EC 285 - Midterm 1 - Practice Test - Answers

EC 285 - Midterm 1 - Practice Test - Answers

Back to the test: EC 285 - Midterm 1 - Practice Test

Part A is 1 mark per question. Part B shows the marks for each part. The total is 30.

If you write it before the Sep 24 class, score the taught questions on their own: they're worth 22 marks. The other 8 marks are Questions 12, 14 and 15 and Question 17 (c) to (f). Questions 12 and 15 and parts (c) to (e) aren't taught yet. Question 14 and part (f) lean on the addition rule, which was only previewed. Part (f) also uses the Venn terms, which were taught, so it's still worth trying now.

Quick answer grid

Question Answer Topic
1 D Parameter and statistic, N and n
2 A Descriptive and inferential statistics
3 B Systematic sampling and simple random sampling
4 C Types of variables, and the chart for each type
5 A Histogram area and merged class intervals
6 C Outliers, shifts and the median with an even count
7 D Skewness
8 D Percentile position
9 A Symbols, centre and spread
10 D Correlation only sees straight lines, the sign of the covariance, and causation
11 A Intersection, union, complement, mutually exclusive, collectively exhaustive
12 B Tree diagram and multiplication rule (not taught yet)
13 B Permutations
14 C Addition rule (not taught yet)
15 B Odds (not taught yet)
16 (a) n=30, mode =4 courses Reading a frequency table
16 (b) x¯=3.8 courses Weighted mean
16 (c) s=1.32 courses Standard deviation from a frequency table
16 (d) Q1=3, median =4, Q3=5, range =5, IQR =2 Percentiles, range and interquartile range from a frequency table
16 (e) CV=34.83% (anything from 34.74% to 34.84% is fine) Coefficient of variation
16 (f) 33.33% "At least" includes the endpoint
17 (a) 0.42 Counting with combinations
17 (b) 0.93 Complement rule
17 (c) 0.30, 0.30, 0.15, 0.25 (table below) Probability table (not taught yet)
17 (d) P(C∣J)=0.50 Conditional probability (not taught yet)
17 (e) Not independent, since 0.30≠0.27 (the extra-practice check: 0.50≠0.45) Independence (not taught yet)
17 (f) Neither: not mutually exclusive (30% have both) and not collectively exhaustive (25% have neither) Mutually exclusive and collectively exhaustive (needs the addition rule, only previewed)

Part A - Multiple choice

Question 1

Worked solution

  1. The 400 surveyed students are a subset of the 19,000 undergraduates. So the 400 are the sample and the 19,000 are the population.
  2. n=400 -- a sample size is written with a lowercase n
  3. N=19,000 -- a population size is written with a capital N
  4. The $85 was calculated from the 400 students in the sample, so it describes the sample. A characteristic of a sample is a statistic.
  5. The researcher may use the $85 to estimate the average for all 19,000 students. That doesn't make it a parameter. The parameter is the true average of all 19,000, and nobody has measured it.

Answer: (D). The $85 is a statistic, and 400 is written n.

Why: A number's name comes from the group it was calculated from. A number calculated from the whole population is a parameter, and the population's size is N. A number calculated from a sample is a statistic, and the sample's size is n. Using a statistic to guess at the population doesn't turn it into a parameter.

Why the other options are wrong:

Where it's covered: Lecture 01: Parameter and statistic and Lecture 01: Population and sample.

Question 2

Worked solution

  1. Descriptive statistics summarizes the data you already have, with charts and summary numbers like the mean and the median.
  2. Inferential statistics uses sample data to make predictions, forecasts and estimates about a bigger population.
  3. Option (A) takes a sample of 1,000 households and uses it to estimate something about all Ontario households. That goes beyond the data in hand, so it's inference.
  4. Options (B), (C) and (D) each summarize data that has already been collected. None of them makes a claim about anything bigger.

Answer: (A). Using a sample of 1,000 households to estimate the average income of all Ontario households.

Why: The test is whether you are going beyond the data you have. Summarizing your data is descriptive. Using a sample to say something about the population is inferential.

Why the other options are wrong:

Where it's covered: Lecture 01: Descriptive and inferential statistics and Lecture 02: Two branches of statistics.

