Evidence for Recommendations
Overview — Attaching the Proof
A recommendation is only as strong as what supports it. The Evidence section on a recommendation is where you gather the interview answers, survey comments, and observations that make the case — and mark which of them matter most.
It has three parts:
1. Search for Evidence — find supporting material across the whole activity and pull it in.
2. The evidence editor — mark each line as Normal, Important, or Critical.
3. Export — produce a PDF or Word document of the recommendation with its evidence.
Open a recommendation and expand Evidence.
|
GOOD TO KNOW Everything here is part of the recommendation itself. It autosaves with the rest of the page and needs Recommendations write access. |
Search for Evidence
Click Search for Evidence to open the Find Evidence dialog. It searches three sources at once across the activity:
· Interviews — interview and focus group answers
· Surveys — survey comments
· Observations — the activity's observations
Filtering the search
|
Filter |
What it does |
|
Source Type |
Tick any combination of Interviews, Surveys, and Observations. All three are on by default. |
|
Rating |
Tick Negative, Neutral, and/or Positive. |
|
Keyword |
Free text — e.g. procedure, compliance. Press Enter or the magnifier to run it. |
|
PRO TIP Start with Rating filtered to Negative and no keyword. That gives you everything critical anyone said anywhere in the assessment, which is usually a short enough list to read — and it surfaces supporting material you had forgotten was recorded. |
|
PRO TIP Then run it again with a keyword and all three ratings. Positive evidence matters too: a recommendation is much stronger when it can say "three departments described the same practice working well; the fourth does not do it". |
|
GOOD TO KNOW The search is scoped to the recommendation's ISF Trait and Attribute where they are set, so classifying the recommendation first gives you closer results. |
Reading the results
Results are listed with a summary line above them — the total, and a breakdown by source: N interviews, N surveys, N observations.
Each result shows:
· An icon for its source type
· Where it came from — the interview name, or the observation
· Its rating, as a coloured chip
· The text itself, quoted
· The question it answered, where there was one
|
PRO TIP Read the source line, not just the text. Three negative comments from one interview is one person's view; the same three from three interviews across two departments is a pattern. The counts at the top of the dialog tell you which you are looking at. |
Adding what you found
Tick the results you want — or use Select All and Clear — and click Add N to Evidence. They are appended to the recommendation's evidence, one per line.
|
HEADS UP The text is copied into the recommendation as it stood at that moment. It does not update if the original interview note or observation is later edited, and it does not disappear if the observation is deleted. That is deliberate — your evidence stays stable — but it means the wording in a recommendation can drift from the source. |
|
PRO TIP Add generously and prune afterwards. It is much easier to delete a line in the editor than to run the search again because you were too selective the first time. |
|
GOOD TO KNOW "No evidence found. Try broadening your filters or using different keywords." usually means the keyword is too specific, or all three Source Type boxes are not ticked. |
Marking What Matters
The evidence editor has two views, switched with a toggle.
Structured view
Each line of evidence is listed with a menu offering:
|
Level |
Use it for |
|
Normal |
Supporting material. The default. |
|
Important |
Evidence a reader should not skip. |
|
Critical |
The evidence the recommendation stands on. |
|
PRO TIP Be sparing with Critical. If eight of twelve lines are Critical, none of them is — you have just made the highlighting meaningless. Two or three Critical lines per recommendation is about right: the ones you would quote if you had thirty seconds to make the case. |
|
PRO TIP Use Edit text to trim a long quotation down to the sentence that matters, but do not change its sense. Trimming makes evidence readable; rewriting it makes it something you invented. |
|
HEADS UP Remove line deletes that evidence from the recommendation. There is no undo and no restore — this is not like deleting an observation. If you might want it back, mark it Normal instead. |
Raw view
The same field as plain text, with importance expressed as tags:
[CRITICAL]The night-shift crew leaves before the brief completes.[/CRITICAL]
[IMPORTANT]Three of four departments described the same gap.[/IMPORTANT]
Supporting detail with no tags is Normal.
|
GOOD TO KNOW Both views edit the same field. Use the structured view for normal work; the raw view is useful for pasting in a block of prepared text or fixing tagging quickly. |
|
HEADS UP Unclosed or mistyped tags are not validated. A stray |
Reading Evidence Back
When you view a saved recommendation, its evidence is displayed with a filter across the top: All (N), Critical (N), Important (N). Each line carries its level's icon and colour.
|
PRO TIP Filter to Critical and read the recommendation as a reviewer would. If the problem statement is not obviously supported by those two or three lines alone, either the evidence is not the right evidence or the problem statement is claiming more than you can show. |
Exporting a Recommendation
Evidence
|
Option |
Effect |
|
Include all evidence (Normal + Important + Critical) |
Everything. |
|
Show only key evidence |
Critical and Important only. |
|
Group evidence by importance level |
Critical first, then Important, then Normal. |
Highlighting
|
Option |
Effect |
|
Highlight Critical evidence |
Marks Critical lines in the output. |
|
Highlight Important evidence |
Marks Important lines in the output. |
Sections
Include or omit Problem Statement, Other Insights, and Consequences.
|
PRO TIP Two exports cover most needs. For the report pack: everything, grouped by importance, with both highlights on. For a briefing or a one-page summary: Show only key evidence, all three sections on. The second is usually the version people actually read. |
|
PRO TIP Export to Word when the recommendation is going into a larger document someone will edit, and PDF when it is going out as it stands. |
|
HEADS UP Evidence contains verbatim quotations and can carry identifying detail — a role, a shift, a location. Check what a filtered export actually contains before it leaves the assessment team. |
Evidence and the Report Builder
There are two routes to the same material, and they suit different moments:
|
Route |
Best when |
|
Search for Evidence in the recommendation |
You are writing the recommendation now and want the supporting lines attached to it. |
|
Evidence for Recommendations template in the Report Builder (admin) |
You want the whole activity's negative findings as a workbook to work through offline. |
|
PRO TIP On a large assessment, ask your administrator to run the Report Builder template first and circulate the workbook. The team can then agree which findings become recommendations before anyone starts attaching evidence one recommendation at a time. |
A Practical Approach
1. Classify the recommendation — Framework, Trait, Attribute — before searching.
2. Search broadly. Negative first, then keyword, then all ratings.
3. Add generously, then prune in the editor.
4. Mark two or three lines Critical, a handful Important, leave the rest Normal.
5. Filter to Critical and read it back. If the case is not made, get more evidence.
6. Export the key-evidence version for the people who will decide, and the full version for the pack.
Common Problems
|
What you see |
What it means |
|
No Evidence section |
You have Recommendations read access without write, or the activity is locked. |
|
Search for Evidence returns nothing |
The keyword is too specific, or Source Type boxes are unticked. Broaden and retry. |
|
Results all from one interview |
That is your finding, not a fault — the evidence is thinner than it looked. |
|
Add button disabled |
Nothing is selected. Tick at least one result. |
|
Evidence text differs from the current interview note |
Expected. Evidence is captured as text when added and does not follow later edits. |
|
Evidence from an observation that no longer exists |
Also expected. Deleting an observation does not remove evidence already added. |
|
A highlight is not appearing |
Malformed tags in raw view. Switch to structured view to see what was recognized. |
|
Removed a line by mistake |
There is no undo. Search for it again and re-add it. |
|
Everything is marked Critical |
Nothing is. Re-read and demote all but two or three. |
|
Export produced a very long document |
Use Show only key evidence, or turn off sections you do not need. |
|
Export failed |
Retry. If it recurs, reduce what you are including and try again. |






No comments to display
No comments to display