Reviews
Part of Five home coffee experiments any curious beginner can run alone
Why do unreported conditions ruin most home coffee comparisons?
A recording method for home coffee testing: what to write down, why unreported conditions ruin comparisons, and how to read your own notes honestly.
The difference between an experiment and a nice morning is the record. This is a method for keeping one that stays useful, and for reading it later without deceiving yourself.
What to take away
- Record the conditions, not only the verdict. Unreported conditions are why results cannot be compared.
- Duration and dose matter as much as the intervention itself.
- Small trials can produce real, specific effects, and they can also produce unwanted ones.
- Five fields a morning is enough. More and you will stop.
The five fields
- Coffee and roast date.
- The bag, and how old it was that morning.
- Dose and water.
- Both in grams. Not "a scoop" and not "a mug".
- Grind setting and brew time.
- The setting as marked, and the total time.
- Three words.
- Written before you think about it.
- One sentence.
- Written after, saying what was different from yesterday.
That takes under a minute. Anything more elaborate gets abandoned in a fortnight, which produces no record at all. Five words a morning for two weeks is enough to separate a fault from a bad day.
Five Fields To Record
- Coffee and roast date
- Dose and water in grams
- Grind setting and brew time
- Three words before thinking
- One sentence after
Why unreported conditions ruin everything
This is the central weakness of the coffee literature. It applies to your notebook too. A review of 26 human dietary intervention studies on coffee and oxidative stress reported that coffee can increase glutathione levels and improve protection against DNA damage, especially with regular intake.
But it noted high heterogeneity between studies in doses and duration, with several failing to report the polyphenol content of the coffee used.
Studies that do not report what they gave people cannot be compared with each other. Neither can your Tuesday and your Thursday if you did not write down the dose.
Duration is a variable, not a detail
A trial's length changes what it can detect. An 8-week randomized trial in 53 healthy adults compared pure arabica with an arabica-robusta blend at 450 to 600 ml of filtered coffee daily.
Both raised serum sirtuin-1 similarly. the blended coffee increased total cholesterol from 4.70 to 5.17 mmol per liter, while pure arabica did not.
Two things carry over. A specific, measurable difference can emerge from a fairly small trial when the intervention is controlled and sustained. And a change may appear in one measure while another stays flat, which is why recording several fields beats recording one verdict.
Reading your own notes
| What you see in the notes | What it usually means |
|---|---|
| Same words every day | You have stopped tasting and started copying |
| Wildly varying brew times | Grind or technique is drifting |
| Verdicts without conditions | The entry is unusable, skip it |
| A clear trend over five days | Worth acting on |
| A single dramatic morning | Worth repeating, not acting on |
The first row is the common failure. When the three words stop changing, take a week off recording and come back. Notes kept in one place are what make a pattern visible, and six pairings will tell you which roasters suit you.
The honesty rule
Write the three words before you look at what you wrote yesterday. Reading yesterday's note first turns today's tasting into confirmation. A week of that produces a record that proves whatever you already believed.
It is the same discipline that makes a structured experiment worth running at all, and it applies just as much to the more ambitious experiments that come later.
