How to Think

This module isn't about aquariums. It's about how to think through any problem: a car that won't start, a phone that won't charge, a leak under the sink. Every later module uses the same method, so learn it here with everyday examples first.

The method is a loop. Observe what's really happening. Check the facts before you believe them. List every cause it could be. Rule causes out with evidence. Act, starting with the safest useful step. Then check whether it worked. If it didn't, go round again with what you've learned.

The goal isn't to remember the answer. It's to be able to work out the answer when nobody has told you.

1. 1. Observe
2. 2. Check the facts
3. 3. List causes
4. 4. Rule them out
5. 5. Act
6. 6. Check

An observation is what you can see, hear, smell or measure: "there's water on the floor next to the dishwasher." A conclusion is what you think it means: "the dishwasher is leaking." Conclusions feel like facts, but they're guesses until you check them.

Good observers look at the whole picture, not just the first thing that catches their eye. Where exactly is the water? How much? Is it clean or dirty? Is it still coming? When did it start? Each detail rules some causes in and others out.

Attention to detail is a skill, not a personality type. It means reading the whole label, noticing the unit (grams or kilograms?), and noticing what's different from yesterday. Small details are often the whole answer.

1. 1. Observe
2. 2. Check the facts
3. 3. List causes
4. 4. Rule them out
5. 5. Act
6. 6. Check

Mopping the floor fixes the symptom. If the cause is a dripping pipe, the water is back tomorrow. Taking a painkiller for a toothache helps today, but the tooth still needs a dentist. Covering a warning light on a car dashboard doesn't fix the engine.

One symptom can have many causes, and one cause can make many symptoms. So whenever you fix a symptom, ask: what caused this, and is it still happening?

Sometimes you have to treat the symptom first because it's dangerous right now: you mop up water near a power socket before you go looking for the leak. That's fine. Just don't stop there.

1. 1. Observe
2. 2. Check the facts
3. 3. List causes
4. 4. Rule them out
5. 5. Act
6. 6. Check

Ask three questions about any fact you're about to act on. Where did it come from? (Did you see it, or did someone tell you?) Could the tool or the person be wrong? (Was the thermometer just in hot tea? Was the scale set to zero?) Does it fit everything else you can see? (A "fever" in a child who's running around happily doesn't fit.)

If a fact doesn't fit, check it again before you do anything. A careful second look costs a minute. Acting on a wrong fact can cost much more.

1. 1. Observe
2. 2. Check the facts
3. 3. List causes
4. 4. Rule them out
5. 5. Act
6. 6. Check

This is called process of elimination. A phone won't charge. It could be the socket, the charger block, the cable or the phone. Test each one on its own: plug a lamp into the socket, try the charger with another phone, try a different cable. The part that fails its own test is the cause.

Order your checks sensibly. Check the most dangerous causes first (a burning smell from the charger comes before anything else), then the most likely ones, then the quickest and cheapest to check. Cross each one off only when you have evidence, not because it "probably isn't that".

Keep track of what you've ruled out and how. Write it down. The next person, or you tomorrow, shouldn't have to check it all again.

1. 1. Observe
2. 2. Check the facts
3. 3. List causes
4. 4. Rule them out
5. 5. Act
6. 6. Check

A new cleaning product, a different brand of food, a moved piece of furniture, a new person doing a job, a power cut, a hot day. Any change just before a problem is a suspect. Look back a day, a week, sometimes longer, and list every change you can find.

A change is a suspect, not a conviction. The cat may have caught a cold that happens to line up with the new spray. Test it: stop using the spray. If the sneezing stops, and comes back if the spray comes back, you've found your cause.

1. 1. Observe
2. 2. Check the facts
3. 3. List causes
4. 4. Rule them out
5. 5. Act
6. 6. Check

When two things rise and fall together, there are several possibilities. A causes B. B causes A. Something else causes both. Or it's coincidence. Ice cream sales and sunburns rise together because hot sunny weather causes both, not because ice cream burns your skin.

Before you decide one thing caused another, ask: could something else explain both? And can you test it? If removing the suspected cause removes the effect, and bringing it back brings the effect back, you have much stronger evidence.

1. 1. Observe
2. 2. Check the facts
3. 3. List causes
4. 4. Rule them out
5. 5. Act
6. 6. Check

When something isn't working and there's no emergency, change one thing, wait long enough to see the effect, and check. Then decide the next step. It feels slower, but it's faster in the end, because you learn something from every step.

Overreacting is the other half of this. Someone turns the shower to hot, doesn't wait, turns it far to cold, doesn't wait, and swings back and forth. Every system takes time to respond. Make a small change, wait, check, then adjust. Doing too much, too often, or doing things that cancel each other out usually makes problems worse.

The exception is an emergency. If something is dangerous right now, do what makes it safe, even if that means doing several things at once. Then go back to one change at a time.

1. 1. Observe
2. 2. Check the facts
3. 3. List causes
4. 4. Rule them out
5. 5. Act
6. 6. Check

This is called triage. A pipe bursts under the sink and water is spreading toward a power strip. First make it safe: switch off the power at the wall if you can do it safely. Then stop the damage: turn off the water. Then find the cause. The mopping comes last.

When you're not sure what to do first, ask: what happens if I leave this for ten minutes? Anything that gets dangerous or much worse in ten minutes comes first.

1. 1. Observe
2. 2. Check the facts
3. 3. List causes
4. 4. Rule them out
5. 5. Act
6. 6. Check

Before you act, decide two things. What will tell me it worked? (The floor stays dry. The phone shows it's charging.) When will I check? (In an hour. Tomorrow morning.) Then actually check, at that time.

If it worked, write down what the problem was, what caused it, and what fixed it. If it didn't, that's useful too: you've ruled something out. Go back round the loop with that new fact.

Some fixes come with their own risks. Pick the step that helps without being able to make things worse, when there's one. Turning off a dripping tap can't hurt. Taking apart the plumbing could.

1. 1. Observe
2. 2. Check the facts
3. 3. List causes
4. 4. Rule them out
5. 5. Act
6. 6. Check

"We've always done it this way" isn't a reason. It may be a good way, but you should know why it's done that way. If nobody knows why, find out.

1. 1. Observe
2. 2. Check the facts
3. 3. List causes
4. 4. Rule them out
5. 5. Act
6. 6. Check

Get help when the problem is dangerous and you don't know how to make it safe, when you've run out of causes to check, or when the next step could cause serious damage if you're wrong. Asking early is better than guessing late.

When you report a problem, say: what you saw (facts, with numbers and times), what you checked and what you ruled out, what you did, and what happened after. "The floor by the dishwasher was wet at 9:00, about a cup of clean water. The hose is dry. There's a slow drip from the pipe behind it. I turned off the valve and put a towel down. Still dry at 9:30." That report lets anyone carry on where you stopped.

Write things down. A problem you solved and recorded only has to be solved once.