STAT1000 · Midterm 2 · Fall 2026
Pleng's Stat Guide
Every idea on Midterm 2, shown with candy and solved with algebra. Almost every calculation on this exam is one equation with one unknown. If you can solve x + 3 = 11, you can do this.
How to use this guide
- Read the chapters in order. Each one builds on the last. Each shows the idea with M&Ms, gives the algebra, then works the real midterm question. Try the question yourself before you open the solution.
- Do the "Try it" problems. They use new numbers, so you can't lean on remembered answers.
- Take the practice test with a timer set to your real exam length. Use only a calculator, the z-table and your cheat sheet.
- Make your cheat sheet from the one at the end. Writing it out by hand is studying.
Jump to: the midterm, solved · practice test · cheat sheet · z-table
Chapter 1 · Midterm Q1 · 8 points
Count it, measure it, or neither
The big idea. A random variable is a number you don't know until something random happens. Once you know it's random, ask one question: would I count it or measure it?
Count it
Discrete
"How many red M&Ms are in this bag?"
The answer is 0, 1, 2, 3, … You can list the possible values, and there are gaps between them. Nobody finds 2.5 red M&Ms.
Algebra: like the integers.
Measure it
Continuous
"How much does this bag weigh?"
47 g, 47.3 g, 47.31 g, 47.312 g… Any value in a range is possible. Time, distance, weight, height and temperature are always continuous, even when you round them.
Algebra: like an interval of real numbers, such as [40, 55].
Already known
Not a random variable
"How many colors are in a standard bag?"
Six. Every bag, every time. If the number can't come out differently when you repeat the situation, nothing about it is random.
Algebra: a constant, not a variable.
The two-step test
- Could the number come out differently next time (another bag, another day, another customer)? If no, it's not a random variable.
- If yes: can you list the possible values with gaps between them (0, 1, 2, …)? Then it's discrete. If the values fill a whole range, it's continuous.
Classify each as a discrete random variable, a continuous random variable, or not a random variable.
- The amount of time a customer spends waiting in line at a grocery store.
- The number of students absent from a class on a given day.
- The distance a delivery driver travels in one day.
- The number of questions on this exam.
Worked solution
(a) Continuous Time is measured.
(b) Discrete You count students: 0, 1, 2, …
(c) Continuous Distance is measured.
(d) Not a random variable The exam already has its questions printed on it (11). The number can't change, so nothing is random.
Sort the candy: 8 quick ones
Pick an answer for each. You'll see right away whether it's correct, and why.
0 of 8 answered
Where points get lost
- "The number of…" usually means discrete, but first check that it's random. "The number of questions on this exam" is fixed.
- Rounding doesn't make something discrete. "Waiting time to the nearest minute" is still time, so it's continuous.
Chapter 2 · Midterm Q7 + Q8(c) · 13 points
Mean and SD of a random variable
The big idea. A probability table is a recipe for a jar of candy. The mean is the average of the numbers printed on the candies. The SD is how far a typical candy sits from that average.
The table shows the probability distribution of X, the number of cups of coffee a college student drinks per day.
| x | 0 | 1 | 2 | 3 |
|---|---|---|---|---|
| P(X = x) | 0.1 | 0.4 | 0.3 | 0.2 |
- Find the probability that a student drinks at most 1 cup of coffee in a day. (2 pts)
- Calculate the mean number of cups consumed daily. (3 pts)
- Calculate the standard deviation. (4 pts)
(a) A probability is a fraction of the jar
"At most 1" means 1 or fewer, so X ≤ 1. The candies that say 0 or 1: that's 1 + 4 = 5 out of 10.
| The words | The symbol | Which values (0–3) |
|---|---|---|
| at most 1 | X ≤ 1 | 0, 1 |
| at least 1 | X ≥ 1 | 1, 2, 3 |
| more than 1 | X > 1 | 2, 3 |
| fewer than 1 | X < 1 | 0 |
For counts, ≤ and < are different questions. Read the words slowly.
(b) The mean is the average number in the jar
Add up the numbers on all 10 candies, then divide by 10:
The formula does exactly the same thing with fractions instead of candies. Multiply each value by its probability and add:
The mean doesn't have to be a value X can take. Nobody drinks exactly 1.6 cups, but that's the long-run average.