Question 3

Worked solution

Statement I (is it a simple random sample?):

  1. Chapter 1 slide 6 says a simple random sample needs three things. Each customer is chosen strictly by chance, each customer is equally likely to be chosen, and every possible sample of n=10 customers is equally likely to be chosen.
  2. j=Nn=15010=15 -- so the random start can only be a number from 1 to 15
  3. Once the start is picked, the other 9 customers are locked in. So only 15 different samples can ever come out.
  4. For example, customers 1 and 2 can never be in the same sample, because every pick is 15 apart. A simple random sample could pick them together.
  5. So not every possible sample of 10 is equally likely, and Statement I is false.

Statement II (the 6th customer):

  1. j=15 from step 2 -- take every 15th customer
  2. The random start, customer 4, is the 1st customer in the sample.
  3. 4, 19, 34, 49, 64, 79 -- keep adding 15 to get the 1st to 6th customers
  4. Shortcut: the 6th pick is 5 steps after the start, so 4+5×15=79.
  5. So Statement II is true.

Answer: (B). Statement II only.

Why: The random start only makes the first pick random. After that, every pick is fixed by the step of 15. That's why systematic sampling is its own method, separate from simple random sampling, and those are the two methods she said you're responsible for. She said it herself right after the 72-and-9 example: "it's not random sampling, it's more organized, it's more systematic, but there is some randomness" (Sep 15 recording, at 18:04). Each customer does have the same chance of being picked, 10150=115, which is what makes the trap tempting. It's the third condition that fails. There are C10150 possible samples of 10, which is about 1.17×1015, and only 15 of them can come out here. For the 6th pick, the random start counts as the first pick, and each later pick adds j. So the k-th pick is the start plus (k−1) steps of j. This is the same pattern as the class example: 3, 11, 19, and so on for N=72 and n=9.

Why the other options are wrong:

Where it's covered: Lecture 02: Simple random sampling, Lecture 02: Systematic sampling and Chapter 1 slide 6 (the three conditions).

Question 4

Worked solution

  1. (i) The number of courses is a count. It's always a whole number (5, never 5.5), so it's discrete numerical.
  2. (ii) Commute time can take any value on the number line, like 23.7 minutes. So it's continuous numerical.
  3. (iii) Program of study puts students into groups (economics, business, and so on). It isn't a number you can average, so it's categorical.
  4. The chart for (iii): categories go in a bar chart or a pie chart. A histogram needs numerical data on an x-axis that runs in order, so it can't show program of study.

Answer: (C). (i) discrete numerical, (ii) continuous numerical, (iii) categorical, and (iii) could be shown in a bar chart or a pie chart.

Why: First ask whether the variable is a number or a category. If it's a number, ask whether it can take decimals. Whole numbers only means discrete. Any value in a range means continuous. The type then decides the chart. Chapter 1 slide 11 lists the frequency distribution, the bar chart and the pie chart for categorical variables, and the histogram among the charts for numerical ones.

Why the other options are wrong:

Where it's covered: Lecture 02: Discrete and continuous variables, Lecture 02: Categorical variables, Lecture 02: Charts for each type and Lecture 02: Histogram versus bar chart. The chart list is on Chapter 1 slide 11.

Question 5

Worked solution

  1. The merged bar runs from 40 to 60, so it is 2 times as wide as the other bars.
  2. 4+3=7 -- the merged bar holds 7 commute times
  3. In a histogram, the area of each bar must be proportional to its frequency. Counting width in 10-minute units, area = height × width.
  4. height=72=3.5 -- divide by the width so the area stays at 7
  5. 3.5×2=7 -- check: the area matches the frequency

Answer: (A). A height of 3.5.

Why: In a histogram the area of a bar carries its frequency, not the height alone. That is exactly what lets you use class intervals of different widths without distorting the picture. The class example was the same move: 20 to 40 held 11, so its height was 5.5. The line in the question about the original bars only sets the scale: on a 10-minute-wide bar, one unit of height stands for one commute. That's why the answer is 3.5, and not a per-minute number like 0.35.

Why the other options are wrong:

Where it's covered: Lecture 02: Histogram versus bar chart.

Question 6

Worked solution

The original data are already sorted: 12, 15, 15, 18, 20, 22, 25, 28, with n=8.

  1. x¯=12+15+15+18+20+22+25+288=1558=19.38 -- the original mean
  2. n=8 is even, so there are two middle values, the 4th and the 5th. They are 18 and 20.
  3. median=18+202=19 -- average the two middle values

Statement I (28 becomes 280):

  1. x¯=155−28+2808=4078=50.88 -- the mean jumps up
  2. Sorted: 12, 15, 15, 18, 20, 22, 25, 280. The 4th and 5th values are still 18 and 20, so the median is still 19 -- the median doesn't move
  3. So Statement I is true.

