A Second Worked Schedule: Sound Arithmetic, Absent Tasks
The library's first schedule failed its own arithmetic. This one passes it — every duration, every roll-up — and then leaves out three tasks a research project cannot be run without. A sound plan of the wrong work is a different failure mode, and this is the worked instance.
What was decoded, and what was not
Stating this first, because it bounds everything that follows. The schedule was supplied as a project-scheduling application's binary file — an OLE compound document. No reader for the format was available in the environment where the extract was made, and no library capable of decoding it could be installed, so a container parser was written by hand.
WHAT THE FILE ITSELF ESTABLISHES, DECODED WITH CERTAINTY
| Item | Value |
|---|---|
| Document title | "Project Schedule" |
| Created / last saved | 29 December 2021, 08:47 / 09:18 — thirty-one minutes apart |
| Revisions | 7 |
| Calendars defined | One, named "Standard" |
| Resource data | Present, but no named resources |
| Scheduled start / finish (project property) | Mon 28/12/20 → Mon 31/07/23 |
| Scheduled duration (project property) | 676d? |
| Work / cost / % complete | 0h · 0.00 · 0% |
676 is the exact Monday-to-Friday working-day count from 28/12/2020 to 31/07/2023 inclusive. Verified. The question mark is the application's estimated-duration marker, uncleared on the project summary itself.
The plan as printed
Twenty-nine rows: one top summary, eight summary tasks and twenty sub-tasks, four of which are zero-day milestones. Durations, starts and finishes exactly as the chart prints them. The spellings are the source's.
THE FULL TASK LIST, TRANSCRIBED FROM THE PRINTED CHART
| WBS | Task name | Duration | Start | Finish |
|---|---|---|---|---|
| 1 | Research Project | 302 days | Mon 28/12/20 | Tue 22/02/22 |
| 1.1 | Developing the research title | 10 days | Mon 28/12/20 | Fri 08/01/21 |
| 1.1.1 | Understanding current issues | 10 days | Mon 28/12/20 | Fri 08/01/21 |
| 1.1.2 | Framing research title | 0 days | Fri 08/01/21 | Fri 08/01/21 |
| 1.2 | Developing research background | 25 days | Mon 11/01/21 | Fri 12/02/21 |
| 1.2.1 | Defining selected problem | 12 days | Mon 11/01/21 | Tue 26/01/21 |
| 1.2.2 | Decribing importance of proposed research | 15 days | Mon 11/01/21 | Fri 29/01/21 |
| 1.2.3 | Problem statement | 5 days | Mon 01/02/21 | Fri 05/02/21 |
| 1.2.4 | Literature gap | 5 days | Mon 08/02/21 | Fri 12/02/21 |
| 1.3 | Forming research aim and objectives | 22 days | Fri 12/02/21 | Tue 16/03/21 |
| 1.3.1 | Stating research aim and objectives | 0 days | Fri 12/02/21 | Fri 12/02/21 |
| 1.3.2 | Framing research question and hypotheses | 10 days | Mon 15/02/21 | Fri 26/02/21 |
| 1.3.3 | Outlining the expected benefits | 12 days | Mon 01/03/21 | Tue 16/03/21 |
| 1.4 | Defining the research method | 10 days | Wed 17/03/21 | Tue 30/03/21 |
| 1.4.1 | Choosing preferred research method | 10 days | Wed 17/03/21 | Tue 30/03/21 |
| 1.4.2 | Justifying research method | 0 days | Tue 30/03/21 | Tue 30/03/21 |
| 1.5 | Analysing research based on literature | 100 days | Wed 31/03/21 | Tue 17/08/21 |
| 1.5.1 | Searching all literature and case studies | 30 days | Wed 31/03/21 | Tue 11/05/21 |
| 1.5.2 | Identifying techniques to study literatures | 40 days | Wed 12/05/21 | Tue 06/07/21 |
| 1.5.3 | Analysing all literature studies | 30 days | Wed 07/07/21 | Tue 17/08/21 |
| 1.5.4 | Putting forward the final idea of research | 0 days | Tue 17/08/21 | Tue 17/08/21 |
| 1.6 | Methodology development | 62 days | Wed 18/08/21 | Thu 11/11/21 |
| 1.6.1 | Planning of data collection | 30 days | Wed 18/08/21 | Tue 28/09/21 |
| 1.6.2 | Planning of data analysis | 32 days | Wed 29/09/21 | Thu 11/11/21 |
| 1.7 | Data analysis | 68 days | Fri 12/11/21 | Tue 15/02/22 |
| 1.7.1 | Survey questionnaire and form development | 20 days | Fri 12/11/21 | Thu 09/12/21 |
| 1.7.2 | Survey analysis | 25 days | Fri 10/12/21 | Thu 13/01/22 |
| 1.7.3 | Secondary research | 23 days | Fri 14/01/22 | Tue 15/02/22 |
| 1.8 | Report development | 5 days | Wed 16/02/22 | Tue 22/02/22 |
Every duration and date belongs to this student's plan for this student's thesis. None is a benchmark, a norm or a recommended allowance.
