Free mini-course · 55 minutes · No account needed
Review a project schedule from dependencies to a changed launch plan
Trace a small dependency network, explain float and critical paths, then evaluate a changed plan before making a delivery commitment. Save a checked schedule-review portfolio.
For a project coordinator who can add whole-day durations and read predecessor IDs. No scheduling software, account or coding is needed.
What you will make
A three-part schedule review containing an earliest-start walkthrough, a baseline critical-path analysis and a revised change recommendation with explicit resource assumptions.
- Calculate earliest start and finish times using the latest predecessor finish.
- Distinguish total float from critical work and evaluate independent baseline changes.
- Recompute a changed critical path and identify the resource confirmation needed before committing a date.
The exercise contract
All records, policies and outcomes in this course are fictional. The supplied contract determines the expected answers.
- Time is elapsed working days starting at 0. A task of duration 2 starting at 0 finishes at 2. No weekends or calendar dates are modeled.
- All links are finish-to-start with zero lag. A task starts only when every predecessor has finished; parallel branches are allowed.
- Assume unlimited parallel resources unless a question explicitly asks about the resource assumption. All durations are fixed fictional inputs, not estimates of a real team.
- Use each unit’s own baseline. For alternative changes within a unit, change only the named duration and reset all other durations first.
- A critical path is a longest start-to-finish dependency path in this fixed model. Total float is how much a task can slip without moving the model’s finish when other inputs stay fixed.
Use the latest predecessor finish instead of adding every duration.
1. Trace the earliest possible finish
A dependency is a condition for starting work, not merely a preferred order in a list. In this exercise a finish-to-start link means that the predecessor must finish before its successor begins. Write down the time boundary before doing arithmetic: elapsed day 0 is the start, and a two-day activity occupies the interval from 0 to 2. Counting both endpoints as whole extra days would create an off-by-one error.
Move through the network from predecessors to successors. A task with no predecessor starts at 0. For any other task, take the largest finish time among its predecessors as its earliest start, then add its duration for its earliest finish. At a join, the later branch controls the start. Adding the branch finish times would wrongly treat parallel work as consecutive work.
Elapsed duration and total effort are different measures. Two independent three-day tasks can run during the same three elapsed days if resources allow them to overlap. Adding every activity’s duration would instead describe a serial schedule. Here we deliberately ignore resource conflicts to isolate dependency arithmetic; that gives a clear model to inspect before introducing a real team’s staffing and calendars.
Maintain an audit trail with task ID, predecessors, duration, earliest start and earliest finish. If a stakeholder disputes the completion time, you can point to the join that controls it rather than arguing from a picture. A schedule with a missing predecessor may look faster while omitting required work, so the calculation is useful only after the listed dependency assumptions are accepted.
Try it yourself
Build a forward pass for this small network. Q and R are independent after P finishes.
Small dependency network
| activity id | activity | duration days | predecessors |
|---|---|---|---|
| P | Agree scope | 2 | |
| Q | Build prototype | 3 | P |
| R | Prepare test data | 4 | P |
| S | Integrate | 2 | Q R |
Numbers use the units in each label. Blank answers are incomplete. ID lists can use commas, spaces or semicolons.
Read this exercise’s solution
P runs 0–2, Q 2–5, R 2–6 and S 6–8. Q and R overlap. R controls the join at S; this is an eight-day elapsed plan under unlimited parallel resources, not an eleven-day serial plan.
Apply the idea
If someone adds a required review between R and S, which predecessor and duration fields must change before you reuse the eight-day result?
Explain which baseline branch controls launch and which changes consume slack.
2. Find float and test a launch proposal
Now use the separate seven-activity rollout network. Perform the forward pass again and retain every finish time. The launch depends on release review and training, so its start waits for the later of those branches. A critical path is a longest dependency chain that drives the finish in this model; it is not simply the task with the longest individual duration or the highest management priority.
To find total float, work backward from the baseline finish. The final task must finish by that boundary. For a predecessor with several successors, its latest allowed finish is the earliest latest-start among those successors. Subtract duration to find latest start, then subtract earliest start to get total float. This backward check helps distinguish a branch that can slip from one that immediately affects launch.
Float is conditional, not a free allowance that every task can spend independently. If successive tasks share the same slack, delaying both can consume more than the branch has available. The questions below change one duration at a time from the baseline. Reset the network between alternatives; stacking unrelated changes would no longer test the proposal described in the label.
A proposal to shorten noncritical training can reduce effort without moving the launch date. That may still have value, but it does not support an earlier delivery promise. Conversely, a delay on another branch may first consume float and only then move launch. Report both the revised finish and the reason, so a stakeholder can distinguish local task savings from an actual schedule gain.