(c) The SD is the typical distance from the mean
Variance is the average squared distance from the mean. Do it as a table, one row per value:
| x | P(x) | x − μ | (x − μ)² | (x − μ)²·P(x) |
|---|---|---|---|---|
| 0 | 0.1 | −1.6 | 2.56 | 0.256 |
| 1 | 0.4 | −0.6 | 0.36 | 0.144 |
| 2 | 0.3 | 0.4 | 0.16 | 0.048 |
| 3 | 0.2 | 1.4 | 1.96 | 0.392 |
| Var(X) = sum = | 0.84 | |||
A faster variance: Var(X) = E(X²) − μ². Here's why. Expand the square inside the sum:
For the coffee jar: E(X²) = 0²(0.1) + 1²(0.4) + 2²(0.3) + 3²(0.2) = 3.4, so Var = 3.4 − 1.6² = 3.4 − 2.56 = 0.84. Same answer. On the exam, use one method and check with the other.
Two rules every probability table follows
Every probability is between 0 and 1, and they all add up to 1. So if one entry is missing, it equals 1 minus the rest. (A jar always holds 100% of its candy.)
Worked solution, written the way the grader wants it
Part B · Midterm Q8(c) · 4 points
Converting units: shift and stretch
The idea. Add the same amount to every candy and the whole pile slides over; its spread doesn't change. Multiply every candy and the pile stretches, so the spread is multiplied too.
This is graphing y = a + bx. The +a shifts the graph; the ×b stretches it. Spread only notices stretching.
Candy example. Bags hold N candies, with mean 55 and SD 2. A shop charges $1.00 per bag plus $0.10 per candy: Price = 1 + 0.10N. Mean price = 1 + 0.10(55) = $6.50. SD of price = 0.10 × 2 = $0.20. The $1 doesn't add spread, because every bag pays it.
November daily highs in Pittsburgh are normal with mean 54 °F and SD 4 °F. The formula C = (F − 32) × 59 converts °F to °C. Find the mean and SD in °C.
Worked solution
Rewrite it as a + bF: C = −1609 + 59F. So b = 5/9.
Trap: (4 − 32) × 5/9 = −15.6. An SD can never be negative. Subtracting 32 shifts every day by the same amount, so it doesn't change the spread.
Where points get lost
- Stopping at the variance. Var = 0.84 is not the SD. Take the square root.
- Dividing by 4 (the number of values) instead of using the probabilities. The values aren't equally likely.
- Applying the shift to the SD. Adding or subtracting never changes spread.
Try it
1. X = the number of cracked M&Ms in a bag. P(0) = 0.5, P(1) = 0.3, P(2) = ?, P(3) = 0.05. Find P(2), the mean and the SD.
Show answer
P(2) = 1 − (0.5 + 0.3 + 0.05) = 0.15
μ = 0(0.5) + 1(0.3) + 2(0.15) + 3(0.05) = 0.75 cracked candies
E(X²) = 0 + 0.3 + 0.6 + 0.45 = 1.35, so Var = 1.35 − 0.75² = 0.7875 and SD = √0.7875 = 0.8874
2. Exam scores have mean 70 and SD 10. The teacher curves them: New = 1.1 × Old + 5. Find the new mean and SD.
Show answer
Mean = 1.1(70) + 5 = 82. SD = 1.1 × 10 = 11. The +5 doesn't touch the SD.
Chapter 3 · Midterm Q11 · 13 points
The binomial: counting blue M&Ms
The big idea. Grab a set number of candies and count how many are blue. That count follows a binomial distribution. Midterm Q11 is exactly this, with contact lenses instead of candy.
The setup
A huge, well-mixed bin of M&Ms is 30% blue. You grab 6. Let X = the number of blue M&Ms in your 6.
Midterm Q11: 30% of students wear contacts (blue candies), you pick 6 students (grab 6), and X = how many of them wear contacts. Same numbers, same math.
Is it binomial? Check BINS
The formula, and why it works
One particular handful, like blue, blue, other, other, other, other, has probability 0.3 × 0.3 × 0.7 × 0.7 × 0.7 × 0.7 = 0.3²·0.7⁴. You multiply because the candies are independent. But the two blues could sit in other spots. C(6, 2) = 15 counts the spots. Every one of those 15 handfuls has the same probability, so you multiply by 15.