Statement II (add 2 to every value):

  1. New data: 14, 17, 17, 20, 22, 24, 27, 30. Then x¯=1718=21.38, which is 19.38+2.
  2. The two middle values are now 20 and 22, so the median is 20+222=21, which is 19+2.
  3. So Statement II is true.

Answer: (C). Both statements are true.

Why: The median only depends on which values sit in the middle positions. The mean uses every value. So one extreme value drags the mean but leaves the median alone. A shift moves every value, including the middle ones, so it moves both by the same amount. With an even count there is no single middle value, so the median is the average of the two middle ones. In class she said she accepts either form for an even count: the average (her example was 3.5), or the two middle values named together ("3 and 4"). So for these data, 19 and "18 and 20" are both fine answers for the median.

Why the other options are wrong:

Where it's covered: Lecture 02: Outliers and shifts, Lecture 02: Median (the even-count rule) and Lecture 01: Outliers.

Question 7

Worked solution

  1. 62<70 -- the mean is less than the median
  2. From the slide: left-skewed means mean < median, symmetric means mean = median, and right-skewed means median < mean.
  3. Mean < median matches left-skewed. A few very low marks (a long tail on the left) pull the mean down below the median.

Answer: (D). Left-skewed, because the mean is less than the median.

Why: The mean gets pulled toward the long tail, and the median doesn't. So the mean ends up on the same side as the tail. A mean below the median means the tail is on the left.

Why the other options are wrong:

Where it's covered: Lecture 02: Shape and skewness.

Question 8

Worked solution

  1. The data are already sorted, and n=11.
  2. position=75100(11+1)=0.75×12=9 -- the 75th percentile is the 9th value
  3. Count along: 5, 8, 10, 12, 14, 15, 18, 20, 25. The 9th value is 25.

Answer: (D). 25 minutes.

Why: The formula (P/100)(n+1) gives a position in the sorted list, not the answer. You then count to that position and read off the value there.

Why the other options are wrong:

Where it's covered: Lecture 02: Percentiles.

Question 9

Worked solution

  1. Population symbols (parameters): μ is the mean, σ2 is the variance, σ is the standard deviation and N is the size.
  2. Sample symbols (statistics): x¯ is the mean, s2 is the variance, s is the standard deviation and n is the size.
  3. Measures of central tendency are the mean, the median and the mode. Measures of variation are the range, the interquartile range, the variance, the standard deviation and the coefficient of variation.
  4. Check (A): s2 is the sample variance, the variance measures spread, and the sample variance divides by n−1. All three parts are true.

Answer: (A). s2 is the sample variance. It is a measure of variation, and it divides by n−1.

Why: Greek letters and a capital N belong to the population. Latin letters and a lowercase n belong to the sample. Her memory trick is "S stands for sample". Separately, centre measures tell you where the middle is, and variation measures tell you how spread out the data are around it.

Why the other options are wrong:

Where it's covered: Lecture 03: Population and sample symbols, Lecture 03: Why divide by n minus 1 and Lecture 03: Center and spread.

Question 10

Worked solution

  1. r=−0.12 is close to 0, so there is no strong straight-line relationship.
  2. The scatter plot shows a clear U: fatigue falls, then rises. So there is a strong relationship. It just isn't a straight line.
  3. Correlation only measures the strength of a linear relationship. The falling half (negative) and the rising half (positive) mostly cancel out, which pushes r toward 0.
  4. r=Cov(x,y)sxsy and both standard deviations are positive, so the covariance has the same sign as r. Here that's negative.

Answer: (D). There is a strong relationship, but it is not linear, and r only measures the strength of a linear relationship.

Why: A correlation near zero means no straight-line relationship, not no relationship. This is her Chapter 2 slide 19 (the U shape with ρ=−0.1540). Nobody in class could answer it, and she said "potentially I ask this on the exam." For the sign, the covariance and the correlation work the same way. Positive means X and Y move in the same direction, negative means they move in opposite directions, and 0 means no straight-line relationship.

Why the other options are wrong:

Where it's covered: Lecture 03: Correlation only sees straight lines, Lecture 03: Covariance and Lecture 01: Correlation and causation.

