Observation & Memory curriculum

Starting to memorize the first line before you have seen the rest wastes the early seconds on one detail. Read the whole description once at an even pace, so you know how many details there are and what kinds. Then spend the remaining time fixing them. The first pass also shows you which details are unusual and need extra attention.

A short list with a generous timer lets you rehearse each detail several times; a long list with a tight timer does not. Count the details as you read, then split the time. With 12 seconds and six details you have about two seconds each, so you cannot linger on one. If you fall behind, keep moving and come back to the gaps at the end.

Descriptions rarely list details in a predictable order, so the list order is a poor memory cue. Pick one order for each kind of scene and file every detail into it as you read. A person goes top to bottom, a vehicle goes from the outside in. A fixed order turns recall into walking a familiar route instead of searching a jumble.

A question names a label, such as footwear or the direction someone left in, and asks for its value. If you stored only values, "green" and "northwest" float free and you cannot tell which question they answer. Rehearse pairs: feet, green sneakers; heading, northwest. The label is the hook the question will pull on.

You cannot rehearse long phrases in a few seconds. Keep the word that separates this value from the likely alternatives. For "a faded mustard poncho" the word is "mustard" if the question asks for a color, but "poncho" if it asks what was worn. Drop the filler, never the part that makes the detail unique.

Repeating only the newest detail gives the earlier ones no attention. Go back to the first slot in your fixed order and run through all of them, adding the newest at the end. In lab span tasks, people who rehearsed this way recalled more, but when one study trained people to do it, accuracy did not rise, so treat it as a habit to try rather than a proven gain.

Every option in a recall question is written to look plausible. If you read them first, a reasonable-sounding one can feel familiar only because you have just read it. Read the question, retrieve the value from your stored list, and only then look for it among the choices. If it is missing, your recall is wrong, not the choices.

A person description covers a predictable set of features: build, hair, anything on the head, the top layer of clothing, footwear, anything carried, a distinguishing mark and a direction of travel. Visit them in one order: build and head first, down to the feet, then the extras. A slot you reach with nothing in it tells you that detail was not given.

Build is usually stated as a height word and a weight word together, such as tall and lanky. Options often keep one half right and change the other, so "tall and stout" sits next to "tall and lanky". Rehearse both words. Remembering only that the person was tall leaves you a coin flip between the options that share it.

A hair value can pack several facts: length, color and a style such as curly or tied back. Wrong options are other hair descriptions, and one may share a part with the answer, such as the same length in a different color. Keep every part, or at least the one no other option shares. Hair and headwear are separate details, so a report can give one without the other.

When a description has several colored things, the colors are easy to remember and easy to misplace. A purple scarf and gray boots can turn into gray scarf and purple boots under pressure. Rehearse the color and the object as one unit, "purple-scarf", never as a loose list of colors. A question about the boots needs the boots' color, not any color you saw.

The mark slot holds whatever sets the person apart: a scar, a tattoo, a birthmark, something worn on the face, or a way of walking. Wrong options are usually other kinds of mark, so the kind matters most, but note the place on the body too, such as a tattoo on the wrist. Give it its own rehearsal: clothes can change, but a mark usually stays with the person.

A description can state an absence: carrying nothing, no hat, no visible damage. These are easy to drop because they feel empty, but a question can ask for them exactly like any other value. Store the absence with its label, "hands: nothing", so you do not later fill the slot with a plausible object you never saw.

"Left heading northwest past the bakery" carries two facts, a compass point and a place, and they travel together. Picture the person moving toward the landmark: the image holds both halves, and when the options come, either half you kept can find the right one. Storing the word "northwest" alone works only if no other option shares it.

A vehicle description has its own standard set of details. The first three, color, type and make, are what you would see from across a street, so file them first and the closer details after. Using the same pass every time means a question about any one detail sends you to a known place in your list.