"At least one": use the opposite
"At least one blue" means 1, 2, 3, 4, 5 or 6 blues. That's six cases to add. Its opposite has only one case: zero blues, the single handful with no blue in it. So subtract that one case from 1:
Algebra: P(A) + P(not A) = 1, so P(A) = 1 − P(not A).
"Given that": shrink the world
"Given" means: throw away every outcome that doesn't fit, then find the fraction among what's left. Picture 1,000 people each grabbing 6 candies:
Mean and SD: shortcut formulas
30% of 6 is 1.8, so on average a handful has 1.8 blues. The long method from Chapter 2 gives the same numbers. The binomial is just a special probability table.
Suppose 30% of college students wear contact lenses. A random sample of 6 students is selected; their contact lens use is independent.
- Find the probability that at least one student wears contact lenses. Clearly define the random variable X and name its distribution. (5 pts)
- Given that at least one student wears contact lenses, find the probability that exactly two do. (4 pts)
- Find the expected number of students who wear contacts and the standard deviation. (4 pts)
Worked solution
Where points get lost
- Writing P(X ≥ 1) = 1 − P(X = 1). The opposite of "at least 1" is "exactly 0".
- Skipping the definition. Q11(a) gave points for defining X in words and naming the distribution. Write both, every time.
- In a "given" problem, forgetting to divide by the probability of the given part.
- Calling something binomial when n isn't fixed. "Pull candies until you get a blue" is not binomial.
Try it
1. A bin is 20% green. You grab 5. Find P(exactly 1 green), P(at least 1 green), and the mean and SD of the number of greens.
Show answer
X ~ Bin(5, 0.2). P(X = 1) = C(5, 1)(0.2)(0.8)⁴ = 5(0.2)(0.4096) = 0.4096
P(X ≥ 1) = 1 − 0.8⁵ = 1 − 0.3277 = 0.6723
μ = 5(0.2) = 1, σ = √(5 · 0.2 · 0.8) = √0.8 = 0.894
2. Binomial or not? (i) Flip a coin 10 times and count heads. (ii) Pull candies until you get a blue and count the pulls. (iii) Ask 50 random adults if they own a dog and count the yeses.
Show answer
(i) Yes, Bin(10, 0.5). (ii) No: the number of pulls isn't fixed in advance. (iii) Yes: fixed n = 50, yes/no answers, independent people, same chance for each.
Chapter 4 · Midterm Q6 · 12 points
The bell curve and 68–95–99.7
The big idea. Weigh 200 bags of M&Ms and stack one candy above each bag's weight. The stack makes a bell. The empirical rule tells you how many candies land within 1, 2 and 3 SDs of the middle.
The half-bell numbers
Memorize this row. It's the whole rule, split into pieces you can add.
The pieces come straight from 68, 95 and 99.7 by subtracting and halving (the bell is symmetric):
The percentile ladder
The pth percentile is the value with p% of the candies below it. Add up the pieces from the left:
Example: 84% = 50 (the whole left half) + 34 (the first piece on the right). So the 84th percentile is 1 SD above the mean.
The method, every time
- Turn each endpoint into SD steps: k = (x − μ)/σ. For this rule they'll be whole numbers.
- Sketch a bell and mark the steps from −3 to +3.
- Add the half-bell pieces between your endpoints.
- "How many?" questions: multiply that percentage by the total count.
Tomato weights are normal with mean 160 g and SD 30 g. Use the 68–95–99.7 rule.
- What percentage weigh between 100 g and 220 g?
- What percentage weigh between 130 g and 250 g?
- What is the 84th percentile of tomato weights?
- Of 200 tomatoes, about how many weigh more than 190 g?
Worked solution
First write the SD ladder: 70 · 100 · 130 · 160 · 190 · 220 · 250 (that's −3σ to +3σ).
(a) 100 is −2σ and 220 is +2σ. Within 2 SD: 95%
(b) 130 is −1σ and 250 is +3σ. Add the pieces: 34 + 34 + 13.5 + 2.35 = 83.85%. The wrong choice 81.5% is what you'd get for −1σ to +2σ.