Question 11

Worked solution

  1. S={1,2,3,4,5,6} -- rolling the die is the random experiment, each face is a basic outcome, and S is the sample space
  2. A={1,2,3} -- "at most 3" includes 3
  3. B={3,4,5,6} -- "at least 3" also includes 3. A and B are both events, because each one is a subset of S.
  4. A∩B={3} -- 3 is in both, so the intersection is not empty and the events are not mutually exclusive
  5. A∪B={1,2,3,4,5,6}=S -- together they cover every face, so they are collectively exhaustive
  6. A¯={4,5,6} -- the complement is every outcome not in A. That isn't B, because B also has the 3.

Answer: (A). A∩B={3} and A∪B=S, so A and B are collectively exhaustive but not mutually exclusive.

Why: "Mutually exclusive" is a question about the intersection: is it empty? "Collectively exhaustive" is a question about the union: is it all of S? They are two separate checks, so a pair of events can pass one and fail the other. An event and its complement always pass both, because they split S with nothing shared and nothing left over.

Why the other options are wrong:

Extra practice: the four basic terms (no marks)

  1. The random experiment is rolling the die once -- a process that leads to an uncertain outcome
  2. One basic outcome is rolling a 4 -- a single possible result of that roll (any face works)
  3. S={1,2,3,4,5,6} is the sample space -- the collection of all the basic outcomes
  4. A={1,2,3} is a subset of S that holds three basic outcomes -- so A is an event, not a basic outcome

Answer: The random experiment is rolling the die. One basic outcome is rolling a 4, though any face works. S is the sample space. A is an event, not a basic outcome, because it's a group of three basic outcomes (1, 2 and 3), while a basic outcome is one single result.

Why: These are the four bold terms on Chapter 3 slide 3. She said the bold terms are "definitely going to be on the midterm test", and that you should know how to explain each one. The terms fit inside each other. The random experiment is the process. Each single result it can give is a basic outcome. All the basic outcomes together make the sample space. Any group of basic outcomes picked out of the sample space is an event. "At most 3" can happen in three different ways, so it's an event, not one outcome.

Where it's covered: Lecture 04: Using the terms together, Lecture 04: Mutually exclusive events, Lecture 04: Collectively exhaustive events, Lecture 04: Complement and Lecture 04: Basic terms.

Question 12

Not taught yet. Expected on Sep 24 (Sep 29 at the latest).

Worked solution

Let O be "lives on campus" and M be "has a meal plan".

  1. P(O)=0.40 and P(O¯)=1−0.40=0.60 -- the complement rule
  2. P(M∣O)=0.75 and P(M∣O¯)=0.35 -- "of the students who live on campus" means "given that they live on campus"
  3. On-campus branch: P(M∩O)=P(M∣O)P(O)=0.75×0.40=0.30 -- the multiplication rule
  4. Off-campus branch: P(M∩O¯)=P(M∣O¯)P(O¯)=0.35×0.60=0.21
  5. P(M)=P(M∩O)+P(M∩O¯)=0.30+0.21=0.51 -- the marginal probability, adding the two branches that end in a meal plan

Answer: (B). 0.51

Why: As a tree, the first split is on campus (0.40) or off campus (0.60). The second split is meal plan or not. Multiply along a path to get the probability of that path. Every student is either on campus or off campus, never both, so the two meal-plan paths don't overlap and you can add them.

Why the other options are wrong:

Where it's covered: Not taught yet. See Chapter 3 slides 19 to 24 (multiplication rule, marginal probability and tree diagram). In the textbook, the multiplication rule is in 3.3, Two Basic Rules of Probability (PDF page 161), and tree diagrams are in 3.4, Contingency Tables and Probability Trees (PDF page 165). The Sep 24 class will link its notes here: EC 285 - 2026-09-24 - Class - Probability Topics. The "above 1" warning is from Lecture 04: Classical probability.

Question 13

Worked solution

  1. Gold, silver and bronze are different prizes, so the order of the top three matters. That makes it a permutation.
  2. n=8 runners, and x=3 medals.
  3. P38=8!(8−3)!=8!5! -- put the numbers into the formula
  4. =8×7×6=336 -- the 5! cancels

Answer: (B). 336 ways.

Why: Settle "order or no order" first. Here, Ana winning gold and Ben winning silver is a different result from Ben gold and Ana silver, so order matters. A ranking or an arrangement means permutations. A plain selection, like a committee, means combinations.

Why the other options are wrong:

Where it's covered: Lecture 04: Order matters or not and Lecture 04: Permutations.

