Reading an address in blocks

Break the address into equal blocks, treat each block as a separate small comparison, and do the blocks in a fixed order. That is all chunking is, and it is the only manual technique on this board that reliably improves a by-eye comparison.

It works because attention has a shape. It peaks at the start and end of whatever you consider to be the current task. Making eight small tasks out of one large one gives you sixteen attention peaks instead of two.

Every block metric on this board cuts an address into seven equal pieces and counts something inside each. The first two tiles do that for digits, on two different addresses.

Look at them side by side. They are different lines with nothing in common, which is the intended lesson: the blocks are a tool for the reader, not a feature of the string. There is no natural place to divide an address, no meaningful boundary, and no block that is more important than another.

The last tile counts distinct symbols per block, and it is almost flat. That flatness is the reassuring part. It means no block is unusually repetitive or unusually varied, so a fixed block size is as good as any clever one.

• Small enough to hold in your head at once. Four to eight characters is the usual range for most people.
• Consistent between the two things you are comparing. Different block boundaries on the two copies defeats the whole method.
• Counted, not eyeballed. If you find blocks by looking for landmarks, you are back to landmark reading and the blocks will not line up.
• The same every time you do this. A routine that changes is a routine you cannot tell you have done correctly.

It does not make a manual comparison equivalent to copying. It reduces the error rate of a bad procedure, which still leaves you with a procedure that has an error rate.

It does not help at all if one of the two copies is wrong at the source. Chunking is a comparison aid, so it inherits every weakness of the thing you are comparing against.

And it does not help with the two regions that are guaranteed to match. Blocks covering the shared opening will always agree, so the first block or so is doing no work regardless of how carefully you read it.

Copy the address. Use the button. That is the whole recommendation and it is what this board says on the copying panel as well.

Chunking is what you do when copying is unavailable: when the address is on a screen you cannot select from, on paper, or in a photograph. Those situations are real and worth having a technique for, and this is the technique. It is a fallback, and treating a fallback as the standard method is how people end up doing careful work in the wrong place.

The physical version is simple and worth writing down. Put the two copies one above the other in a fixed width font, cover everything except the current block with two pieces of card, compare, then move both pieces of card together to the next block.

Covering matters more than it sounds. An uncovered string invites the eye to wander ahead, and wandering ahead is how you end up comparing a block against the wrong block. Physically hiding the rest removes the temptation instead of resisting it.

Mark off blocks as you complete them. A comparison you abandoned halfway and a comparison you finished look identical afterwards, and the only way to tell them apart is a record you made at the time.

Why seven blocks in the tiles?

Seven divides the address length evenly, which keeps every block the same size. Nothing about the number seven matters beyond that.

Does reading blocks backwards help?

It breaks the habit of skimming forward, which helps some people. It also makes losing your place easier, so it is worth trying rather than adopting.