Test it yourself: does the arithmetic hold up?

This lab loads Party A's and Party B's vote share in each of five seats as two dot plots, with live mean, median, mode and range as you edit any value. Starting values match the worked example: A has 51%, 51%, 51%, 10%, 10% (winning three seats narrowly, losing two heavily); B has the complements.

Two challenges: find Party A's mean vote share (well under half, despite winning most seats), and find the range of A's results (the size of the gap between its narrow wins and heavy losses). Try dragging A's losing seats from 10% up towards 40% one at a time and watch the mean rise while the seat count stays exactly the same until a seat actually flips — seats change in steps, means change smoothly.

Two modes in one lab session. The bill mode lets you flag a bill as a money bill before starting, which locks out the Rajya Sabha's amendment and joint-sitting options and instead shows the 14-day "deemed passed" countdown; running it as an ordinary bill and forcing a rejection instead opens the joint-sitting branch, matching what you tested with predictions above.

The election mode lets you set up your own toy country: choose how many seats, and enter a vote split for two or three parties per seat, either uniform across all seats or varied seat by seat. It then shows the seats won, the vote share and the seat share side by side, exactly like the worked examples in this chapter, so you can build your own test cases beyond the ones already given.

All 18 general-election turnout percentages, from 1951-52 to 2024, load as one dot plot with live mean, median, mode and range readouts.

Three challenges: find the mean (about 58.8%), the median (about 59.1%, the average of the two middle values since there are an even number of elections), and the range (about 21.96 percentage points, between the lowest, 1957 at 45.44%, and the highest, 2019 at 67.4%).

Then try the open-ended part: drag the dots around to invent a data set with the same mean as the real one but a much larger or smaller range, and notice that mean alone hides a great deal about how bumpy the real history actually was.

A dot plot of the three per-person figures — ₹1,600, ₹500 and ₹5,000 — with a challenge to find the range between the highest and lowest. Beneath the plot, the underlying fund and population for each village stay editable, so you can change either number and watch the per-person figure update immediately, testing your own "what if the fund doubled" or "what if the population halved" questions directly.

A three-bin sorting game — Local, State, Union — covering twelve real situations: a pothole, a delayed train, hospital overcrowding, a post office, market cleanliness, school curricula, a passport, police response, property tax, interest rates, a state highway and panchayat accounts. After each card, a short note explains the reasoning, deliberately including cases that are easy to get backwards, such as police (State, not Union) and railways (Union, not State).

A seven-pair matching game covering the vocabulary this lesson introduced: wasted vote, percentage point, seat share, vote share, coalition, the unequal-constituency-size problem, and proportional representation. Each term reappears in the closing quiz, so this is good rehearsal.