(c) 84% = 50 + 34, so 1 SD above: 160 + 30 = 190 g
(d) Above +1σ is 100 − 84 = 16%. 16% of 200 = 0.16 × 200 = 32 tomatoes
Where points get lost
- Uneven intervals like −1σ to +3σ. Don't reach for 68, 95 or 99.7. Add the pieces on each side separately.
- "How many" questions want a count, not a percentage. Multiply by the total.
- Using this rule when the endpoints aren't whole SD steps. Then you need the z-table (Chapter 5).
Try it
1. Bag weights are normal with mean 48 g and SD 1.5 g. (a) What % weigh between 45 and 51 g? (b) What % weigh more than 51 g? (c) What % weigh between 46.5 and 51 g? (d) What is the 16th percentile? (e) Of 1,000 bags, how many weigh under 45 g?
Show answer
Ladder: 43.5 · 45 · 46.5 · 48 · 49.5 · 51 · 52.5
(a) ±2σ → 95%. (b) 51 g is +2σ; above it is 2.35 + 0.15 = 2.5%. (c) −1σ to +2σ → 34 + 34 + 13.5 = 81.5%. (d) 16th percentile = −1σ = 46.5 g. (e) Below −2σ is 2.5%, and 0.025 × 1,000 = 25 bags.
Chapter 5 · Midterm Q3, Q8(a)(b), Q10 · 15 points
z-scores and the z-table
The big idea. A z-score says how many SDs a value is from its mean. Once every value is in SD steps, one table works for every normal curve.
This equation is behind Q3, Q6, Q8, Q9 and Q10. Each rearrangement is one exam question type:
| You want | Solve for it | On the midterm |
|---|---|---|
| How unusual a value is | z = (x − μ) / σ | Q3, Q8(a), Q9(a) |
| A cutoff value | x = μ + zσ | Q8(b), Q9(b) |
| The SD | σ = (x − μ) / z | Q10 |
| The mean | μ = x − zσ | Practice test |
The only "stats" step is the z-table, which turns a z into an area, or an area back into a z. Everything else is algebra.
Comparing apples and peanuts (Q3)
A plain M&M and a peanut M&M aren't on the same scale, so asking "which is heavier for its kind?" means comparing z-scores. Same with SAT and ACT scores. Convert both to SD steps and they share one ruler:
Reading the z-table
The table gives the area to the left of z. In candy terms: line up every bag from lightest to heaviest. The table tells you what fraction of the line is lighter than you.
To look up z = 0.75: go to row 0.7, column .05. The entry is 0.7734. So 77.34% of values are below z = 0.75.
For a continuous variable, "at least" and "more than" give the same answer: P(X ≥ 57) = P(X > 57). A single exact value has zero area under the curve.
Going backwards: from a percentage to a value
- Turn the percentage into an area to the left. Bottom 15% → 0.15. Top 3% → 1 − 0.03 = 0.97.
- Find the number inside the table closest to that area.
- Read z off that entry's row and column.
- Solve: x = μ + zσ.
Handy ones to know: top 10% → z = 1.28 · top 5% → 1.645 · top 2.5% → 1.96 · top 1% → 2.33. For a bottom percentage, use the same number with a minus sign.
November daily highs in Pittsburgh are normal with mean 54 °F and SD 4 °F.
- What is the probability that the high on a random November day is at least 57 °F?
- Find the high temperature below which the coldest 15% of November days fall.
Worked solution
Sanity check: the coldest days must be below the mean of 54. 49.84 is.
Delivery times are normal with mean 35 minutes. A delivery over 50 minutes is late, and about 18.4% of deliveries are late. Find the standard deviation.
Worked solution
Check-your-work calculator
It uses the same 4-decimal z-table you'll have on the exam and writes the steps the way the grader wants them. It starts loaded with Q8(a). Leave n = 1 for one individual; set n above 1 for a sample mean (Chapter 6).
Where points get lost
- Reading the table as the area to the right. It's always the area to the left.
- "Top 3%": look up 0.97, not 0.03. (Or look up 0.03 and flip the sign of z.)
- A cutoff for a bottom percentage must come out below the mean. If yours doesn't, your z has the wrong sign.
