First Day of AP Statistics: A Complete Plan
By Jude Wallis · Published
Run the first ten minutes as a two-question class survey that puts one categorical and one quantitative variable on the board, not a syllabus read-aloud. Read the numbers back before naming vocabulary, then point out the class is a convenience sample, not a random sample of anyone beyond the room.
AP Statistics: Unit 1 (topics 1.1 Introducing Statistics: What Can We Learn from Data?, 1.2 Variables, 1.10 The Investigative Question Revisited and Data Collection). Day one previews Practice 1, formulating a question worth asking, and Practice 2, collecting data to answer it, informally before either is named. It also gets ahead of topic 1.2's categorical-versus-quantitative distinction and topic 1.10's investigative-question framing by having the class produce and classify its own data before either topic is taught directly.
Minutes 0 to 10: the two-question survey
Materials: index cards or scrap paper, or just the whiteboard. No devices required.
Before students settle in, hand out one card per student, or point at two columns already sketched on the board. Ask two questions and nothing else. Question one, quantitative: to the nearest half hour, how many hours did you sleep last night? Question two, categorical: how did you get to school this morning, choosing from walk, drive yourself, get a ride, or bus.
Collect both answers as fast as the room can move. The fastest version: students write both answers on one card and stick it to the board under columns you have already drawn, sleep hours as a running number line, transportation as four labeled bins. Take attendance while the cards go up rather than as a separate step; nobody notices, and the period never stalls for roll call.
By minute ten the board holds two real datasets nobody had to invent: one column of numbers and one set of four counts, generated entirely by the people sitting in the room.
Minutes 10 to 20: read the numbers before you name anything
Do not introduce a single vocabulary word yet. Just look at the board and ask what the class notices.
On the sleep-hours side, someone will point out a cluster, most answers sitting close together, and probably one point sitting far from the rest, a student who slept two hours or eleven. Ask what a fair single number would be if you had to describe the whole board with one value. Whatever the class proposes, average, middle, most common, all three are reasonable answers and the disagreement between them is the entire subject of the year in miniature.
On the transportation side, ask which bin is tallest and why that might be true of this particular school. A bar graph and a dotplot side by side make the difference visible before either is named: one is counting categories, the other is placing numbers on a line, and the histogram vs bar graph mix-up, one that catches many students later in the unit, is sitting on the board already correctly built.
Now attach the words. Sleep hours is a quantitative variable; how a student got to school is a categorical variable. The students in the room are a sample. A number that summarizes the sample, such as the most common sleep total, is a statistic. Four words, and every one of them was pointed at before it was defined.
If a projector happens to be available, this is the moment to open the descriptive statistics sandbox and click points onto its dotplot to rebuild the shape the board just built by hand, then hand a student the controls to drag the outlier and watch the mean move while the median holds still. It takes under two minutes and it is entirely optional; the plan works without a screen in the room.
Minutes 20 to 28: why this room is not a random sample of anything
This is the question that makes the day distinctive rather than generic. Ask directly: can we say anything from this board about how much sleep teenagers get in general? The honest answer is no, and working out why is worth eight minutes on day one.
The students in the room are whoever happens to be enrolled in this class period. That is a convenience sample: easy to reach, not chosen at random from any larger group. The board describes this room and nothing beyond it with any confidence. Making a real claim about the whole school's sleep would require selecting students from the whole school at random, a specific procedure the course teaches properly rather than an assumption anyone gets to make for free.
Name the two remaining words without expecting mastery: the room is a sample, the school or the wider group of students it might stand in for is a population, and a sample chosen for convenience does not automatically represent the population it was drawn from. Nothing here needs to be exact yet. The point of the eight minutes is that a student leaves already suspicious of any claim that starts with a show of hands, a genuinely useful habit that how to choose a sampling method will make precise later in the unit.
