EC 285 - 2026-10-06 - Lecture 08

EC 285 - Lecture 08 - 2026-10-06

Introductory Statistics · Zara Liaqat · SB 106

Week 5 · 16:01-17:17 · 76 min · Full transcript · Flashcards

Overview

This lecture opened with Midterm 1 logistics and the results of the mid-semester feedback survey from another section, then returned to Chapter 4, Discrete Random Variables. After a quick recap of last week's two-coin distribution and a fuller preview of portfolio analysis, it covered the Bernoulli distribution, a single trial with two outcomes, and derived its mean and variance. It then built the binomial distribution, which is the Bernoulli trial repeated n times, and worked three binomial problems, including how to recognise a binomial question when the word "binomial" never appears. The Poisson distribution was left for Thursday.

Picking Chapter 4 back up

This recaps Lecture 06: the two-coin table (X = 0, 1, 2 heads with probabilities 1/4, 1/2, 1/4), its expected value and variance, and linear functions of random variables. The lecturer added detail to the portfolio example.

Portfolio analysis preview

You have money to invest, for example $1,000, and split it between two stocks X and Y in some ratio, such as 0.5 and 0.5 or 0.7 and 0.3. The portfolio's return is a linear function W=aX+bY, where a and b are the shares. Because the returns on X and Y are random, W is random too.

Why: Historical average returns are publicly available for publicly listed companies, so you can use them to estimate the expected return on your whole portfolio, and the covariance of X and Y. A risky stock may pay a higher average return, while a safer stock with a smaller standard deviation usually pays less. How someone splits their money depends on how much risk they like, but that is microeconomic theory, so the course skips it. "I will give you the allocation ratios, so you don't have to." Worked numbers come next class.

The Bernoulli distribution

This is the first named distribution on the Chapter 4 slides. It follows the three-step pattern from Lecture 06: the probability function, the expected value and the variance.

Bernoulli trials

A Bernoulli experiment is done once and has only two mutually exclusive outcomes, called "success" and "failure". X = 1 for a success and X = 0 for a failure. Let p be the probability of success, so the probability of failure is 1 − p by the complement rule.

Why: Any yes-or-no event becomes a number you can average once you code it as 0 or 1.

P(x)=px(1−p)1−x,x=0,1

where x is 0 (failure) or 1 (success) and p is the probability of success. This gives P(X=1)=p and P(X=0)=1−p, which is how the lecturer wrote it on the board.
On the formula sheet for: Midterm 2, final.

Bernoulli mean and variance

The average of a Bernoulli variable is the probability of success, and its variance is p(1−p).

Why: X is 1 a fraction p of the time and 0 the rest of the time, so on average it equals p. The spread is largest when success and failure are equally likely, at p=0.5, and shrinks toward zero as the outcome becomes certain.

E[X]=pVar[X]=p(1−p)

where p is the probability of success.
On the formula sheet for: Midterm 2, final.

Example: Deriving the Bernoulli mean and variance

The lecturer derived the mean on the board from the table x=0,1 with P(0)=1−p and P(1)=p. For the variance, the lecturer wrote step 3 and said that simplifying it gives p(1−p). Steps 4 to 6 fill in that simplification.

  1. E[X]=∑xP(x)=0⋅(1−p)+1⋅p -- multiply each value by its probability and add
  2. E[X]=0+p=p -- the first term is zero
  3. Var[X]=∑(x−μ)2P(x)=(0−p)2(1−p)+(1−p)2p -- average squared distance from the mean, using μ=p
  4. =p2(1−p)+p(1−p)2 -- (0−p)2=p2, and reorder the second term
  5. =p(1−p)[p+(1−p)] -- factor out p(1−p)
  6. =p(1−p)⋅1=p(1−p) -- the bracket equals 1

Answer: E[X]=p and Var[X]=p(1−p) -- these match the formula sheet.

Made-up example: Rolling a 6

One die roll, where a 6 counts as success, so p=1/6.

  1. E[X]=p=16≈0.167 -- the mean is the chance of success
  2. Var[X]=p(1−p)=16×56=536≈0.139 -- put p into the variance formula
  3. σ=5/36=56≈2.2366≈0.373 -- the standard deviation is the square root of the variance

Answer: Mean about 0.167 and standard deviation about 0.373 -- on average a single roll produces 0.167 "sixes".

Exam

"Obviously, I won't ask you for proofs of these formula that you see on the formula sheet." The formulas are given, but "you still need to know what n is, what p is." Bernoulli questions are limited because it is a one-time experiment, but "sure, I can ask you a question on this."

The binomial distribution

The binomial distribution is the Bernoulli experiment repeated. In the lecturer's words: "You don't toss a coin one time, you do it multiple times. That's the only difference."

Binomial distribution

Repeat a two-outcome experiment n times, with the same success probability p each time. X is the number of successes, so x can be 0, 1, ..., n. The binomial formula gives the probability of exactly x successes.

Why: The first part, n!x!(n−x)!, is the combinations formula from Chapter 3. It counts how many ways the x successes can be placed among the n trials, and order doesn't matter. The rest, px(1−p)n−x, is the probability of any one of those arrangements. It is an intersection, like "pass, fail, fail, fail, fail", so you multiply x success probabilities by n − x failure probabilities.

P(x)=n!x!(n−x)!px(1−p)n−x

where n is the number of times the experiment is repeated (the number of observations), x is the number of successes, p is the probability of success, and 1 − p is the probability of failure. The slides write the success probability as capital P. The formula sheet and the lecturer use small p.
On the formula sheet for: Midterm 2, final.

