 boarding events at A and 56 at B. That is 140 boarding events in the specified sample. It excludes weekends and other stops. HF-05 was published on August 19 and refers to the same observed period. It says a presentation wrongly called the 140 events “140 riders.” It identifies 102 distinct fare cards among those events, but cash riders were not linked to cards. A card is not automatically a person, and even 102 cards is not the total number of unique riders.

Sign in and link WeChat to read

Sign in, then link your WeChat mini program account on the account page. Linked accounts share one free article per day. Unlock additional articles with rewarded videos in the mini program.

Scan in WeChat to open Dayward AI. Open the same course and chapter, choose the rewarded video, then refresh this page after completion.

Unlocked — refresh to readSign in / scan with WeChat →