- Skipping the z-score. The exam instructions say answers without the z-score and table step lose credit. Write P(Z > 0.75) on the page.
Try it
1. Made-up numbers: plain M&Ms weigh 0.86 g on average (SD 0.04 g); peanut M&Ms weigh 2.55 g (SD 0.15 g). A plain one weighs 0.95 g and a peanut one weighs 2.80 g. Which is more unusually heavy for its kind?
Show answer
Plain: z = (0.95 − 0.86)/0.04 = 2.25. Peanut: z = (2.80 − 2.55)/0.15 = 1.67. The plain one is further above its mean.
2. Bag weights are normal with mean 48 g and SD 1.5 g. Find P(bag < 46 g) and P(46 g < bag < 50 g).
Show answer
z = (46 − 48)/1.5 = −1.33 → table 0.0918.
50 g gives z = +1.33 → 0.9082. Between: 0.9082 − 0.0918 = 0.8164.
3. Same bags. The heaviest 10% weigh more than what?
Show answer
Top 10% → area 0.90 on the left → closest entry 0.8997 at z = 1.28. x = 48 + 1.28(1.5) = 49.92 g.
4. A machine fills bags with mean 48 g. 2.5% of bags weigh under 45 g. Find the SD.
Show answer
Area 0.025 → z = −1.96. Then −1.96 = (45 − 48)/σ, so σ = −3 / −1.96 = 1.53 g.
Chapter 6 · Midterm Q2, Q4, Q5, Q9 · 23 points, the most
Handfuls: sampling distributions and the CLT
The big idea. One candy can be weird. The average of a handful is much less weird. The sampling distribution is the pile you'd get by grabbing handful after handful and writing down each handful's average.
The five words, in candy
| Word | In candy | Symbol |
|---|---|---|
| Population | Every M&M in the giant bin. | |
| Parameter | A fixed number about the whole bin, like its true average weight. You usually don't know it, but it never changes. | μ, σ, p |
| Sample | One handful of n candies. | n |
| Statistic | A number you compute from your handful, like its average weight. It changes from handful to handful. | x |
| Sampling distribution | The pattern of x over all possible handfuls. It's the distribution of a statistic. |
Memory trick: Parameter goes with Population. Statistic goes with Sample.
Three facts about the sample mean
1 · Center
Handful averages center on the bin's true average, whatever the handful size.
2 · Spread
This is the standard error. Bigger handfuls give a tighter pile, always tighter than single candies.
3 · Shape
Normal population: normal for any n. Otherwise, approximately normal once n ≥ 30. That's the Central Limit Theorem (CLT).
What happens to the standard error if you take 4 times as many candies?
Multiply n by 4 and the SE is cut in half. Multiply by 9: one third. By 100: one tenth. To halve the SE you need 4 times the data, not 2 times.
Is x normal? Ask in this order
The lopsided jar: watch the CLT happen
From the 30%-blue bin in Chapter 3, pull candies one at a time until you get a blue. X = how many you pulled. Most people get a blue within 1–3 pulls, but a few unlucky ones need 10 or more. That gives a right-skewed population with μ = 3.33 and σ = 2.79.
Now get n people to do it and average their pull counts. Repeat 2,000 times. Pick n:
Q2. BMI among teenage girls is right-skewed. With a random sample of 20, can you conclude the sampling distribution of the sample mean is approximately normal?
Q4. Soda per can is normal, mean 16 oz, SD 0.3 oz. Smallest sample size among 2, 20, 30, 50 that ensures x is normal?
Q5. True or false: (a) If n is multiplied by four, the standard error of the sample mean is cut in half. (b) The sampling distribution of the sample mean has greater variability than the population. (c) The mean of the sampling distribution of the sample mean stays the same as n increases. (d) A sampling distribution is the probability distribution of a population parameter.
Worked solution
Q2 (b) No The population is skewed and 20 < 30, so the CLT doesn't apply yet.
Q4 (a) 2 The population is already normal, so x is normal for any n. The smallest choice wins.
Q5 (a) True √4 = 2, so SE halves. (b) False Less: σ/√n ≤ σ. (c) True It's always μ. (d) False It's the distribution of a statistic. A parameter is one fixed number, so it has no distribution.
