A technical report presents its conclusions in a uniform voice. Each is a sentence of similar length, in the same typeface, under the same heading. Read at speed, they all carry the same authority.
They should not. Behind those sentences sit different kinds of evidence, and the difference decides how much weight a conclusion can take. A reader who cannot tell them apart is relying on the author’s private judgement about which conclusions are firm, rather than forming their own.
Five kinds of basis
Almost every statement in a building report rests on one of five things.
- Observed. Seen by the author during inspection. The strongest basis available for a physical condition, and still limited to what was accessible on the day.
- Documented. Recorded in a document the author has reviewed. Strong evidence of what was intended or recorded, weaker evidence of what exists now. A drawing shows a design. It does not confirm that the design was built.
- Reported. Stated by someone else and not independently checked. This covers information from an owner, a managing agent, a contractor, a previous consultant or a vendor’s data room. It may be entirely accurate. It has not been tested.
- Assumed. Adopted where information was not available. Assumptions are legitimate and often necessary. They become a problem only when they are not identified, because an unlabelled assumption reads as a finding.
- Not verified. Could not be confirmed within the scope. This is a finding in its own right rather than an admission. Knowing what remains unknown is part of knowing where you stand.
Why the distinction is commercial, not academic
Consider a report that concludes: the compartment walls between the residential floors and the basement car park are of masonry construction and separate the two.
If that is observed, someone gained access above the ceiling line and saw the construction. If it is documented, someone read a drawing. If it is reported, someone was told. If it is assumed, someone inferred it from the age and type of the building. The sentence in the report may be word for word the same in all four cases. The decision it can support is not.
An owner asked to sign off a remediation scope, a fund asked to accept a building at a price, and a lawyer asked to rely on a technical position all need to know which of the five they are working from. That is not a matter of doubting the author. It is a matter of knowing which parts of the position would move if someone lifted a ceiling tile.
Where assumptions hide
Assumptions rarely announce themselves. They tend to sit in four places.
- Extrapolation from a sample. Ten doors inspected out of ninety, with a conclusion written about ninety. The sample may be entirely reasonable. The conclusion is still an assumption about the eighty that were not inspected, and the report should say how the sample was chosen and what it is taken to represent.
- Photographs. A photograph is evidence of what was in front of the camera. A caption that says what it means is interpretation. Both belong in a report, and the reader should be able to tell them apart.
- Quantities and costs. A cost is a rate applied to a quantity. Where the quantity was estimated rather than measured, the cost inherits that uncertainty, and an accuracy range should say so.
- Age and type. “A building of this age and construction will typically have” is one of the most common assumptions in building reporting, and often a sound one. It is still an assumption.
Qualified is not the same as vague
Qualified language is often read as an author protecting themselves. Sometimes it is. More often it is doing necessary work, and the difference lies in whether the qualification is specific.
“It appears that the cavity is unventilated” is vague. “The cavity could not be inspected. There are no visible ventilators on the elevations, which indicates it is unventilated and would be consistent with the construction date. Opening up at two positions would confirm it” is qualified and specific. The second tells the reader exactly what the author knows, how they know it, and what would settle the point. The first tells the reader only that the author is uncertain.
Measured confidence means being firm where the evidence supports it and qualified where it does not, and being specific in both directions. A report that qualifies everything to the same degree has not done the work of distinguishing.
The limitations section is part of the findings
Limitations are usually read last, if at all, and are often written as protection rather than as technical information. They deserve better from author and reader alike, because they define the shape of what the report can support.
The useful test is whether the section says what was not done and what that means. “No opening up was undertaken” is a statement of scope. “No opening up was undertaken, so the external wall build-up has been taken from the drawings listed in Appendix B and has not been confirmed on site” is a statement of scope and its consequence. Only the second allows a reader to judge reliance.
What to do when the basis is not stated
Many reports do not label their evidence. That does not make them unusable, but it does move the work to the reader. Three questions usually recover the position.
- Which conclusions were seen, and by whom? Ask for the inspection record: what was accessed, what was opened up, and what was not reached.
- What information was provided rather than found? Ask for the document schedule, and note which conclusions depend on it.
- What would change if a particular assumption were wrong? This is the question that finds the assumptions that matter. Most are immaterial. A few carry the conclusion, and those are the ones worth testing.
An author who can answer those three questions quickly has usually written a sound report with a presentation problem. An author who cannot is a different matter.
The point of the discipline
Labelling evidence is not a stylistic preference. It is how a technical report stays useful to two readers at once: the specialist who wants to test the reasoning, and the decision-maker who needs to know what the findings mean and how much weight to put on them.
It has a quieter benefit as well. An author who has to record the basis of each conclusion discovers, while writing, which conclusions are thinner than they felt on site. That is a better thing to find out before a report is issued than after.
EMTED works from evidence and states the basis of every material statement, using the labels observed, documented, reported, assumed and not verified. We help clients understand what the reports in front of them do and do not establish, and what to ask their authors.