Testing a lift

A lift (an elevator in a building) has a sign: maximum 8 people. How would you test it?

Try 3 people. Try exactly 8. Try 9. Try 0. What about a chair, or a dog? Each of these is a different kind of data for a test. The values you use to test are called test data (the data you test with). Good testers choose it carefully.

Key words

Word	What it means	Example
Test data	The values you give to a program to test it.
Valid data	Data the program should accept: the right kind and inside the limits.	55 for a mark of 0 to 100
Invalid data	The right kind of data, but outside the limits. The program should reject it.	150
Boundary data	Values on the edges of the limits, and the first values just outside them.	0, 100, -1, 101
Erroneous data	The wrong kind of data. The program should reject it.	"abc"
Test plan	A table that lists each test, its data and its expected result.
Expected result	What should happen, decided before the test.
Actual result	What really happened when you ran the test.

The cells students fill in are left blank in this preview.

Four kinds of test data

Kind	Example for a mark from 0 to 100	Expected result
Valid	55	Accepted
Invalid	150	Rejected: it is outside the limits
Boundary	0, 100, and the values just outside: -1 and 101	Accepted on the edge. Rejected just outside
Erroneous	"abc" or nothing at all	Rejected: it is not a whole number

The cells students fill in are left blank in this preview.

Why edges matter

Many bugs are at the edges of the limits. A programmer writes age > 13 when they should write age >= 13. Test data of 5 or 15 never finds it. Only 13 does. (With age > 13, the value 13 is wrongly rejected.)

So always test the lowest and highest valid values, and the first value outside each one.

Try it: run the test data

Code: python

def mark_ok(text):
    return text.isdigit() and int(text) >= 0 and int(text) < 100

for data in ["55", "0", "100", "101", "abc"]:
    print(data, mark_ok(data))

# One answer is wrong. Which data found the bug? What is the bug?

The test plan

A test plan is written before the testing starts. It says what you will test, why, with what data, and what should happen.

You fill in the last column during the test. If the actual result is not the expected result, the test fails and you have found a bug.

A test plan for a mark

Test	Purpose	Test data	Expected result	Actual result
1	Valid data in the middle is accepted	55	Accepted	Fill in when you run it
2	The lowest valid mark is accepted	0	Accepted	Fill in when you run it
3	The highest valid mark is accepted	100	Accepted	Fill in when you run it
4	A mark just above the top is rejected	101	Rejected	Fill in when you run it
5	A mark just below the bottom is rejected	-1	Rejected	Fill in when you run it
6	Text instead of a number is rejected	abc	Rejected	Fill in when you run it

The cells students fill in are left blank in this preview.