EC 285 - 2026-10-01 - Lecture 07

EC 285 - Lecture 07 - 2026-10-01

Introductory Statistics · Zara Liaqat · SB 106

Week 4 · 16:00-17:15 · 74 min · Full transcript · Flashcards

Overview

This class was a review session for Midterm 1 instead of new Chapter 4 material. It opened with the exam logistics and a study plan. The class then worked through a 15-question multiple-choice practice set from Chapters 1 to 3, which led to a walk through the Midterm 1 formula sheet. It ended with the harder numerical practice question on the median, quartiles and the interquartile range.

Probability tables and the addition rule

This revisits the probability table and addition rule from Lecture 05, through one of the practice multiple-choice questions.

Probability table

A grid that splits the whole group by two characteristics at once, with row totals, column totals and a grand total. You fill the empty cells from the few numbers the question gives you, using the fact that each row and column must add up to its total.

Why: The lecturer called it the "best way to do this." Once the table is full, every probability you need (each event alone, and both together) can be read straight off it.

Exam

On a numerical question, "you have to show the table and you have to show your calculations." It's fine if you don't remember the rule's name. On a multiple-choice question you can do the table on scrap paper and just pick the option.

Addition rule

The probability that A or B (or both) happens is the probability of A plus the probability of B, minus the probability that both happen.

Why: Outcomes in both A and B get counted once in P(A) and again in P(B). Subtracting the overlap means they are counted only once.

P(A∪B)=P(A)+P(B)−P(A∩B)

where A∪B means "A or B" and A∩B means "A and B".
Not on any formula sheet, so memorise it.

Example: Probability a student earned a B or is a freshman

A class has 20 freshmen and 30 sophomores. 15 students earned a B, including 5 freshmen. Find P(B∪F), where B is "earned a B" and F is "freshman".

  1. 20+30=50 -- total number of students
  2. B∩Sophomore=15−5=10 -- the B students who are not freshmen
  3. B¯∩F=20−5=15 and B¯∩Sophomore=30−10=20 -- the students without a B in each year
  4. B¯ total=15+20=35 -- check: 15+35=50
  5. P(B)=1550, P(F)=2050, P(B∩F)=550 -- read straight off the table
  6. P(B∪F)=1550+2050−550 -- addition rule with the numbers in
  7. =3050=0.6 -- simplify

Answer: 0.6 (option C) -- 60% of the students either earned a B or are freshmen.

Exam

Some students answered 0.7. That is 3550, which comes from forgetting to subtract the overlap. "You're supposed to subtract the intersection part, because you don't want to count it twice."

Frequency tables

Two practice questions used frequency tables, like the shoe example from Lecture 03.

Cumulative percentage

The cumulative percentage through a class is the share of all observations that fall in that class or any class before it. "Class" means one row or interval of the table: first class, second class and so on.

Why: It works the same way as the cumulative probability function from Lecture 06, where the last value has to be 1. Here the last cumulative frequency equals n, which is 100%.

Cumulative % through class k=cumulative frequency through class kn×100

where n is the total number of observations, which is the sum of all the frequencies.
Not on any formula sheet, so memorise it.

Example: Cumulative percentage through the second class (practice question 8)

The class frequencies are 5, 9, 4 and 2. The values of X themselves are not given. Find the cumulative percentage through the second class.

  1. n=5+9+4+2=20 -- total observations
  2. Cumulative frequencies: 5, 5+9=14, 14+4=18, 18+2=20 -- running totals; the last one equals n
  3. 1420×100=70% -- cumulative frequency through class 2 as a percentage

Answer: 70% (option C) -- 70% of the observations are in the first two classes. In the lecturer's words, if X were income brackets, "70% of the population earns this much or less."

Exam

It's the X values (the classes) that must be in order, not the frequencies. Don't re-sort the frequencies into 2, 4, 5, 9. "There's a numerical question later on, that does the same thing."

Weighted mean

The mean of data given as a frequency table: multiply each value by its frequency, add these up, and divide by the total number of observations.

Why: The values are not equally common, so each one has to count as many times as it occurred. Just averaging the distinct values (0+1+23) would ignore the frequencies.

x¯=∑fixin,n=∑fi

where xi is each value, fi is how many times it occurs, and n is the total number of observations.
Not on any formula sheet, so memorise it.

Example: Mean number of errors

The number of errors x is 0, 1 or 2, with frequencies 4, 7 and 9. Find the mean number of errors.

  1. n=4+7+9=20 -- total observations
  2. ∑fixi=0⋅4+1⋅7+2⋅9 -- each value times its frequency
  3. =0+7+18=25 -- add them up
  4. x¯=2520=1.25 -- divide by n

Answer: 1.25 errors -- on average, each observation had 1.25 errors.

Exam

The word "weighted" may or may not appear: "even if I skipped the term weighted, it's still a valid question and it's the same formula." Whenever the data come with frequencies, the frequency matters.

The Midterm 1 formula sheet

The formula sheet is attached to the exam. You don't bring your own. It's posted on MyLS under Content. The lecturer stressed that it gives formulas with "no explanation whatsoever": it never says what n or x means, so you have to know that yourself.