Question 14

Not taught yet, only previewed at the end of the Sep 22 class. The full rule is expected on Sep 24 (Sep 29 at the latest).

Worked solution

Let A be "reads the campus newspaper" and B be "listens to the campus radio station".

  1. P(A)=0.35, P(B)=0.50 and P(A∩B)=0.15 -- "or" means the union, A∪B
  2. P(A∪B)=P(A)+P(B)−P(A∩B) -- the addition rule
  3. =0.35+0.50−0.15=0.70

Answer: (C). 0.70

Why: Adding P(A) and P(B) counts the students who do both twice, once in each. Subtracting P(A∩B) once fixes the double count.

Why the other options are wrong:

Where it's covered: Lecture 04: Addition rule preview, then the full rule on Chapter 3 slides 14 and 15 (slide 14) and textbook 3.3, Two Basic Rules of Probability (PDF page 161).

Question 15

Not taught yet. Expected on Sep 24 (Sep 29 at the latest).

Worked solution

Let A be "the car needs new tires".

  1. P(A)=0.20 and P(A¯)=1−0.20=0.80 -- the complement rule
  2. odds=P(A)1−P(A)=P(A)P(A¯) -- the odds in favour of A
  3. =0.200.80=0.25

Answer: (B). 0.25, which is odds of 1 to 4.

Why: Odds compare the chance an event happens with the chance it doesn't. Here, for every car that needs tires, four don't.

Why the other options are wrong:

Where it's covered: Not taught yet. See Chapter 3 slide 25 (odds) and textbook 3.1, PDF page 151. The book writes odds as "numerator to denominator", so a probability of 0.60 gives odds of 3 to 2, the same way 0.20 gives 1 to 4 here.

Part B - Numerical questions

Question 16

Part (a): sample size and mode (1 mark)

  1. n=∑fi=2+3+6+9+8+2=30 -- each f counts students, so add up the f row
  2. The largest frequency is 9, and it sits under x=4 -- read from the biggest f up to its x

Answer: n=30 students, and the mode is 4 courses.

Why: The f row counts people, so the sample size is the sum of the f row, not the number of columns (6). The mode is the most common answer, so it's an x value in courses, not the count 9. "Your mean, median, mode must be in units of x."

Where it's covered: Lecture 03: Reading a frequency table.

Part (b): the mean (1 mark)

  1. ∑fixi=1(2)+2(3)+3(6)+4(9)+5(8)+6(2) -- weight each value by how many students gave it
  2. =2+6+18+36+40+12=114 -- add the products
  3. x¯=∑fixin=11430=3.8 -- divide by n=30, not by 6

Answer: x¯=3.8 courses.

Why: The table is shorthand for the raw list of 30 answers: two 1s, three 2s, six 3s, and so on. Multiplying each x by its f is the same as adding that value as many times as it appears.

Where it's covered: Lecture 03: Weighted mean.

Part (c): the standard deviation (2 marks)

The students were a random sample, so use the sample formula, which divides by n−1:

s=∑fi(xi−x¯)2n−1
  1. x=1: 2(1−3.8)2=2(7.84)=15.68 -- squared gap times its frequency
  2. x=2: 3(2−3.8)2=3(3.24)=9.72
  3. x=3: 6(3−3.8)2=6(0.64)=3.84
  4. x=4: 9(4−3.8)2=9(0.04)=0.36
  5. x=5: 8(5−3.8)2=8(1.44)=11.52
  6. x=6: 2(6−3.8)2=2(4.84)=9.68
  7. 15.68+9.72+3.84+0.36+11.52+9.68=50.8 -- the numerator
  8. s2=50.830−1=50.829=1.7517 -- the sample variance
  9. s=1.7517=1.3235 -- take the square root

Answer: s=1.32 courses.

Why: The standard deviation is "the average of squared deviations from the average", then square-rooted. Each squared gap is counted as many times as its value appears, which is why it's multiplied by f. The divisor is n−1=29 because this is a sample. A likely mark split is 1 mark for the numerator of 50.8 and 1 mark for dividing by 29 and taking the root. Dividing by 30 (the population formula) gives 1.30, which would lose a mark here. Forgetting to multiply by f is another common slip. A third one is dividing by 6−1=5, the number of different x values minus 1, instead of n−1. That gives 50.8/5=3.19, which is far too big. The divisor counts students, not columns: 30 students answered, so it's 29. Heads up: in the shoes problem, your Sep 17 class note writes the divisor as 7−1, where it should be 40−1=39. The 1.29 written next to it is right, because it comes from dividing by 39. Dividing by 6 would give 3.29.

