A Prettier Graph Can Still Lie Data Handling and Presentation

Navya and Naresh disagree about which game their class likes best -- and the only real way to settle it is to survey every single classmate and count the actual votes, rather than guessing from a handful of loud opinions. Collecting everyone's answer and tallying them (using simple tally marks, four straight lines and a fifth struck through to mark each group of five) turns a room full of individual opinions into one clear frequency table, and sorting a list of raw values (like a class's shoe sizes) from smallest to largest makes questions like 'what's the most common size?' or 'how many wear a size above 4?' answerable at a glance, rather than requiring the whole list to be re-scanned every time.

A pictograph represents data using a repeated symbol, with each symbol standing for a fixed number of items (a chosen scale) rather than just one -- so if 1 picture always equals 10 children, then 5 full pictures mean 50 children, and a half-picture means 5 more. Picking a good scale matters: a scale of 1 symbol per student works fine for small classes, but becomes unworkable once the numbers grow into the hundreds, and choosing a scale that doesn't divide the data evenly (like a scale of 10 when a class has exactly 33 students) creates the genuine, unavoidable headache of drawing a partial symbol.

A bar graph replaces repeated symbols with a single bar per category, whose height (read against a chosen scale on an axis) directly shows the frequency -- genuinely useful once real, messy data enters the picture, like Delhi traffic counts across a morning (rising sharply between 6 and 8am, staying high through the morning rush, then tapering off) or India's population by decade (climbing steadily from 36 crore in 1951 to 102 crore in 2001). Drawing one from scratch always follows the same steps: choose axes, pick a sensible scale for the data's actual size, compute each bar's height by dividing the real value by the scale, then draw -- exactly the process used to turn Imran's family budget (rent, food, education, electricity, transport, and other costs) into a single bar graph where a bar's height can be checked directly against another (confirming, for instance, that the electricity bar is genuinely about half the height of the education bar, matching the real numbers 400 and 800).

Taking the seven tallest mountains (one per continent, from Everest's 8848 metres down to Kosciuszko's 2228 metres) and showing them first as an ordinary bar graph, then as a fancier set of triangular 'mountain-shaped' bars whose WIDTH is also allowed to grow along with their height, reveals something genuinely important: once width is allowed to change too, the picture becomes visually misleading even though not one single number has changed -- a taller-and-wider triangle looks disproportionately more impressive than the plain bar heights alone would suggest. The lesson is not that artistic data presentation is wrong, but that it needs to be checked carefully against what the numbers actually say, since a prettier picture can quietly distort the very comparison it's supposed to make clear.