Portfolio evidence / 3 MIN READ
Reconcile conflicting GRC evidence without choosing a convenient record
Work through a fictional access-review chronology where a dashboard, ticket and export appear to disagree.
On this page
Two records can look inconsistent because they describe different populations or times. They can also expose a real unresolved exception. This original fictional exercise practices distinguishing those possibilities before writing a conclusion. It is not an assessment of a real system or a compliance determination.
Establish the question
The practice objective is narrow: did the supplied packet establish completion of the September review for all 40 listed privileged accounts by September 30? The deadline and population are scenario assumptions, not a universal regulatory requirement.
You receive three records:
| Record | What it says | Limitation |
|---|---|---|
| Dashboard captured October 2 | Review complete, 40 of 40 | Does not show individual completion dates |
| Ticket closed September 30 | 38 accounts reviewed; two awaiting owner response | Closing the ticket does not resolve the two accounts |
| Account export generated October 1 | 40 account identifiers | Shows membership, not review decisions |
The dashboard may reflect work completed after the deadline. Alternatively, the ticket narrative may be stale. The export cannot settle either explanation because it contains no review decision history.
Write a discrepancy record
Record the claim, the conflicting sources, the relevant period and the missing evidence. A useful note is: “Completion by September 30 is not established for two of the 40 accounts in the supplied packet. The October 2 dashboard reports current completion but does not establish the earlier state.”
Do not write “the control failed” from this packet alone. Also do not write “passed, 100 percent” merely because the newest screen is green. Both conclusions go beyond what the supplied records demonstrate.
Ask for evidence that distinguishes the explanations
Request the review event history for the two account identifiers, including decision timestamps, reviewer identity and the scope of each decision. Confirm that the dashboard and ticket refer to the same review cycle and population. Keep any real request within authorized access and data-handling procedures.
Now add a fictional response: one account was approved September 29; the second was approved October 1. Under the exercise's stated deadline, the packet now supports timely completion for 39 accounts and late completion for one. The later completion does not rewrite the September 30 state. Your conclusion should still identify the evidence's scope rather than generalize to every access control.
Preserve the reasoning trail
Keep both the initial and revised note, with the new evidence that caused the revision. Do not delete the discrepancy simply because the current dashboard is complete. For a public portfolio, use only synthetic identifiers and label the entire packet fictional.
NIST's definition of the examine assessment method describes inspecting and analyzing assessment objects to obtain evidence. It provides methodological context, not a prescribed outcome for this case.
Use the conflicting-evidence preparation comparison to choose between this bounded drill and connected portfolio practice. The GRC uncertainty guide covers explaining a bounded recommendation when evidence remains incomplete.
PUT IT INTO PRACTICE
Risk & Evidence Portfolio Kit
Connect the individual document to a complete fictional case, with registers and a versioned decision memo.