The arithmetic, independently verified
Every date range above was recomputed against a Monday-to-Friday working calendar for the extract this page draws on. This is the part the library's other schedule failed, and this one passes it.
That matters because it makes the plan a calculation rather than a drawing. A schedule whose summaries roll up and whose durations follow a calendar can tell you something you did not already know — that a slip here moves a date there. A schedule whose numbers were typed cannot, and that is the condition the library's other example is in.
The two schedules side by side
The library now holds one defective schedule and one sound one, written by two different students for two different theses. The comparison is the teaching content, and it is worth reading whole. The first schedule is transcribed at a worked thesis schedule and its faults are worked through at what goes wrong in a thesis schedule.
THE LIBRARY'S TWO WORKED SCHEDULES COMPARED
| The first schedule | This schedule | |
|---|---|---|
| Rows | 25 (1 project summary + 24 tasks) | 29 visible (1 top summary + 8 summaries + 20 tasks) |
| Working calendar | None applied — weekend starts and finishes | Standard five-day calendar, applied consistently |
| Durations against date ranges | Do not match on at least four tasks | Match on 22 of 23 |
| Summary roll-up | Broken — children finish after parents | Correct in all eight cases |
| Milestones | None | Four zero-day milestones |
| WBS numbering | None | Full three-level breakdown |
| Uncleared estimated-duration markers | 17 of 24 | 1, on the project summary |
| Unexplained gap | 160 days with no task | None |
| Ethics task | Present, before data collection | Absent |
| Editing or revision task | Absent | Absent |
| Tracking | Not recorded | 0% throughout |
The single row both schedules share is the one that matters most for the write-up: neither allocates any time to revision.
Where the defects actually are
The arithmetic is sound, so the faults are all in what the plan contains. Six are demonstrable, and the first three are absences of work that this project demonstrably did.
Six defects in the plan
No ethics task
Nothing in the twenty-nine rows covers ethics approval, an application, or a human-research committee. Approval was in fact obtained: the project's participant information sheet names the committee and quotes an approval reference. The library's other schedule scheduled ethical approval as a discrete task before data collection — the one thing it clearly got right, and the thing this plan omits.
No data-collection task
Instrument development runs 20 days and survey analysis runs 25 days, and nothing sits between them for distributing the survey or waiting for responses. Fieldwork has no duration in this plan at all, which is the one part of a research schedule whose length you do not control.
Instrument design filed under analysis
"Survey questionnaire and form development" is a child of "Data analysis". Designing an instrument is not analysis, and putting it there leaves the plan with no design phase for the primary research.
No editing, revision or proofreading
The last row is "Report development", and it is the end of the plan. The subject's own writing-up guidance names three stages and says of the last that you should take as much time as is needed. The library's other schedule has the identical omission: two schedules, two students, the same gap.
Five days for a 62-page thesis
The plan allows five days to produce a document of 11,741 words. The thesis's own file metadata records 2,145 minutes — about thirty-six hours — of editing time, and a last modification ten days after the plan's finish date. The plan's final task and the document's own record of how it was made do not agree.
Nothing was ever tracked
0% complete, 0 hours of work, after seven saves. This is a plan and never a tracker, so it could not have shown a slip while there was still time to respond to one.
The plan against the artefacts it planned
This project is the only one in the library where a plan and the things it planned both survive — the full apparatus is walked through at a complete research project, end to end — so the comparison can be made from file dates rather than recollection.
PLANNED WINDOW AGAINST WHAT THE FILES RECORD
| The plan says | What the artefacts record |
|---|---|
| Questionnaire and form development, 12/11/21 → 09/12/21 | The questionnaire file is dated 19 Jan 2022 and the instrument was printed 7 Feb 2022 — both after the window |
| Survey analysis, 10/12/21 → 13/01/22 | The analysis workbook was created 27 Dec and last modified 29 Dec 2021 — inside the window |
| Secondary research, 14/01/22 → 15/02/22 | The thesis document was created 27 Jan 2022 — inside the window |
| Report development, 16/02/22 → 22/02/22 | The thesis was last modified 4 Mar 2022 — ten days after the planned finish |
These are observations about file save dates, not about anyone's working pattern, and must be read as such.