"A Mazda" is a make and "a station wagon" is a type. Under time pressure the two merge into a single image, and a question about the make can pull out the body style instead. Rehearse them with their labels: make, Mazda; type, station wagon. Check which one the question asks for before you answer.

A plate fragment is only a few characters, and the wrong options are other short runs that also mix letters and digits, so a vague sense of "a letter and a number" will not choose between them. Say the characters one at a time, in order: "five, S". Note which is the letter and which is the digit, and spend a little more of your time budget here.

"A scraped passenger door" names both the damage and the part of the vehicle it is on. Picture the scrape on the door: the part is often the easier half to keep, and either half can pick out the right option. If the description says there is no visible damage, store that as the value rather than leaving the slot empty.

One-word values like northeast or five people take no effort to read, so they are easy to skim past without rehearsing. Attach them to something: the car drove off northeast, toward the hills; five people, a full car. A small image costs a second and keeps the detail from fading.

A building report covers how tall the building is and what it is used for, where smoke shows and its color, what stands next to it, any hazard, and the water supply. File the details in that order: the building itself, what the fire is doing, what it could reach, and what the crew needs to know. The same route each time makes every detail findable.

"Pale blue smoke from the basement" holds two separate details, and a question asks for only one: where the smoke showed or what color it was. Rehearse both words with their labels. Picture the building with smoke coming from one specific spot in one specific color, so a single image answers either question and a bare color option can be checked against it.

On the fireground, an exposure is nearby property the fire threatens; in national incident reporting the word is narrower and covers only property that actually caught fire. A report may say what stands to the left, the right or across the street. Sketch the building in your mind and set the neighbor on its side, so the question becomes a look at your picture.

"A cistern behind the church" and "a cistern a block away" share the word cistern; the distance is the actual detail. The same holds for a height given as a number of floors. Options often repeat the noun and change the qualifier, so rehearse the qualifier: cistern, behind. Convert a height to a digit, five floors, which is quicker to say.

A hazard is the detail that changes what people should do, such as a gas meter on the wall or a chained gate. Turn it into a short warning in your head: "watch the gas meter". A warning is more memorable than a list item because it carries a reason, and it reminds you to treat the hazard as a separate detail from the building itself.

If the question asks the color of a scarf, every option will be a scarf color, and every option could have been true. Reasoning about which is likely does not help. The answer comes only from what you stored. That is why encoding during the study time matters more than cleverness at the question: there is nothing in the options to work out.

If the footwear will not come to mind, start from the top of your order: build, hair, head, jacket, then feet. Recalling the details around a slot often brings back the one in it, because you rehearsed them together. This only works if you used the same order while studying, which is the main reason to choose one and keep it.

A question asking which detail was NOT in the description flips the task. Three of the options are details you saw, and one was never mentioned. Rather than hunting for an absence, go through the options and mark each one you can place in your stored list. The one left over is your answer.

The missing option is often a different value of a kind the description did include. If the person had short red hair, an option offering curly gray hair belongs to a category you remember seeing, which makes it feel present. Check the value, not the category: hair was mentioned, but not this hair. Matching the label and stopping is the mistake this kind of option invites.

An option may read only "teal" or "a guitar case". Before you judge it, ask which label it would belong to: teal could be a jacket or a car; a guitar case is something carried. Then check that slot in your list. Without the label, a bare color can seem familiar because some other item in the scene had a similar color.

Not every description includes every slot. If a person report gave no footwear at all, an option naming shoes is absent for a simpler reason. Knowing which slots were filled is part of what you store. When you walk your fixed order, note the empty slots too, so a question can be answered by "that kind of detail was never given".

Start with any option you can place with confidence; each one removes a candidate. If two options remain, compare them against your list one at a time, checking the exact wording. Guessing between the last two is better odds than guessing among four, and the confident marks keep you from talking yourself out of a detail you really saw.

