Description
Objective
In this lab, you will expand your skills in developing test cases.
Working alone or in a group of 2, you will create test cases for a program that creates a leader
board for a sports league.
You are not required to develop code for the test cases or to develop test input data for your test cases.
Preparation
You are going to help manage the leader board for a local sports league. The teams play against one
another and the leader board is created by ordering the teams from most winning to least winning.
Specifically, the first team on the leader board has the most game wins. If a tie happens between
teams then the order of tie-breaking, in order of precedence, is:
• most games won
• most games tied
• most points scored by the team
• highest difference of points scored to points lost
• most games played
• lexicographic order of the team names
You will be writing test cases for a class for this problem. The methods in the class are as follows:
public boolean addTeam(String teamName) Adds the given team name as one team in the league.
Returns true if the team was added to the league. Returns false if the team is already in the league or
if any problem arose with the addition.
public boolean recordGameOutcome(String team1, String team2, int scoreTeam1, int scoreTeam2)
Record that team1 played team2 and that the ending score was scoreTeam1 for team 1 and scoreTeam2
for team 2. Return true if the game was recorded. Return false if the game was not recorded or if any
problem arose with the recording.
1
public String createLeaderBoard() Create a string that is the leader board for the league. The
leader board lists all of the teams in the league in the order of most-successful team to least-successful
team, based on game outcomes, as described earlier. Each line of the leader board string is as follows:
• team name (use 15 columns)
• number of games won (use 2 columns)
• number of games lost (use 2 columns)
• number of games tied (use 2 columns)
• points scored by the team (use 4 columns)
• points scored against the team (use 4 columns)
Include one space between each column.
There will also be a header row, with titles Team, W, L, T, +, and -, all right justified except “Team”.
The league has space for a maximum of 24 teams.
Given the following data for recordGameOutcome (4 method invocations)
“Anchor”, “Salty”, 15, 12
“RC”, “Kitchen Party”, 9, 11
“Anchor”, “RC”, 12, 7
“Salty”, “Kitchen Party”, 6, 0
We get the following leader board
Team W L T + −
Anchor 2 0 0 27 19
S al t y 1 1 0 18 15
Kitchen Party 1 1 0 11 15
RC 0 2 0 16 23
Resources
• None required
Procedure
Set-up
1. Ensure that all in your group understand and agree upon the intention of the problem. Do not
make assumptions on how someone will implement a program to solve this problem.
2
Lab steps
Part 1 – Understand the problem
1. Clarify any ambiguities that your team may have about the problem. Record these clarifications.
2. Identify the boundary conditions that exist in the problem statement. Record these boundaries.
3. Outline the type of control flow that any implementation for the problem will need to follow,
based on the problem statement. Record these flows.
4. Define what would be perceived as the “normal” order in which methods would be invoked for
the problem.
Part 2 – Create test cases
1. Identify a set of input validation tests for each method as well as the expected outcome for each
case.
2. Identify a set of boundary tests for these boundary cases.
3. Identify a set of control flow tests.
4. Identify a set of data flow tests based on your “normal” order.
Questions
How, if at all, would input validation change if you were getting input from a user interface rather
than as parameters to methods?
Explain whether or not boundary cases exist beyond checking the incoming input values.
How does the idea of “control flow” change between black box tests and white box tests?
Should all permutations of methods result in data flow tests? Explain.
Reporting
1. In one file, list
• The members of your team.
• The clarifications and reporting from part 1.
• The tests from part 2, divided by test type.
• How complete you think that your test cases are.
• The answers from the Questions section of the lab.
2. Generate a PDF from the document.
3. Submit the PDF in Brightspace in the Lab/Lab 3 section of the course page in Brightspace.
3
Assessment
The assessment will be on a letter grade and will reflect how well you broke down the problem,
developed tests, and, most importantly, saw how to take the work done in this lab to broaden it to
other contexts (the Questions section).
Letter grade Problem clarifications Extent of the test
cases and classification (main weight)
Awareness of the capabilities of current
tests
Awareness on using
the test structures in
broader contexts
A
(Excellent) Problem is now clear
and good assumptions were made
Considerable nonoverlapping tests
made for all elements
of the program and
correctly classified
Demonstrates a good
sense for what the
tests and testing
approaches can do.
Knows how variations
happen
Clear idea of changes
to control and data
flows for testing
B
(Good) Confusing problem
aspects identified.
Not all with good
conclusions
Considerable tests
made for the whole
problem, though
some overlap, and
well classified
Demonstrates a good
sense for the testing
approaches. May not
appreciate the variations
Knows that the
control or data flow
changes might impact testing, but not
spelled out well
C
(Satisfactory) Main problem elements identified for
clarification
Generally well-tested
methods and classification is 75%
Has a reasonable
sense of how well the
testing is going
Appreciates that control and data flow
tests can change, but
unsure how
D
(Marginal Pass) Just a few problem elements were clarified
A small percentage of
tests made for most
methods. Unclear
if classification helped
guide them
Has a rough idea of
how well the testing
would be, but some of
the logic is flawed
Seeing little change in
the effects of the flows
F
(Inadequate) Lots of ambiguity
remains about the
problem
Poor testing of the
methods and/or no
structure to case creation
Little sense on how
well testing is happening
Unsure of even the
context of the question
4
