Clerical Operations curriculum

Every item can be settled with certainty by checking, so a miss comes from a slip, not from a hard idea. Settle on one way of checking each kind of item and use it every time. When you speed up and your misses climb, drop back to the last pace at which you were accurate.

Nine characters are hard to hold in mind at once, but three are easy. Read 7KD49TXR2 as 7KD, 49T, XR2 and compare one group at a time. Grouping also makes it obvious when a group has one character too many or too few, which a single long glance hides.

Moving your eyes back and forth between two codes makes it easy to land one character off, and an offset of one can make two different codes look alike. Point at the character you are reading in each code, or move the cursor along, and advance both together.

Agreement in the first several characters tells you nothing about the rest, because a changed character can sit at any position, the last one included. Run each comparison left to right all the way to the final character, and do not call a match just because the start looked right.

Two codes are either exactly alike or not, and one changed character makes them different. When the question only asks whether codes match, record the verdict and move on. The exception is a question about what kind of error a copy has: check the next character too, since a swap shows up as two differences side by side.

Finish your own count before you look at the options. Counting items tend to offer neighboring numbers, so a count that is off by one still lands on an option and gives no warning. If your count is not among the options, recount rather than picking the nearest one.

When each wrong choice changes one character, the only positions worth checking are the ones where the choices differ. Given FW38KD5N, FW38KB5N, FW36KD5N and FV38KD5N, the choices disagree only at the W, the 8 and the D. Check those three positions against the reference and the match stands out.

Instead of reading each choice in full, take the first group of the reference and compare it across all four choices, then the second group, and so on. You hold one short group in mind at a time, and each choice drops out the moment one of its groups fails.

Elimination is fast, but a slip can knock out the right answer and leave a wrong one standing. When one choice is left, read it against the reference once more from start to finish. If it fails that check, one of your earlier eliminations was a mistake, so go back.

When you line two codes up, it is easy to start at the second character because the first was taken in at a glance, or to stop one short of the end. A changed character is as likely to sit first or last as anywhere else, so make the first and last positions part of every check.

In these codes a changed character stays the same kind: a digit is replaced by another digit and a letter by another letter. So the confusions that matter are within a kind, such as 3 and 8, 1 and 7, M and N, or V and W. Saying each one in your head makes you decide what it is, rather than accepting a shape that only looks right.

Five pairs and six pairs look much alike in a list. Count them first. When you finish, the pairs you called alike and the pairs you called different must add up to that total. If they do not, you skipped a line or counted one twice.

Keeping a running count in your head while you also compare codes splits your attention, and that is where a tally slips. Put a tick or a cross beside each pair, or keep a count on your fingers, the moment you decide. Then count the marks at the end, when nothing else is competing for your attention.

A pair can be called different at its first mismatch, but it is alike only after a full check, so a difference is the quicker verdict to reach. If 2 of 6 pairs differ, then 4 are exactly alike. Read the question once more before you answer, to be sure whether it asks for the alike pairs or the different ones.

A pair can agree in eight characters out of nine and still count as different. The pull is to call a near-identical pair alike because it feels the same. Treat "exactly" literally, and give a pair the alike verdict only after checking it to the last character.

Reading the whole left code and then the whole right one asks you to remember nine characters. Instead, read the first three on the left, then the first three on the right, then the next three on each side. The comparison happens while the characters are still fresh.

Nothing guarantees that a list contains at least one difference, or at least one match. If every pair checks out, the answer is the full count, and if none does, it is zero. Expecting a mix tempts you to invent a difference that is not there.

The quickest check is the count. If the original has six characters and the copy has five, something was dropped, and you know that before you compare a single character. Equal lengths rule out a missing or extra character and leave a replaced or swapped one to look for.

When one character is copied as a different one, the codes match everywhere else. TK47MD copied as TK47ND has the same length, and only the fifth position differs. That is a plain error in the copy: not a swap, and not a missing character.