Where it's covered: Lecture 03: Standard deviation from a frequency table and Lecture 03: Why divide by n minus 1.

Part (d): the quartiles, the median, the range and the interquartile range (2 marks)

  1. Running totals of f: 2,5,11,20,28,30. So positions 1 to 2 hold x=1, 3 to 5 hold x=2, 6 to 11 hold x=3, 12 to 20 hold x=4, 21 to 28 hold x=5, and 29 to 30 hold x=6.
  2. Q1: 25100(30+1)=0.25×31=7.75 -- the 7th and 8th values are both in the block from 6 to 11, so Q1=3
  3. Median: 50100(31)=0.5×31=15.5 -- n=30 is even, so the median sits halfway between the 15th and 16th values. Both are in the block from 12 to 20, so the median is 4.
  4. Q3: 75100(31)=0.75×31=23.25 -- the 23rd and 24th values are both in the block from 21 to 28, so Q3=5
  5. Range=xmax−xmin=6−1=5 -- the largest and smallest answers anyone gave (both have f>0)
  6. IQR=Q3−Q1=5−3=2 -- using the quartiles from steps 2 and 4

Answer: Q1=3 courses, the median is 4 courses, and Q3=5 courses. The range is 5 courses, and the interquartile range is 2 courses.

Why: (P/100)(n+1) gives a position, not the answer. The running totals tell you which x value sits at that position, because the table is already in ascending order. 7.75, 15.5 and 23.25 are positions, so writing them as answers loses the marks. For the median, the 15th and 16th values happen to match here. If they'd been different, say 4 and 5, you'd average them to get 4.5. The range and the IQR are both gaps between two x values. The range uses the two extremes, so one odd answer can stretch it. The interquartile range is the spread of the middle 50% of the data. Using frequencies instead (like 9−2=7) is the mistake to avoid. A likely mark split is 1 mark for the three quartiles and 1 mark for the range and the IQR.

The IQR formula, Q3−Q1, comes from the textbook. In class she only named the IQR as one of the five measures of variation, but on Sep 22 she said one of her review questions asks for Q1, Q3 and the IQR. Take only IQR=Q3−Q1 from the textbook page. Find Q1 and Q3 with her (P/100)(n+1) position rule, not the book's "middle of each half" method. The two happen to agree for these data, but not always. On the book's own 14-value example, her rule puts Q3 at position 11.25, which falls between the values 9 and 10, while the book says Q3=9.

Where it's covered: Lecture 03: Percentiles from a frequency table, Lecture 02: Quartiles, Lecture 02: Median, Lecture 03: Range and Lecture 02: Measures of variation. The IQR formula is in textbook 2.2, PDF page 78.

Part (e): the coefficient of variation (1 mark)

  1. CV=(sx¯)⋅100% -- the sample version
  2. =1.32353.8⋅100% -- keep the unrounded s from part (c) in your calculator
  3. =0.3483⋅100%=34.83%

Answer: CV=34.83%. Anything from 34.74% to 34.84% is fine. The only difference is when you rounded: 34.83% keeps s unrounded, 34.84% rounds s to 1.324 first, 34.81% rounds the variance to 1.75 first, and 34.74% rounds s to 1.32 first.

Why: The CV puts the spread in proportion to the mean, so it has no units. Here the typical gap from the mean is about 35% of the mean itself. Keep four decimals in s/x¯ (0.3483) before you turn it into a percent, the same way the CV example in your Lecture 03 notes keeps 0.2693 to get 26.93%. If you round the ratio to 0.35 first, you get 35%. The method is right, but the answer is rougher than it needs to be, so round only at the end.

Where it's covered: Lecture 03: Coefficient of variation.

Part (f): the percentage taking at least 5 courses (1 mark)

  1. "At least 5" means 5 or more, so it includes 5. That's x=5 and x=6.
  2. 8+2=10 students
  3. 1030=0.3333, which is 33.33% -- divide by n=30

Answer: 33.33% of the students take at least 5 courses.

Why: "At least" and "at most" both include the endpoint. "More than 5" would leave out the 5s and give only 230=6.67%.

Where it's covered: Lecture 03: Percentiles from a frequency table (the "at least five pairs" example is at the end) and Lecture 04: Using the terms together (the "at least" warning).