Resting heart rates of adult runners are normal with mean 72 bpm and SD 10 bpm. A random sample of 36 runners is selected.
- What is the probability that the sample mean resting heart rate is less than 70 bpm? (5 pts)
- Find the cutoff above which the highest 3% of sample means fall. (4 pts)
Worked solution
About notation: the answer key writes X ~ N(72, 10²/36). There the second number is the variance (SD squared). Some books write N(mean, SD) instead. Check which one your instructor uses. Either way, the number you divide by in z is the SD, 10/6 = 1.667.
Where points get lost
- Using σ = 10 instead of σ/√n = 1.667 when the question says "sample mean". With 10 you've answered a question about one runner, not the average of 36.
- "The sampling distribution has more spread than the population" is false. Averages are always less spread out.
- "The CLT works for any n" is false for a skewed population with n < 30.
- When the population is normal, the sample size doesn't matter for shape.
Try it
1. Bag weights are normal with mean 48 g and SD 1.5 g. A box holds 9 bags. Find P(average weight of the 9 bags < 47 g). Compare it with P(one bag < 47 g).
Show answer
SE = 1.5/√9 = 0.5. z = (47 − 48)/0.5 = −2.00 → 0.0228.
One bag: z = (47 − 48)/1.5 = −0.67 → 0.2514. One bag is light fairly often; the average of 9 rarely is.
2. With n = 25, the standard error is 2. What n makes the standard error 1?
Show answer
Halving the SE takes 4 times the data: n = 100. With algebra: σ = 2 × √25 = 10, then 10/√n = 1 gives √n = 10, so n = 100.
Answer key, checked
The midterm, solved
Every answer below was worked out independently and matches the official key (84 points). Full work is in each chapter.
| Q | Answer | The move | Pts | Chapter |
|---|---|---|---|---|
| 1 | (a) Continuous (b) Discrete (c) Continuous (d) Not random | Measure → continuous, count → discrete, fixed → not random | 8 | 1 |
| 2 | (b) No | Skewed population and n = 20 < 30 | 3 | 6 |
| 3 | (c) Ann | z = 1.00 beats z = 0.50 | 3 | 5 |
| 4 | (a) 2 | Normal population → x normal for any n | 3 | 6 |
| 5 | T, F, T, F | SE = σ/√n; center stays μ; distribution of a statistic | 8 | 6 |
| 6 | 95% · 83.85% · 190 g · 32 | Half-bell pieces 34 / 13.5 / 2.35 / 0.15 | 12 | 4 |
| 7 | 0.5 · 1.6 cups · 0.9165 cups | ΣxP, then √Σ(x − μ)²P | 9 | 2 |
| 8 | 0.2266 · 49.84 °F · 12.22 °C and 2.22 °C | Right tail = 1 − table; x = μ + zσ; SD ignores the shift | 12 | 5, 2 |
| 9 | 0.1151 · 75.13 bpm | Use SE = 10/√36 = 1.667 | 9 | 6 |
| 10 | σ = 16.67 min | Area 0.816 → z = 0.90 → solve for σ | 4 | 5 |
| 11 | Bin(6, 0.3) · 0.8824 · 0.3673 · 1.8 and 1.122 | Complement; "given" → divide; np and √np(1 − p) | 13 | 3 |
Same skills, new numbers
Practice test
Question for question, this mirrors Midterm 2, with the same point values (84 total). Time yourself, show every step, then open the solutions. Anything you miss: go back to that question's chapter.
For normal probabilities, show the z-score and use the z-table, except when told to use the 68–95–99.7 rule.
1. (8 pts) Classify each as discrete, continuous, or not a random variable. (a) The time it takes a pizza to be delivered. (b) The number of text messages you get in an hour. (c) The number of days in a week. (d) The height of a randomly chosen student.
Solution
(a) Continuous: time is measured. (b) Discrete: you count texts. (c) Not a random variable: always 7. (d) Continuous: height is measured.
2. (3 pts) Household incomes in a city are strongly right-skewed. A random sample of 100 households is taken. Is the sampling distribution of the sample mean income approximately normal? (a) No, because the population is skewed. (b) Yes, because n = 100 is large enough for the CLT. (c) Yes, because random samples always give normal sampling distributions. (d) No, only normal populations give normal sampling distributions.