What to take to your own schedule
The supplied teaching material never teaches scheduling, names no tool and supplies no plan, so the checks below are this library's, assembled from what two student schedules did and failed to do.
Run these before you commit to a research plan
- Every artefact your method promises — instrument, approval, dataset, transcript, workbook, draft — has a task that produces it.
- Fieldwork has a duration, and it is longer than you want it to be. It is the phase whose length other people control.
- Ethics approval is a predecessor of data collection, not a parallel activity, and it appears on the chart by name.
- Instrument design sits in design, not in analysis.
- There is an editing task after the last writing task, and it is not the shortest task on the chart.
- Summary durations are calculated from children, not typed. Check one by hand: if it does not roll up, none of them do.
- Estimated-duration markers are cleared as estimates firm up, including on the project summary.
- The plan is opened and updated at least at each milestone, so a slip shows while you can still respond.
Every fault above is a fault in the schedule, and several are ordinary consequences of how a scheduling tool gets driven under time pressure — not judgements about a researcher whose thesis passed examination. Both of the library's schedules are single artefacts from single projects: two plans are not a sample, and nothing here is a claim about how research is generally scheduled. What makes the pair worth studying is that between them they show both failure modes — arithmetic that does not hold, and arithmetic that holds over the wrong list of work. The constructive version is at scheduling a research project.
What to carry forward
- Sound arithmetic is necessary and not sufficient. This plan calculates correctly over a task list missing three phases the project actually performed.
- Ethics approval, data collection and editing are the three tasks most often missing from a research schedule. Two student plans, and between them they omit all three.
- Fieldwork with no duration is the clearest sign a plan was written from a document structure rather than from the work.
- A plan nobody updates cannot warn you. This one was saved seven times and never showed a single per cent complete.
- Where a plan and its artefacts both survive, compare them: the file dates will tell you what the plan could not.
Frequently asked questions
Is this schedule a better model than the library's first one?
It is better built and it is not a better plan. Its calendar, roll-ups, milestones and breakdown numbering are all correct, which the first schedule's are not. But it omits ethics approval, data collection and editing, and the first schedule scheduled ethics approval properly. Take the mechanics from this one and the task list from neither.
Why does it matter that the file was only partly decoded?
Because it bounds what can be claimed. Per-task durations, dates, dependencies and outline levels were not read from the binary; the task table comes from the chart printed inside the thesis. The file-level properties were decoded with certainty, and one of them — the project finish — disagrees with that chart by seventeen months, with no explanation available in either artefact.
How long should I allow for writing up?
Nothing in this material supports a number, and the five days in this plan is that plan's own allowance rather than any kind of guide. What the evidence supports is a structural rule: writing and editing are separate tasks, the editing task comes last, and a plan whose final row is the production of the whole document has no capacity to absorb a bad draft.
What is the practical difference between a plan and a tracker?
A tracker records progress against the plan, so a slip becomes visible while there is still time to act. This file records 0% complete and zero hours of work after seven saves, so it could show only what was intended, never what was happening. Updating a schedule at each milestone is what converts one into the other.
Should every research schedule use a work breakdown structure?
This one shows what a three-level breakdown buys: every task has a parent, summaries roll up, and milestones mark the transitions. That structure is what made the arithmetic checkable from outside. It is not a requirement in anything the supplied material teaches, and it costs very little.
Can I generalise from two student schedules?
No. Two artefacts are not a sample, and both are visible in this detail only because unusual amounts of raw material survived. Use them as worked examples of two distinct failure modes, and check your own plan against the checklist rather than against these plans' numbers.
References and source attribution
- A complete research dataset supplied as student work: one examined master's thesis with its project schedule (a partly decoded scheduling-application binary, plus the Gantt chart printed inside the thesis), its analysis workbook, its raw response file, its survey instrument and its participant information sheet. Researcher, supervisor, institution, network user names and file paths scrubbed; software product names written generically. Used as observed practice, not as a model answer.
- A second student thesis schedule supplied with an earlier batch, transcribed and audited in this library as its first worked plan; used here only for comparison.
- The supplied teaching source: weekly study notes, slide decks and assessment activities for a master's-level research methods subject in project management, which requires a proposed timeline for the research and teaches no scheduling. Author, institution and year not stated in the supplied files.
Suggested questions for Ask KEVOS
- What tasks must appear on a research project schedule?
- How do I check whether my summary tasks roll up correctly from their children?
- How much time should I plan between finishing data collection and submitting?
- Where should ethics approval sit in a research schedule?
- How do I compare my plan against what actually happened?
- What is the difference between a plan and a tracker for a thesis project?