Question 17

Part 1, parts (a) and (b), is built from Chapter 3 slides 9 to 14 and was taught on Sep 22. Parts (c) to (e) are built from Chapter 3 slides 16 to 23 and textbook 3.1 to 3.4, and aren't taught yet. Part (f) uses the Venn terms from slides 4 to 7, taught on Sep 22, plus the addition rule from slides 14 and 15, which was only previewed.

Part (a): exactly two economics majors (1 mark)

  1. The order of committee members doesn't matter, so use combinations.
  2. C412=12!4!8!=12×11×10×94×3×2×1=1188024=495 -- all possible committees
  3. Favourable: 2 of the 5 majors and 2 of the 7 others. Check: 5+7=12 and 2+2=4.
  4. C25=5×42=10 and C27=7×62=21
  5. P=C25×C27C412=10×21495=210495=0.4242 -- favourable over total

Answer: 0.42

Why: This is the officers problem from class with new numbers. Each way to pick the two majors can pair with each way to pick the two others, so the counts multiply. Then classical probability divides by all the possible committees.

Where it's covered: Lecture 04: Splitting the sample into groups and Lecture 04: Combinations.

Part (b): at least one economics major (1 mark)

  1. "At least one" means 1, 2, 3 or 4 majors. "None" is a single case, so use the complement rule: P(at least one)=1−P(none).
  2. P(none)=C47×C05C412 -- all 4 from the 7 non-majors and 0 from the 5 majors
  3. C47=7×6×5×44×3×2×1=84024=35 and C05=1
  4. P(none)=35×1495=0.0707
  5. 1−0.0707=0.9293 -- don't stop at 0.0707

Answer: 0.93

Why: This is the purple-car problem from class with new numbers. The complement of "at least one" is "none", and the two probabilities add to 1. She warned: "you have to remember that you have to subtract that from 1."

Where it's covered: Lecture 04: Complement rule.

Parts (c) to (e) are not taught yet. They are expected on Sep 24 (Sep 29 at the latest).

Part (c): the probability table (2 marks)

  1. Given: P(J)=0.60, P(C)=0.45 and P(J∩C)=0.30. That's one cell and two totals.
  2. P(J¯)=1−0.60=0.40 and P(C¯)=1−0.45=0.55 -- the complement rule gives the other two totals
  3. P(J∩C¯)=0.60−0.30=0.30 -- the J row has to add up to 0.60
  4. P(J¯∩C)=0.45−0.30=0.15 -- the C column has to add up to 0.45
  5. P(J¯∩C¯)=0.40−0.15=0.25 -- the J¯ row has to add up to 0.40
  6. 0.30+0.30+0.15+0.25=1.00 -- check: the four cells cover everyone

Answer: P(J∩C)=0.30, P(J∩C¯)=0.30, P(J¯∩C)=0.15 and P(J¯∩C¯)=0.25. The totals are 0.60 and 0.40 for the rows, and 0.45 and 0.55 for the columns.

The completed table:

C C¯ Total
J 0.30 0.30 0.60
J¯ 0.15 0.25 0.40
Total 0.45 0.55 1.00

Why: Each cell is a joint probability ("and"), and each total is a marginal probability. J and J¯ split everyone into two groups with no overlap, so the cells in a column add up to its total. The same goes for the cells in a row. So once you have one cell and the totals, the rest is subtraction.

Where it's covered: Not taught yet. See Chapter 3 slide 16 (probability table), slides 22 and 23 (joint and marginal probabilities) and textbook 3.4, Contingency Tables and Probability Trees (PDF page 165).

Part (d): conditional probability (1 mark)

  1. "Owns a car, given that they have a part-time job" is P(C∣J). The event you're told has happened goes after the bar.
  2. P(C∣J)=P(C∩J)P(J) -- the conditional probability formula
  3. =0.300.60=0.50

Answer: 0.50

Why: Knowing the student has a job shrinks the group you're looking at to the 60% with jobs. Of that 60%, the 30% who also own a car make up half. The common slip is flipping it: P(J∣C)=0.300.45=0.67 answers "has a job, given they own a car", which is a different question. This is the same setup as her used-car example (AC and CD player).

Where it's covered: Not taught yet. See Chapter 3 slides 17 and 18 (conditional probability) and textbook 3.1, Terminology (PDF page 150, where the section explains conditional probability).