Minutes 28 to 35: what this year is actually about
One paragraph, delivered directly, using the board the class just built as the running example. Say it close to this: statistics is not a course about computing averages, it is a course about deciding what a set of numbers can and cannot tell you, and how confident you get to be about it. Every one of the four skills the course actually tests happened in the last twenty minutes in a small way: you asked a question worth asking, you collected data to answer it, you looked at what the data actually showed, and you argued about how far the conclusion could travel. The rest of the year adds the tools, probability, sampling distributions, confidence intervals, significance tests, regression, that make each of those four steps rigorous instead of a guess.
Keep the unit tour to one sentence per unit: describing and collecting one-variable data first, then probability and random variables, then inference for categorical proportions, then inference for means, then regression between two quantitative variables last. Naming five units in a few seconds tells students the course has a shape without turning day one into a syllabus read-through.
Minutes 35 to 43: the syllabus, in one paragraph
Read nothing aloud. Say the policy once, in one paragraph, and point to where it lives in writing for anyone who wants the detail later. Adapt the specifics to your own school, but the shape below covers what students actually need on day one: how grades are weighted, roughly how much homework there is and whether it is graded for completion or correctness, what happens if an assessment is missed, and where the written syllabus will be posted.
An example paragraph a teacher can adapt directly: assessments make up most of the grade, homework is checked for completion rather than correctness and exists to prepare for those assessments, a missed test is made up within a set number of days rather than skipped, and the full written syllabus will be posted online for anyone who wants the detail today's paragraph left out. Whatever your actual policy is, keep the spoken version about that length. The written version can hold everything else.
Minutes 43 to 50: the first homework, and closing
Assign the homework the discussion in minutes 20 to 28 already set up: ask three people who are not in this class the same two survey questions, hours of sleep last night and how they got to wherever they were going this morning. Bring the six answers tomorrow along with one written sentence: does this new group of three look like a random sample of anyone, or is it a convenience sample the same way this class was?
That sentence is the actual assignment. It needs no materials, is hard to copy usefully from a classmate because everyone's three people are different, and hands day two a real, larger dataset to reopen the same board discussion with, now comparing two datasets instead of describing one. Close by pointing at where the syllabus will be posted and dismiss on time; a class that ends exactly on the bell on day one is itself a small piece of evidence that the course runs on a plan.
The no-tech variant
Nothing above requires a device, which is deliberate: the first day is the wrong moment to discover that a cart of laptops will not connect to the projector. The only technology named anywhere in the plan is the optional step in minutes 10 to 20 where the sleep-hours shape gets clicked into the descriptive statistics sandbox for a live dotplot and a draggable outlier.
Skip that step entirely and the plan is unchanged. The board already holds a hand-built dotplot from the sticky cards, and dragging an outlier by hand is just erasing one card and asking the class which way the middle of the data would move without it. Index cards, a whiteboard, and a marker run the entire fifty minutes without a single screen, which also means the plan survives a projector failure, a fire drill that eats ten minutes, or a first week schedule that has not settled into its final room yet.
Worked walkthrough: reading a real day-one board
A class of 24 students runs the survey. The sorted sleep-hours cards read: 2, 4, 5, 6, 6, 6, 7, 7, 7, 7, 7, 7, 7, 8, 8, 8, 8, 8, 9, 9, 9, 9, 10, 10. The transportation tally reads: walk 3, drive yourself 5, get a ride 9, bus 7. Turn this board into the first four vocabulary words of the course without giving a lecture.
Separate the two columns by what kind of answer they hold. Sleep hours are numbers that could sensibly be averaged, a quantitative variable. Transportation is a label sorted into one of four fixed categories, a categorical variable.
Look for the shape of the numeric column before computing anything. Most of the 24 values sit between 6 and 9, with 7 appearing seven times, the single most common value on the board. The value of 2 sits well apart from that cluster and is worth asking the student about directly rather than silently averaging in.
Compute one statistic out loud so the word has something to point at. The mode of the sleep-hours column is 7 hours, appearing 7 times out of 24. That number is a statistic: a value computed from the sample sitting in the room, not yet a claim about anyone outside it.
Read the categorical tally the same way. Get a ride is the tallest bin at 9 of 24 students, more than double walk's 3. Ask why that might be true of this specific school, for example distance from campus, before it gets treated as a fact about students in general.