Permutations

A permutation counts the ways to choose x items out of n when order matters, like arrangements or distinct positions. A combination counts selections where order does not matter.

Why: A permutation counts every ordering of the same people as a different outcome. That is why it is bigger than the combination by a factor of x!, the number of ways to order the x chosen items.

Pxn=n!(n−x)!Cxn=Pxnx!=n!x!(n−x)!

where n is the number of items to choose from and x is how many are chosen. The sheet labels the combination Ckn but uses x on the right side; they mean the same thing.
On the formula sheet for: Midterm 1, Midterm 2 (not the final).

Example: Choosing a president, vice president and treasurer (practice question 12)

Three people are chosen from 5 for three distinct positions: president, vice president and treasurer. How many ways can this be done?

  1. Order matters -- the same person could be president, vice president or treasurer, and each is a different outcome
  2. P35=5!(5−3)!=5!2! -- permutation formula with n = 5, x = 3
  3. =5⋅4⋅3⋅2⋅12⋅1=5⋅4⋅3 -- cancel the 2!
  4. =60 -- multiply
  5. Wrong route for comparison: C35=603!=606=10 -- this would ignore order

Answer: 60 (option E) -- there are 60 ways to fill the three posts. The combination answer, 10, is a trap.

Exam

The question had no keyword such as "arrangements" or "order matters." You have to use "common sense": distinct positions mean order matters. "Please study the formula sheet beforehand so that you're not surprised when you actually see these letters." A calculator's nPr or nCr button is fine. On a numerical question you don't have to copy the formula out, but showing work earns partial credit.

Coefficient of variation

The standard deviation as a percentage of the mean.

Why: It is a unit-free measure of spread, so it lets you compare variability between data sets with very different means or units.

CV=(sx¯)⋅100%

where s is the sample standard deviation and x¯ is the sample mean.
On the formula sheet for: Midterm 1 only.

Made-up example: Coefficient of variation

A sample has mean x¯=16 and standard deviation s=4.

  1. CV=(416)⋅100% -- put the numbers in
  2. =0.25⋅100% -- divide
  3. =25% -- convert to a percentage

Answer: 25% -- the standard deviation is a quarter of the mean.

Exam

"I don't have any sample question in my slides today on CV or coefficient of variation, but you should know what that is." It "could appear on the exam."

Covariance

A measure of whether two variables move together. It is positive when they tend to be above (or below) their means at the same time, and negative when one tends to be high while the other is low.

Why: Each product of deviations is positive when x and y are on the same side of their means, and negative when they are on opposite sides. The average of those products shows which pattern dominates.

σxy=∑i=1N(xi−μx)(yi−μy)Nsxy=∑i=1n(xi−x¯)(yi−y¯)n−1

where the left one is the population covariance (N values, population means μ) and the right one is the sample covariance (n values, sample means x¯ and y¯, dividing by n − 1 as with the sample variance in Lecture 03).
On the formula sheet for: Midterm 1, Midterm 2, but only the population version. The sample version is not on any sheet, so memorise it. The lecturer said the sample covariance is "the exact same formula" and the sheet "helps to remind you," but the sheet's version uses μ and divides by N.

Made-up example: Covariance of three pairs

The pairs (x, y) are (1, 2), (2, 4) and (3, 6).

  1. μx=1+2+33=2, μy=2+4+63=4 -- the means
  2. Deviations of x: −1,0,1; deviations of y: −2,0,2 -- subtract the means
  3. Products: (−1)(−2)=2, (0)(0)=0, (1)(2)=2; sum =4 -- multiply the pairs and add
  4. Population: σxy=43≈1.33 -- divide by N = 3
  5. Sample: sxy=43−1=2 -- divide by n − 1 = 2

Answer: 1.33 as a population, 2 as a sample -- both are positive, so x and y move together.

Exam

"You might need sample covariance." The sheet only prints the population one.

Correlation coefficient

Covariance divided by the product of the two standard deviations. It always lies between −1 and 1.

Why: Dividing by the standard deviations removes the units, so the size of the number means something. Values near 1 or −1 are a strong straight-line relationship, and values near 0 are no straight-line relationship.

ρ=Cov(x,y)σxσy

where Cov(x,y) is the population covariance and σx and σy are the population standard deviations.
On the formula sheet for: Midterm 1, Midterm 2 (population version only).

Made-up example: Correlation of the same three pairs

Use the pairs (1, 2), (2, 4), (3, 6) from the covariance example, where σxy=43.

  1. σx2=(−1)2+02+123=23 -- population variance of x
  2. σy2=(−2)2+02+223=83 -- population variance of y
  3. σxσy=23⋅83=169=43 -- product of the standard deviations
  4. ρ=4/34/3=1 -- divide

Answer: ρ=1 -- a perfect positive straight line, which makes sense because y = 2x exactly.

Normal density formula

The probability density function of the normal distribution, a Chapter 6 topic for continuous random variables.

f(x)=12πσ2e−(x−μ)2/2σ2

where μ is the mean and σ2 is the variance.
On the formula sheet for: Midterm 1, Midterm 2, final.