Try it yourself
A fictional software rollout has seven activities. A stakeholder wants an earlier launch and proposes shortening training. First reconstruct the baseline dependency schedule, then evaluate a test-data delay and the proposed training reduction separately. Your deliverable is a schedule review with its critical path and a supported change recommendation.
Rollout activities
| activity id | activity | duration days | predecessors |
|---|---|---|---|
| A | Confirm requirements | 2 | |
| B | Prototype | 3 | A |
| C | Prepare test data | 2 | A |
| D | Build and verify | 4 | B C |
| E | Release review | 1 | D |
| F | Prepare training | 3 | B |
| G | Launch | 1 | E F |
Numbers use the units in each label. Blank answers are incomplete. ID lists can use commas, spaces or semicolons.
Read this exercise’s solution
Earliest finish times are A=2, B=5, C=4, D=9, E=10, F=8 and G=11. The critical path is A-B-D-E-G. C has one day of float and F two. A two-day increase in C’s duration pushes launch to day 12; reducing F by one day leaves launch at day 11. Resource availability could change this simple model and needs a separate check.
Apply the idea
If two tasks on the same branch each claim the branch’s two days of float, what must be recalculated before accepting both slips?
Evaluate a new duration baseline and compare two independent recovery ideas.
3. Recompute a changed critical path before committing
The final unit replaces the baseline: preparing test data now takes five days. The table below is the new planning version, with all other durations and links retained. Recalculate from its inputs rather than adding a delay directly to the old finish. A previously shorter branch can become the controlling branch, changing which improvement would affect the date.
First find the revised earliest times and a longest chain. Then evaluate two alternatives separately from that revised baseline: reduce B by one day, or reduce D by one day. An improvement to B can be absorbed by waiting for C, while a reduction to a controlling D can advance the final join. State which scenario a result belongs to; a list of dates without version labels is difficult to use safely.
The one-day D reduction is a hypothetical duration input, not evidence that the work can actually be compressed. Before proposing it operationally, confirm an achievable method, available people and retained acceptance checks. A fixed network does not price acceleration, quantify risk or authorize dropping quality work. Report the calculated gain as conditional on those confirmations.
Finally challenge the unlimited-resource assumption. If B and C need the same person and cannot overlap, the dependency-only result is insufficient even when every arithmetic answer is correct. Resource order, availability and calendars need to be added before a delivery commitment. Your handoff should name that missing confirmation rather than presenting the model’s finish as a guaranteed launch date.
Try it yourself
This revised baseline has C=5 days. For the B and D alternatives, change exactly one duration by −1 day and reset the others to this revised table. Choose the conclusion for a possible shared B/C resource.
Revised rollout baseline
| activity id | activity | duration days | predecessors |
|---|---|---|---|
| A | Confirm requirements | 2 | |
| B | Prototype | 3 | A |
| C | Prepare test data | 5 | A |
| D | Build and verify | 4 | B C |
| E | Release review | 1 | D |
| F | Prepare training | 3 | B |
| G | Launch | 1 | E F |
Numbers use the units in each label. Blank answers are incomplete. ID lists can use commas, spaces or semicolons.
Read this exercise’s solution
Revised baseline: A=2, B=5, C=7, D=11, E=12, F=8, G=13. Critical path A-C-D-E-G. Reducing B alone leaves finish 13; reducing D alone makes it 12. Verify D’s compression method and acceptance checks. If B/C share an exclusive person, plan their resource order and recalculate before a commitment.
Apply the idea
What would you ask to confirm before replacing a duration estimate with a shorter number in the schedule sent to the delivery team?
Save your three-part portfolio
Check all three exercises and include an explanation for each. You may download a portfolio with mismatches for later review; it clearly records which checks passed. Editing an answer or explanation invalidates that exercise’s previous check.
Complete and check all three exercises to prepare your portfolio.
Inspect the portfolio before downloading
Your portfolio will appear here after each exercise has been checked.
Download before leaving. Course answers stay in this page and are cleared on refresh; they are not sent to a server or added to the older course progress tracker.
Source and scope
Occupational task context comes from the O*NET® 31.0 Database, USDOL/ETA, under CC BY 4.0. View the occupation.
Task 16157 · 08/2019 · Monitor or track project milestones and deliverables.
Task 16170 · 08/2019 · Develop or update project plans for information technology projects including information such as project objectives, technologies, systems, information specifications, schedules, funding, and staffing.
Method reference: Microsoft Support: Manage your project’s critical path. Background for the relationship between the finish-driving task sequence, float and a critical path that can change. The calculations here use an original simplified network, not Microsoft Project settings.
Skill Workbench selected this context and authored the teaching, examples, practice rules, data and feedback. USDOL/ETA has not approved, endorsed or tested these additions. O*NET® is a trademark of USDOL/ETA. Matching the fixed checks does not certify occupational competence.