Solution
(b). The population is skewed, but 100 ≥ 30. Compare Q2 on the midterm, where n = 20 was too small.
3. (3 pts) Maya scored 85 on a chemistry exam (mean 75, SD 8). Leo scored 80 on a history exam (mean 68, SD 12). Who did better relative to their class?
Solution
Maya: z = (85 − 75)/8 = 1.25. Leo: z = (80 − 68)/12 = 1.00. Maya.
4. (3 pts) Lengths of adult trout in a lake are normal with mean 30 cm and SD 4 cm. What is the smallest sample size, among these, that ensures the sampling distribution of the sample mean is normal? (a) 5 (b) 15 (c) 30 (d) 100 (e) Not enough information
Solution
(a) 5. The population is normal, so any n works. If the lengths were skewed, the answer would be (c) 30.
5. (8 pts) True or false. (a) If the sample size is multiplied by 9, the standard error of the sample mean is divided by 3. (b) As the sample size increases, the SD of the sampling distribution of x increases. (c) The sampling distribution of x is centered at μ, whatever the sample size. (d) The population mean μ is a statistic.
Solution
(a) True: σ/√(9n) = σ/(3√n). (b) False: σ/√n shrinks. (c) True. (d) False: μ is a parameter; x is a statistic.
6. (12 pts) Phone battery life is normal with mean 40 hours and SD 5 hours. Use the 68–95–99.7 rule. (a) What % last between 30 and 50 hours? (b) What % last between 30 and 45 hours? (c) What is the 97.5th percentile? (d) Of 400 phones, about how many last less than 30 hours?
Solution
Ladder: 25 · 30 · 35 · 40 · 45 · 50 · 55.
(a) ±2σ → 95%. (b) −2σ to +1σ → 13.5 + 34 + 34 = 81.5%. (c) 97.5% = +2σ → 50 hours. (d) Below −2σ is 2.5%; 0.025 × 400 = 10 phones.
7. (9 pts) X = number of pets in a household: P(0) = 0.3, P(1) = 0.4, P(2) = 0.2, P(3) = 0.1. (a) Find P(at least 2 pets). (b) Find the mean. (c) Find the SD.
Solution
(a) P(X ≥ 2) = 0.2 + 0.1 = 0.3.
(b) 0(0.3) + 1(0.4) + 2(0.2) + 3(0.1) = 1.1 pets.
(c) E(X²) = 0 + 0.4 + 0.8 + 0.9 = 2.1. Var = 2.1 − 1.1² = 0.89. SD = √0.89 = 0.9434 pets. (Long way: (−1.1)²(0.3) + (−0.1)²(0.4) + (0.9)²(0.2) + (1.9)²(0.1) = 0.363 + 0.004 + 0.162 + 0.361 = 0.89 ✓)
8. (12 pts) Commute times are normal with mean 28 minutes and SD 5 minutes. (a) Find the probability a commute takes at least 31 minutes. (b) Find the time below which the fastest 10% of commutes fall. (c) A rideshare charges $3 plus $0.50 per minute. Find the mean and SD of the cost.
Solution
(a) P(X ≥ 31) = P(Z ≥ (31 − 28)/5) = P(Z ≥ 0.60) = 1 − 0.7257 = 0.2743.
(b) Area 0.10 → closest entry 0.1003 at z = −1.28. x = 28 + (−1.28)(5) = 21.6 minutes.
(c) Cost = 3 + 0.5X. Mean = 3 + 0.5(28) = $17.00. SD = 0.5 × 5 = $2.50.
9. (9 pts) Systolic blood pressure is normal with mean 120 and SD 15. A random sample of 25 people is taken. (a) Find the probability that the sample mean is above 125. (b) Find the cutoff above which the highest 1% of sample means fall.
Solution
SE = 15/√25 = 3, so X ~ N(120, 3).
(a) P(Z > (125 − 120)/3) = P(Z > 1.67) = 1 − 0.9525 = 0.0475.
(b) Area 0.99 → closest entry 0.9901 at z = 2.33. x = 120 + 2.33(3) = 126.99.