A statement in sentences hides its details among connecting words. As you read, pull each fact into a slot: who, what they wore, what they did, which way they went, what vehicle. A sentence such as "a cyclist in a purple jersey took the parcel and rode toward the bridge" holds four slots. Rehearse the slots, not the sentence.

With two or three people in a statement, the actions are easy to remember and hard to assign. Tag each person with one feature, such as "jersey" and "tall one", and rehearse each action with its tag: jersey took the parcel; tall one held the door. A question about who did something then has a tag to land on.

Questions may ask what someone did after another event, or which happened first. Keep a short numbered sequence as you read: one, the taxi stopped; two, a man got out; three, a crate was left. Order is a separate fact from the events themselves, so a list of events without numbers often fails these questions even when every event is remembered.

Procedure passages are lists in disguise, and questions about them often ask for a position: the second step, or what comes before a named step. Count the steps as you read and rehearse them with their numbers. Watch for words that mark order, such as first, then, before and only after, because they set the position even when the sentence order differs.

A passage might say a drill lasts 12 minutes or that a team checks the dials every 3 hours. A bare 12 or 3 is easy to misplace or swap with another number in the text. Say the number with its unit and its subject: drill, 12 minutes. Options can keep the number and change the unit, or keep the unit and nudge the number.

An image is hard to rehearse once it is hidden. Describe it to yourself in a few words while you look: three benches, a newsstand, a lamp post. The words give you something to rehearse after the picture is gone, and they force you to notice details you would otherwise only glance at.

A quick glance gives an impression, not a number. Split the picture into rows, sides or shelves and count each part, then add: four porch lights on the left, three on the right, seven in all. Keep the part counts too, because a question may ask about one part alone rather than the total.

A picture of a row of houses may show some porch lights on and some off. A question might ask how many are lit or how many there are in all, and you will not know which in advance. Count both while the picture is visible: nine houses, four lit. It costs a few seconds and covers either question.

A face looks like a whole, which makes it hard to recall in parts. Go from the top: hair, eyes, nose, mouth, ears, facial hair, then anything unusual such as a mark. Name each feature briefly as you pass it. A question about one feature then has a named detail to retrieve instead of a general impression.

Simple list spans average about seven items, while tasks that block grouping and rehearsal average about four. One account is that the seven is really about four chunks of several items each, and one study found chunking helps in both kinds of task. A description of eight details is at or past that limit, so group details, such as hair and headwear as one "head" unit.

Memory for words improves when they form familiar groups. In one series of experiments, familiar word pairs seemed to help rebuild fading traces, while longer familiar groups seemed to be stored as single units. Knowing which kinds of details a report usually holds puts the same idea to work, though the studies used word lists: new details fill known slots.

When a list is heard or read one item at a time, people recall the first and last items better than the middle ones, a U-shaped pattern called the serial position effect. It is usually explained by the first items getting more rehearsal and the last still being fresh. Apply it to your own fixed order: give the middle slots extra passes.

Proactive interference builds when you study several lists of similar items, and recall of the later ones drops. A run of person descriptions is that kind of series. In one study, students given 25 minutes of instruction in placing words in imagined locations showed less of this buildup. Picturing each scene somewhere specific is a way to try the same idea.

The production effect is a recall difference between words read aloud and words read silently, mostly measured with word lists. It shows up most clearly when only some words are spoken; in one study where most of a list was read aloud, the silent words were recalled better. Part of the gap is a cost to the silent words beside the spoken ones, so treat it as a way to make a few key details stand out, not a general boost.

Many studies find that retrieving information strengthens memory more than studying it again. Most used college students, and one series found the same benefit in 10-year-olds recalling words from a category cue. In a study window, that suggests looking away to recall each value from its label, then checking, rather than reading the lines again.

In eyewitness research, misinformation repeated several times sways later memory more than misinformation seen once, and the authors of one such study read this as people answering from how easily a detail comes to mind. An option can feel familiar because you met it in earlier questions, not in this scene. Place a detail in this description before trusting the feeling.