Example: One pass among five students

Suppose the midterm was so hard that the chance of passing is only 10%, so p=0.1. Among 5 students (n=5), what is the probability that exactly one passes (x=1)?

  1. P(X=1)=5!1!4!(0.1)1(0.9)4 -- put the numbers into the binomial formula, with 1−p=0.9 and n−x=4
  2. 5!1!4!=5×4!4!=5 -- five ways to choose which student passes
  3. (0.9)4=0.9×0.9×0.9×0.9=0.6561 -- the probability of four failures
  4. 5×0.1×0.6561=0.32805 -- combinations times the probability of one arrangement

Answer: P(X=1)≈0.328 -- there is about a 33% chance that exactly one of the five passes.

Exam

The first part of the formula is "combinations. Not permutations, combinations," because "I just want one to pass out of five, so order doesn't matter." The powers must add up to n: here there is one success and four failures, out of five students.

Binomial mean and variance

The average number of successes in n trials is np, and the variance is np(1−p). A Bernoulli variable is the special case of a binomial with n=1.

Why: Each trial adds p to the expected count and p(1−p) to the variance, so n trials give n times each. The lecturer stressed the link: both distributions have two mutually exclusive outcomes with p and 1 − p, and "if n is 1, you get the exact same formula, p and p times 1 − p, because you're only doing it once."

E[X]=npVar[X]=np(1−p)σ=np(1−p)

where n is the number of trials and p is the probability of success.
On the formula sheet for: Midterm 2, final. The square root for σ is not printed, so take it yourself.

Example: Standard deviation of throws made

A sports player makes 80% of their throws in a season, so p=0.8. Over the next 8 throws (n=8), what is the standard deviation of the number of throws they make?

  1. Var[X]=np(1−p)=8×0.8×0.2 -- put the numbers into the binomial variance, with 1−p=0.2
  2. 8×0.8=6.4, then 6.4×0.2=1.28 -- the variance
  3. σ=1.28≈1.131 -- the standard deviation is the square root of the variance

Answer: σ≈1.13 throws -- the number made in 8 throws typically varies by about 1.13 from its average.

Exam

Several students answered 1.13. One answered 1.28, and the lecturer said "I know where you made a mistake." The value 1.28 is the variance: when the question asks for the standard deviation, take the square root. The lecturer also advised starting every problem by stating the random variable: "Always a good idea to start with that."

Spotting a binomial question

An exam question may not say "binomial", so you have to recognise the setup. Look for these four features:

Why: "Once you know, you have the formula, you can apply it." Recognising the distribution is the real skill, especially by the final, when about 10 distributions have been covered. Note that "success" only means the outcome you are counting. Finding your job stressful is not nice, "but that's how the question is framed."

Example: Stressed professors

A survey found that 40% of professors consider their job extremely stressful. If 12 professors are selected at random, what is the probability that exactly 5 consider their job stressful? Here n=12, x=5, p=0.4 and 1−p=0.6.

  1. P(X=5)=12!5!7!(0.4)5(0.6)7 -- five find it stressful and the other seven do not
  2. 12!5!7!=12×11×10×9×85×4×3×2×1=95040120=792 -- the number of ways to choose 5 of the 12
  3. (0.4)5=0.01024 -- the probability of 5 successes
  4. (0.6)7=0.0279936 -- the probability of 7 failures
  5. 792×0.01024×0.0279936≈0.2270 -- multiply the three parts

Answer: P(X=5)≈0.227 -- about a 22.7% chance that exactly 5 of the 12 find their job stressful.

Exam

"When you read the question, it's not very obvious that it's binomial distribution. So I may not mention that in the question." In class, a student guessed that it was binomial "because there are more than two people or two outcomes". That is wrong. It is binomial because there are only two outcomes (stressful or not) and the count is discrete: "I cannot have 5.2 professors finding their job to be stressful."

Binomial tables

The slides include a binomial table that lists P(x) for given n, x and p. You will not get these tables in exams.

Why: Printing them would be about 10 sheets of paper per exam "when you would only look for this one number". The table values are just the binomial formula worked out in advance, so you can get the same numbers by hand.

Example: Slide table lookup, n = 10, x = 3, p = 0.35

The slide reads 0.2522 from the table. Here is the same value from the formula, with 1−p=0.65.

  1. P(X=3)=10!3!7!(0.35)3(0.65)7 -- put the numbers into the formula
  2. 10!3!7!=10×9×83×2×1=120 -- the combinations
  3. (0.35)3=0.042875 and (0.65)7≈0.049022 -- the success and failure parts
  4. 120×0.042875×0.049022≈0.2522 -- multiply

Answer: P(X=3)≈0.2522 -- this matches the table.

Example: Slide table lookup, n = 10, x = 8, p = 0.45

The slide reads 0.0229 from the table. Here is the same value from the formula, with 1−p=0.55.

  1. P(X=8)=10!8!2!(0.45)8(0.55)2 -- put the numbers into the formula
  2. 10!8!2!=10×92×1=45 -- the combinations
  3. (0.45)8≈0.0016815 and (0.55)2=0.3025 -- the success and failure parts
  4. 45×0.0016815×0.3025≈0.0229 -- multiply

Answer: P(X=8)≈0.0229 -- this matches the table.

Exam

"Don't rely on these tables. Just work out the probabilities using the formula... these will not be provided, although they're in the slides." The syllabus says statistical tables are provided during exams, but the lecturer was clear that binomial tables are not among them.

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 - 2026-10-06 - Lecture 08.opus.