Part (e): independence (1 mark)

  1. Chapter 3 slide 21: two events are statistically independent if and only if P(J∩C)=P(J)P(C).
  2. P(J)P(C)=0.60×0.45=0.27
  3. P(J∩C)=0.30, and 0.30≠0.27 -- the check fails, so the events are not independent
  4. Extra-practice check (from textbook 3.2, not on the slides): if the events were independent, knowing about the job wouldn't change the chance of owning a car. So P(C∣J) would equal P(C).
  5. P(C∣J)=0.50 from part (d), but P(C)=0.45, and 0.50≠0.45 -- the second check fails too

Answer: No. Having a part-time job and owning a car are not independent, because P(J∩C)=0.30 is not equal to P(J)P(C)=0.27. The extra-practice check agrees: P(C∣J)=0.50 is not equal to P(C)=0.45.

Why: Independent means that knowing one event tells you nothing about the other. Both checks say the same thing: students with jobs are a bit more likely to own a car (50% against 45% overall). The two checks always agree. The mark is for the slide 21 check: work out P(J)P(C) and compare it with P(J∩C). The textbook says either check alone proves it. The second one is here for practice, because slide 21 only shows the first. Re-check this after the Sep 24 class: if she shows the P(C∣J)=P(C) version in class, it's fair game too. Don't mix independence up with mutually exclusive. That's a different question, and it's part (f).

Where it's covered: Not taught yet. See Chapter 3 slide 21 (statistical independence) and textbook 3.2, Independent and Mutually Exclusive Events (PDF page 154). The slide gives the first check. The textbook gives both, and says you only need to show one.

Part (f) is half taught. The Venn terms were taught on Sep 22, but the "neither" share needs the addition rule, which was only previewed, like Question 14. So it's scored with the not-taught marks. You don't need part (c) for it: steps 1 to 5 use the table, and steps 6 to 10 get the same answer from a Venn diagram.

Part (f): mutually exclusive and collectively exhaustive (1 mark)

With your table from part (c):

  1. Mutually exclusive means J∩C=∅: no student has both. So look at the J∩C cell of the table.
  2. P(J∩C)=0.30, which isn't 0 -- 30% of students have a job and a car, so J and C are not mutually exclusive
  3. Collectively exhaustive means J∪C=S: every student has a job or a car, or both. So look at the "neither" cell, J¯∩C¯.
  4. P(J¯∩C¯)=0.25, which isn't 0 -- 25% of students have neither, so J and C are not collectively exhaustive
  5. P(J∪C)=1−0.25=0.75 -- check: the union is less than 1, so it doesn't cover all of S. The addition rule gives the same thing: 0.60+0.45−0.30=0.75.

Without the table (if you skipped part (c)):

  1. Draw two overlapping circles, J and C, inside the rectangle S. The overlap is J∩C, and it holds 0.30.
  2. Job only: 0.60−0.30=0.30 -- the whole J circle minus the overlap
  3. Car only: 0.45−0.30=0.15 -- the whole C circle minus the overlap
  4. P(J∪C)=0.30+0.30+0.15=0.75 -- job only, plus both, plus car only, so the overlap is counted once
  5. Neither: 1−0.75=0.25 -- what's left of the rectangle outside both circles. The overlap (0.30) and the outside (0.25) are both above 0, which is the same result as steps 2 and 4.

Answer: Neither. J and C are not mutually exclusive, because 30% of students are in both. They are not collectively exhaustive, because 25% of students are in neither.

Why: These are the same two checks as Question 11, with probabilities in place of lists of outcomes. "Mutually exclusive" asks whether the intersection is empty. "Collectively exhaustive" asks whether the union is all of S. A probability above 0 means that part of the Venn diagram has students in it. The common slip is adding 0.60+0.45=1.05, seeing a total above 1, and calling the events collectively exhaustive. That sum counts the 30% who do both twice, and a probability above 1 is a sign something went wrong. Compare J and J¯: they are both mutually exclusive and collectively exhaustive, like any event and its complement. That's why the two cells in each column of the table add up to the column total. The Venn route in steps 6 to 10 is the addition rule drawn out: adding the two whole circles counts the overlap twice, so you take it off once.

Where it's covered: Lecture 04: Mutually exclusive events, Lecture 04: Collectively exhaustive events, Lecture 04: Using the terms together and Lecture 04: Addition rule preview (for step 5 and the Venn route).

If you got these wrong