Name the sample. All 24 students who happen to be enrolled in this class period are the sample; nothing computed from them yet describes anyone outside the room.
Hand back the four words attached to what the class already pointed at: quantitative variable for sleep hours, categorical variable for transportation, statistic for the mode of 7 hours, sample for the 24 students who produced the board.
Sleep hours is a quantitative variable with a mode of 7 hours and one likely standout value at 2 hours worth asking about directly. Transportation is a categorical variable with four bins, tallest at get a ride with 9 of 24 students. The 24 students are a sample, and every number computed from their answers is a statistic describing that sample, not yet a claim about any wider group.
Worked walkthrough: making the convenience-sample point concrete
The same class of 24 students sits inside a school of roughly 900 students. A student asks, reasonably, why the board's sleep data does not just tell us how much sleep students at this school get. Answer using the numbers rather than the general rule.
State what was actually done: 24 students were surveyed because they happen to be enrolled in this particular class period, not because 24 names were drawn at random from a list of all 900 students.
Name what that method is. A sample chosen because it is easy to reach rather than because it was drawn at random is a convenience sample, and one specific class period systematically includes and excludes certain kinds of students, for example whoever's schedule happens to have a free elective slot at that hour.
Point at a concrete way that could bias the result. If this section meets first thing in the morning, it may quietly select for students who function fine on less sleep, or exclude students with an early sports practice who slept even less. Either would shift the board's numbers away from the true value for all 900 students, in a direction nobody can determine just by looking at the board.
Contrast it with what a random sample would need to look like: numbering all 900 students and using a random mechanism to choose, say, 30 of them regardless of what class period they are in, so every student had an equal chance of ending up in the sample.
State the honest scope of the claim available today: the board describes this class period's sleep habits reasonably well and says nothing reliable yet about the other roughly 876 students in the building.
The 24 students are a convenience sample of the school, gathered because they share a class period rather than because they were chosen at random, and a class period can plausibly relate to sleep habits in ways nobody on the board can see. The data is trustworthy as a description of this room and not yet trustworthy as a description of the school's roughly 900 students; reaching that wider claim would require randomly selecting students from the full enrollment, the skill the sampling topics later in the unit teach how to do.
Frequently asked questions
What should happen in the first ten minutes of AP Statistics?
A short data-collection activity, not the syllabus. A two-question class survey, one categorical question and one quantitative question, puts a real dataset on the board within ten minutes and gives the rest of the period something concrete to discuss instead of a list of rules.
What survey questions work well for a day-one data-collection hook?
One quantitative question with a numeric answer, such as hours of sleep last night, and one categorical question with a small fixed number of categories, such as how a student got to school. Together they produce one dotplot-shaped column and one bar-shaped tally, which sets up the categorical-versus-quantitative distinction the course opens with before either variable type has a name.
Should the syllabus be read aloud on the first day?
No. Post it, summarize the grading and homework policy in one spoken paragraph, and point students to the full written version. Reading a syllabus line by line spends the one period where students decide what kind of class this will be, and it teaches nothing that a written page does not already say.
What is a good first homework assignment for AP Statistics?
Have students ask three people outside the class the same two survey questions used in the day-one activity, then bring the answers back with one sentence judging whether that small group looks like a random sample of anyone or just another convenience sample. It needs no materials, is hard to copy since everyone surveys different people, and gives day two a second dataset to compare against the first.
How do you run this plan with no technology?
The entire plan is already built around index cards and a whiteboard. The only optional technology is clicking the sleep-hours points into a descriptive statistics interactive for a live dotplot, and skipping it changes nothing else: the board already holds a hand-built dotplot from the survey, and an outlier can be removed by hand just as easily as it is dragged on a screen.
Why point out that the class is not a random sample on day one, before sampling is even taught?
Because treating a convenient group as if it represents everyone is the single habit the sampling and inference units spend the most time correcting later. Raising the question on day one, with the class's own data, makes topics like random sampling and scope of inference land as the answer to something students already noticed rather than a new rule to memorize.