For comparison, one person above 125: z = 0.33 → 1 − 0.6293 = 0.3707. Individuals land above 125 often; averages of 25 rarely do.
10. (4 pts) Delivery times are normal with mean 30 minutes, and 10% of deliveries take longer than 40 minutes. Find the SD. Bonus: test scores are normal with SD 4, and 20% score below 60. Find the mean.
Solution
90% are below 40 → area 0.90 → closest entry 0.8997 at z = 1.28. 1.28 = (40 − 30)/σ, so σ = 10/1.28 = 7.81 minutes.
Bonus: area 0.20 → 0.2005 at z = −0.84. −0.84 = (60 − μ)/4 → μ = 60 + 0.84(4) = 63.36.
11. (13 pts) 25% of students take the bus to campus. A random sample of 8 students is taken; assume independence. (a) Define X, name its distribution, and find the probability that at least one takes the bus. (b) Given that at least one takes the bus, find the probability that exactly two do. (c) Find the expected number who take the bus and the SD.
Solution
(a) X = the number of the 8 students who take the bus. X ~ Bin(8, 0.25). P(X ≥ 1) = 1 − 0.75⁸ = 1 − 0.1001 = 0.8999.
(b) P(X = 2) = C(8, 2)(0.25)²(0.75)⁶ = 28(0.0625)(0.1780) = 0.3115. P(X = 2 | X ≥ 1) = 0.3115/0.8999 = 0.3461.
(c) μ = 8(0.25) = 2 students. σ = √(8 · 0.25 · 0.75) = √1.5 = 1.225 students.
One page of everything
Cheat sheet
The exam allows one cheat sheet. Copy this onto yours by hand: writing it is part of learning it. Colors match the chapters.
Download the one-page PDF (letter size)Random variable?
- Fixed, already known → not a random variable
- Count it → discrete
- Measure it (time, distance, weight, temperature) → continuous
Probability table
- ΣP(x) = 1 missing one = 1 − rest
- μ = Σ x·P(x)
- Var = Σ(x − μ)²P(x) = Σx²P(x) − μ²
- σ = √Var (don't forget the root)
- at most ≤ · at least ≥ · more than > · fewer than <
Shift and stretch: Y = a + bX
- μY = a + bμX
- σY = |b|σX Var(Y) = b²Var(X)
- °C = (°F − 32)·5/9 → mean: plug in; SD: × 5/9 only
Binomial (BINS)
- Binary · Independent · n fixed · Same p
- Write: "X = number of … out of n" and "X ~ Bin(n, p)"
- P(X = k) = C(n,k) pk(1 − p)n−k
- μ = np σ = √np(1 − p)
- P(X ≥ 1) = 1 − (1 − p)n
- P(A | B) = P(A and B) / P(B)
68–95–99.7
- Pieces from the middle out, each side: 34 · 13.5 · 2.35 · 0.15
- Percentiles at −3σ…+3σ: 0.15 · 2.5 · 16 · 50 · 84 · 97.5 · 99.85
- Uneven interval → add pieces. "How many" → % × total
z-scores
- z = (x − μ)/σ ⇔ x = μ + zσ
- σ = (x − μ)/z ⇔ μ = x − zσ
- Table = area left. Right = 1 − table. Between = table(b) − table(a)
- Top p% → look up 1 − p. Bottom p% → look up p
- Top 10% 1.28 · 5% 1.645 · 2.5% 1.96 · 1% 2.33
- Write: P(X > a) = P(Z > (a − μ)/σ) = …
Sample mean x
- mean = μ SE = σ/√n (n × 4 → SE ÷ 2)
- Shape: normal population → normal, any n. Otherwise n ≥ 30 → approx. normal (CLT). Else: can't say
- z = (x − μ) / (σ/√n)
- Spread of x is always less than the population's
Words
- Parameter ↔ Population (μ, σ, p): fixed
- Statistic ↔ Sample (x): changes every sample
- Sampling distribution = distribution of a statistic
Standard normal · area to the left of z
z-table
The standard left-area table, the same layout most courses hand out. Row = z to one decimal, column = the second decimal. Tap any entry to see what it means.
Example: z = 0.75 is row 0.7, column .05 → 0.7734