Reading a US Transcript: What Letter Grades Mean for International Students

Students who move to the United States from a system based on marks out of 100, a 1–10 scale or division classes often find their first American report card confusing. Scores turn into letters, letters turn into grade points, and the same percentage can produce a different letter at two schools in the same city. This guide explains how the pieces fit together.

From percentage to letter

Most US schools convert a percentage into one of five letters: A, B, C, D or F. E is usually skipped, although a few schools use it instead of F. The most common pattern is the ten-point scale, where 90–100 is an A, 80–89 a B, 70–79 a C, 60–69 a D and anything below 60 an F. Many schools add plus and minus steps, so that an 88 may appear as a B+ and an 81 as a B−.

This scale is a convention, not a national rule. Each district, school or college sets its own cut-offs. Some use a stricter scale in which an A starts at 93 or 94. Before comparing your results with anyone else’s, look for the grading scale printed on the transcript or in the student handbook. A detailed percent to letter grade chart shows the common variants side by side.

From letter to GPA

Each letter carries grade points: usually 4 for an A, 3 for a B, 2 for a C, 1 for a D and 0 for an F. The grade point average (GPA) is the average of those points across all courses, often adjusted for how many credits each course is worth. A student with mostly B grades will have a GPA close to 3.0.

Two details catch newcomers out. First, an A+ rarely adds anything: at many institutions it counts as 4.0, the same as an A. Second, some high schools publish a weighted GPA that gives extra points for honors or Advanced Placement courses, which is why a GPA above 4.0 can appear on a transcript.

Passing is not always enough

A D is normally a passing grade, but passing a course and being able to use it are different things. Colleges often require a C or better for courses in the major, for prerequisites and for transfer credit. Scholarships and good academic standing, which schools also expect of international students, usually depend on the overall GPA, not on individual passes.

Why direct conversion from your home system fails

It is tempting to translate grades by percentage alone: 65 percent at home equals 65 percent in the US, which is a D. That is usually wrong. In several systems a mark of 60 or 70 percent is a strong result, while on the common US scale it is a D or a C. For this reason, universities rarely convert foreign grades by percentage alone. They use credential evaluation services or their own admission staff, who compare the grading culture of each country.

What to do with your first grades

  • Find the official scale of your school before judging a result.
  • Check the syllabus: it states how homework, tests and participation are weighted.
  • Ask which minimum grade your program requires for prerequisite courses.
  • Keep your home-country transcripts and their official grading key for later applications.

Once the scale is clear, American grades are easy to read. The letter tells you where you stand in one course; the GPA tells everyone else where you stand overall.

Describe a failure that, looking back, was actually one of the best things that ever happened to you.

View all responses

Free sports datasets for students and data science classes

Students learn data work faster when the data means something to them. Sales figures from an unknown company teach the mechanics. A season of match results teaches the same mechanics while the class argues about who should have won. Sport gives a teacher a subject most students already have opinions about.

This guide covers what makes a sports dataset useful in class, what to check first, and exercises that work at each level.

7a1cda26db.webp

What a teaching dataset needs

A dataset for a class has different needs from one used in research. It should open on a laptop and explain every column. Any mess in it should teach something, not waste a lesson.

Look for a clear licence first. If the terms are vague, students cannot share their work or publish a project built on it. A dataset released for academic use, with the terms written down, saves a conversation later.

Then read the documentation. Every column needs a description, a unit and a note on how missing values are written.

Finally, check the size. A few thousand rows suits a first course. Tens of thousands suit a course on databases or modelling, where performance starts to matter.

Why tennis works well for teaching

Tennis is a good sport for a data class because its structure is strict. A match is made of sets, a set of games and a game of points, and every point has one winner. That hierarchy gives students practice with grouping and aggregation without the confusion of team substitutions or a running clock.

The scoring also produces interesting questions. Why does winning more points not always mean winning the match? How much does serving first in a set matter? These questions have answers students can work out from the data, and the answers are not always what they expect.

A point-by-point record adds another layer. Instead of one row per match, students get one row per point, with the server, the score before the point and who won it. That is enough to rebuild the score of any match and to study pressure moments such as break points.

For a class that wants this data, Live Tennis API publishes a free tennis dataset for research that teachers and students can request. Its historical archive holds point-by-point records from 2013, which gives a class several seasons to compare.

Exercises by level

For an introductory course, start with counting and grouping. Ask students to find the share of service games each player held in a season, then to compare two players. The work teaches filtering, grouping and percentages, and the result is easy to check against common sense.

For an intermediate course, move to joins and time. Give students a table of matches and a table of points, and ask them to rebuild the score after every point of one match. Students who get this right understand keys, ordering and running totals.

For a course on modelling, ask a question with a probability in it. Given the score, how likely is the server to win the game? Students can estimate this from the data, compare their estimate with a simple model and discuss where the two disagree. It is a good first project in logistic regression, with no need for advanced methods.

Checking the data before class

Run a few checks yourself before you hand a dataset to students. Count the rows per season to see whether coverage is even. Look for duplicate matches. Check that the scores add up, since a set should not end at an impossible score.

Write down what you find, even when it is a small gap. Telling students about a known gap is better than letting them find it at midnight before a deadline. It also teaches a useful habit, which is to describe your data before you draw conclusions from it.

Keeping projects honest

Sports data invites bold claims. A student who finds that a player wins more often after a long rally may be tempted to call it a skill. Remind the class that a pattern in one season can be noise. A claim has to hold up on data the student did not use to find it.

Splitting the seasons is the simplest way to teach this. Students build their analysis on earlier years and test it on the most recent one. If the pattern survives, it is worth writing up.

Where to start

Pick one question the class cares about, find a dataset that is documented and licensed for teaching, and build the course around answering it well. A single good question carries a term of work further than a dozen small exercises on data nobody minds about.

Describe a failure that, looking back, was actually one of the best things that ever happened to you.

View all responses