How to read placement statistics before choosing a course

Ask who is counted before asking how high the number is

“Ninety per cent placed” is incomplete without a denominator. Does it refer to everyone who graduated, students eligible under the institution’s rules, students who registered for placements or a particular programme? Ask for the number of students in each group, the reporting year and the date outcomes were collected. Higher study, self-employment and people still seeking work may be reported separately. Do not assume every institution uses the same categories. Keep the original wording rather than silently translating it into a stronger claim.

Separate offers, people and jobs started

One person can receive more than one offer. An offer may be withdrawn, declined or begin later. An internship, a temporary contract and a permanent job may also be grouped differently. Ask how the institution counts each and whether the role is related to the course. A list of employer logos does not tell you how many students they hired, for which jobs or in which year. Seek a clear cohort-level record and treat unexplained gaps as uncertainty, not proof of either success or failure.

Read salary summaries with the right population

The NIRF data-definition manual illustrates a crucial distinction: median salary is calculated for placed graduates, not automatically for every graduate. The linked manual is historical, so use the current cycle’s definitions for a live comparison. Ask whether a reported figure is median or mean, domestic or international, and total compensation or fixed pay. Currency, location and living costs also affect interpretation. Do not use the highest package as a prediction of what you personally will receive.

A worked example with invented numbers

Imagine a programme with 120 graduates. Of these, 90 register for placement and 72 receive at least one offer. The placement rate is 80% of registered students, but 60% of all graduates. Both calculations can be correct while answering different questions. If 110 offers were made, that still does not mean 110 people were placed. The example is invented to show the arithmetic; it is not a statistic about any university. Your next question is what happened to the other graduates and how the institution defines each outcome.

Compare like with like across more than one year

Request programme-level information where available, with the same definitions across recent cohorts. University-wide data can combine very different courses. A small cohort may produce a volatile percentage; show the underlying counts. Look at job titles and destinations as well as salary, and ask what support students received before applying. The NIRF methodology provides a route to institutional submissions, but a ranking or submitted dataset cannot guarantee an individual outcome. Check the source and any limitations before using the figures in a cost decision.

• Same programme, level, campus and reporting period.
• Graduates, eligible students, participants and placed people separately counted.
• Offers distinguished from people and jobs started.
• Salary definition, currency and population stated.
• Higher study, unknown outcomes and exclusions visible.

Turn unclear claims into precise questions

Instead of asking “Are placements good?”, ask for the latest three cohorts’ graduating numbers, participating numbers, unique students placed and the definition of the salary statistic. Ask whether off-campus outcomes are included and how they are verified. Speak with graduates about the application process while keeping their accounts separate from cohort data. If the institution cannot supply a detail, mark it unknown in your comparison. Choose with an explicit understanding of that uncertainty rather than building your future budget around a promotional headline.

Next actions

1. Rewrite each placement percentage with its denominator.
2. Ask for unique people, role types and salary definitions.
3. Compare defined cohorts and keep unknown outcomes visible.

Continue this career investigation