A transposition trades two adjacent characters, so both positions look wrong but the characters are all present. TK47MD copied as TK74MD has the same length and the same characters, with the 4 and the 7 traded. Check whether the two wrong positions hold each other's character before you call it a replaced character.

When a character is left out, every position after it looks wrong, even though the rest was copied correctly. TK47MD copied as TK4MD puts the M where the 7 belonged. A run of mismatches that starts at one point and continues to the end usually means a missing character.

When the options name the error, finding a difference is only half the job. Decide which kind it is: one character replaced, two neighbors swapped, or one character missing. Use the length check and the character check together, and choose the option that describes what you saw.

A copy that looks right at a glance is the most tempting place to stop early. Run the length check and then the full character check before you choose the match option. If you cannot point to the position that differs, the copy matches; if you can, it does not.

Names written surname first, such as Lindqvist, Rosa, put the filing key at the front. Compare surnames and ignore the given names. Lindqvist, Rosa files before Moreau, Albert, because L comes before M, even though A comes before R. The given name is used only to separate equal surnames.

Line two names up from the left and move along until a letter differs. Kettering and Kettle agree in K, E, T, T, then have E against L, so Kettering files first. Nothing after that point matters, so a name's length or its later letters cannot change the order.

Fenn and Fennimore agree for four letters, and then Fenn runs out. A name that has ended comes before any name that continues, so Fenn files first. The same holds for given names: Lou files before Louise. Treat the end of a name as earlier than every letter.

Corben and Corbin differ only in the fifth letter, E against I, so Corben files first. Names that share most of their letters are where a quick look goes wrong, because two names with the same outline pass for one. Slow down and compare letter by letter whenever two names in a set look almost identical.

Quill, Noel and Quill, Nora share a surname, so the given names decide. They agree in N and O, then have E against R, so Quill, Noel files first. Use the full test on the given names too, including the rule that a shorter given name files before a longer one it begins.

Sort the names by first letter, then spend your attention only on names that share one. A set often holds two or three names with the same initial, or even the same surname, so expect at least one close comparison. A set holding Garza, Lindqvist and two names beginning with Q needs only the two Q names compared closely.

The names are listed in a mixed order, so it is easy to answer with whatever sits in the asked position of the list instead of the filing order. Write the sorted order down, even as four initials, and count to first, second, third or fourth from there.

Finding the first name does not require sorting all four: scan for the earliest initial, and compare closely only if two names share it. The last works the same way from the other end. For second, find the two earliest names and take the later one; for third, find the two latest names and take the earlier one.

To count 4 just after 3, when that 3 follows an even digit, find every 4 first. In 234734634139 the 4s sit in three places. Each has a 3 before it, and before those 3s come 2, 7 and 6. The 7 is odd, so the count is 2. A named digit is one of ten, while even fits five, so anchor on the digit.

A pattern with a condition has several parts, and each must be true at the same spot. A 3 then a 4 is not enough if the digit before the 3 has the wrong parity. Check each part in order, and if one fails, discard that spot and move on rather than counting it as half a match.

Parity is a property of a single digit, so it can be read at a glance once it is automatic. The even digits are 0, 2, 4, 6 and 8. Zero is even because it divides by 2 with nothing left over, so a 0 meets an even condition and fails an odd one. Decide parity from memory, never by working it out.

A long string is easy to lose your place in, and a running count is easy to lose when you backtrack to check a condition. Put a tick under each confirmed match, or note its position. At the end, count the ticks; you never have to trust a number you carried in your head.

Reading a long string of digits in groups of four keeps your place. A three-digit pattern can still begin at the end of one group and finish in the next, and a chunk-by-chunk search misses it. After each group, look at its last two digits together with the start of the next.

In a long string the pieces of a pattern often appear without ever lining up correctly, and the right count is then zero. A string full of the right digits can tempt you to call a near miss a hit. Trust the check: if no spot passes every condition, choose zero.