Exam

"We never use this. Ever, like even for the final exam. It's just there to confuse you, basically." Normal probabilities come from the table instead. Don't try to plug numbers into it.

Terms from the practice questions

Several multiple-choice questions tested Chapter 1 and 3 vocabulary. The multiple-choice section can be conceptual as well as numerical.

Data terms

Sampling methods

The sampling methods covered in class, as in Lecture 02. For the sampling question, the lecturer said of two options: "I just made up these terms. Convenience sampling, no such thing. Judgment sampling, no." (Outside this course, convenience sampling and judgment sampling are real non-random sampling methods that most textbooks describe. For this exam, go with the lecturer: they were not taught in this course, so treat them as wrong options.) She also named stratified random sampling as an example of a slide term to look up in the book if you missed that class.

Exam

Expect made-up method names as wrong options. Stick to the methods on the slides and in your notes.

Collectively exhaustive events

Events are collectively exhaustive when together they cover the whole sample space. In the lecturer's words, "there's no element in the sample space which is not part of either A or B."

Why: If the events cover everything, at least one of them must happen, so their union has probability 1.

Exam

A student asked about this during the practice. Her answer was: "It's been defined in the class. It's on the lecture slides. Chapter three. So you should know it."

Median, quartiles and the IQR

This is numerical practice question 2. The lecturer went through it because it is harder than the version done in Lecture 03. It extends the percentile method from that class to data with an even number of values.

Median

The middle value of the sorted data. When n is even there are two middle values, and you take their average.

Why: Averaging the two middle values gives one specific number. The same in-between idea is what makes the quartile method below work.

Median position=n+12

where n is the number of observations, and the data are sorted in ascending order.
Not on any formula sheet, so memorise it.

Example: Median of 8 values

8 values are sorted in ascending order. Positions 2 to 7 are 18, 21, 24, 25, 29, 31. The first and last values were not read out in class.

  1. 8+12=4.5 -- the median sits halfway between the 4th and 5th values
  2. 4th value =24, 5th value =25 -- read from the sorted data
  3. 24+252=24.5 -- average the two middle values

Answer: 24.5 -- half the values are below 24.5 and half are above.

Exam

"Some books allow you to write both" middle numbers, but "my preferred method is to take the average, one specific number."

Quartiles

Q1 is the 25th percentile and Q3 is the 75th percentile. First find the position in the sorted data. If the position is not a whole number, move that fraction of the way between the two neighbouring values.

Why: The position tells you where to look, not the answer itself. A position like 2.25 means a quarter of the way from the 2nd value to the 3rd value.

Lp=p100(n+1)

where p is the percentile (25 for Q1, 75 for Q3), n is the number of observations, and Lp is the position in the sorted data.
Not on any formula sheet, so memorise it.

Example: Q1 and Q3 of the same 8 values

Same sorted data: positions 2 to 7 are 18, 21, 24, 25, 29, 31, and n = 8.

  1. L25=0.25×(8+1)=0.25×9=2.25 -- Q1 is a quarter of the way from the 2nd value (18) to the 3rd value (21)
  2. Q1=18+0.25×(21−18)=18+0.75=18.75 -- add a quarter of the gap to the 2nd value
  3. L75=0.75×9=6.75 -- Q3 is three quarters of the way from the 6th value (29) to the 7th value (31)
  4. Q3=29+0.75×(31−29)=29+1.5=30.5 -- add three quarters of the gap to the 6th value

Answer: Q1=18.75 and Q3=30.5 -- a quarter of the data are below 18.75, and three quarters are below 30.5.

Exam

"Step number one is to arrange your data in ascending order. That's important. Otherwise all your numbers will be wrong." The position (2.25, 6.75) "is not the answer. It's just telling us where to look." Chapter 2 of the textbook has many more examples if you need practice.

Interquartile range

The third quartile minus the first quartile. It measures the spread of the middle 50% of the data.

Why: It ignores the bottom and top quarters, so a few extreme values can't distort it.

IQR=Q3−Q1

where Q3 is the 75th percentile and Q1 is the 25th percentile.
Not on any formula sheet, so memorise it.

Example: IQR of the same 8 values

From the quartiles example, Q1=18.75 and Q3=30.5.

  1. IQR=30.5−18.75 -- subtract Q1 from Q3
  2. =11.75 -- arithmetic

Answer: 11.75 -- the middle half of the data spans 11.75 units.

Exam

"You should know what IQR is. I don't define it here," meaning the exam question won't define it. "The most commonly made mistake is 6.75 − 2.25": subtracting the positions instead of the quartile values.

Exam checklist

Admin & deadlines


Transcribed automatically from the lecture recording. Slides used: EC 285 - Ch. 04 - Discrete Random Variables (slides).pdf. Course info used: EC 285 - Formula Sheet - Final.pdf, EC 285 - Formula Sheet - Midterm 1.pdf, EC 285 - Formula Sheet - Midterm 2.pdf, EC 285 - Syllabus - Fall 2026.pdf. Audio archived at /mnt/porsche/configs/lectures/archive/2026-10/EC 285 - 2026-10-01 - Lecture 07.opus.