Chapters 1, 2, 3, 4 and 6. Chapter 5 is not on it.

• Chapter 6 (probability) is the hardest chapter, and it’s always on the test.
• Chapters 2 and 3 are mostly graphing, and you won’t draw graphs on the test.
• Parts of some chapters are left out. My lessons skip them for you.

My rule: if it’s not covered on my website, then it’s not on the exam.

How long to spend on each chapter: Term Test 1 study checklist.

• Closed book. All written questions, no multiple choice (true for at least the last five years).
• Usually three short-answer questions, each with several parts.
• Some parts are quick: a definition or one line of theory.
• Yes, there are part marks. Markers accept answers within a range, so small rounding differences won’t cost you.

To collect every part mark, write each step separately:

1. Write the formula.
2. Plug in the numbers.
3. Calculate.
4. Finish with one sentence saying what the answer means.

If the steps run together, the marker can’t give you part marks.

• Sunday 18 October, 10:00 to 11:30am, on the Keele campus.
• Every section writes the same test at the same time.
• Bring your student ID. It is required.
• Term Test 2 is Sunday 8 November, same time.

• Yes. One 8.5 x 11 sheet, one side, for each term test (two sides for the final). The course outline calls it a reference sheet.
• Printed sheets are fine.
• Yes, you can use mine. Students have brought it in for more than 44 semesters in a row without a problem, and a professor confirmed it again this term.

▶ Download the free Term Test 1 cheat sheet

• A basic scientific calculator (non-programmable). About $25 is plenty.
• No graphing calculators. Students have been made to put them away.
• Make sure it has an x-root key (ˣ√y). You need it for the geometric mean on this test.
• Nice to have for later: a Σxy readout (it adds up lists of x and y values for you). It saves a lot of time on the final.

Set-up tips: fixed decimals (4 places), not scientific notation. Work one step at a time instead of typing the whole formula at once. Lost the manual? Search the model number plus “manual”.

• 4 decimal places. The test usually says so on page 1.
• Keep more decimals while you work. Round only the final answer.
• Off by one in the 4th decimal? That is normally still marked correct.

When marks come back, and any adjustment, is up to your professor.

• Yes. It is worth 5%.
• It is taken at random during lectures. How often depends on your professor.

• Stats Doesn’t Suck has 17 solved past ADMS 2320 tests and exams: 8 past Term Test 1s, 5 past Term Test 2s and 4 past finals.
• For Term Test 1 alone that is over 100 solved past test questions, most with video solutions.
• Every chapter lesson is built from questions taken from past tests.

My advice: focus on past tests. A lot of textbook questions have nothing to do with the kinds of questions that will be on your exam.

▶ Try a full solved past Term Test 1, free · Everything for ADMS 2320

• Yes: PASS (Peer-Assisted Study Sessions), free, run by York.
• Fall 2026: Monday and Wednesday, 1 to 3pm, Founders College 015A.
• Both days use the same worksheet, so go to whichever fits. No sign-up needed.
• In person only (not online or recorded).

▶ ADMS 2320 PASS page

• The course uses Keller, Statistics for Management and Economics, 11th edition.
• Students mostly use it for practice questions. Many skip it and study from past tests instead.

“I didn’t use the textbook and his videos described each topic really well.” (student, 2025)

“sds practice exams set you up better than textbook questions” (student, 2023)

• Free: the chapter 1 to 4 lessons, the cheat sheet and one full solved past Term Test 1.
• Term Test 1 Prep, $49: adds the chapter 6 lessons and all the solved past Term Test 1s.
• Full-course bundle, $199: everything for Term Test 1, Term Test 2 and the final.

▶ Term Test 1 Prep · Full-course bundle

Stats Doesn’t Suck is an independent tutoring service, not affiliated with York University.

• Your section doesn’t change what’s tested. Everyone writes the same tests.
• Fall 2026 sections: Ali (A, I), Bartel (B, C, K), Salman (D, G), Rochon (E, H).
• For years, the most popular professor with students, by far, has been Professor Rochon.
• The deadline to switch sections this term (last day to add, 22 September) has passed.

• There are no make-up dates.
• A missed term test moves its 30% onto the final, which then covers the whole course.
• Talk to your professor right away about what documentation they need.

Don’t skip a test on purpose to “write the final instead”. It looks like an escape. It isn’t:

• It’s the whole course on one exam. Not many people pass it.
• No extra cheat sheet room. Still one page, two sides.
• Studying harder later doesn’t fix it. There is just too much to cover at once.

My advice: go and get whatever marks you can. Then buckle down and do as well as you can on the remaining exams. At least that way you spread it out.

• Tuesday 10 November, two days after Term Test 2.
• After that, the course stays on your record with a final grade.
• If you retake the course later, your Stats Doesn’t Suck access renews at half price.

• No. It is a few definitions. A few hours is enough.
• Make sure your cheat sheet has: population vs sample, parameter vs statistic, descriptive vs inferential, confidence vs significance level.

▶ Chapter 1 lessons (free)

• Population: the whole group you want to know about. Sample: the part you actually collect data from.
• Parameter describes a population. Statistic describes a sample.
• Wording trick: “it is known” or “it is understood” means a population value.

▶ Parameters vs statistics · Populations vs samples

• Significance level = alpha (α). Confidence level = 1 − α. So 95% confidence goes with a 5% significance level.
• Confidence level, in plain words: how confident you are that your range has captured the true value. In other words, how often this method would produce a range that contains the true value if you repeated the study many times. A 95% confidence level means that about 95 out of 100 such ranges would contain the true value.
• Significance level, in plain words: how much risk you are willing to accept of concluding that there is a real effect when there actually isn’t one. A 5% significance level means accepting about a 1-in-20 risk of this kind of false positive.
• For Test 1 you only need the definitions. You use them properly later in the course.

▶ Confidence vs significance levels

• Not very likely. Past tests have never asked students to draw a graph, though they sometimes ask you to interpret one.
• Expect a few short questions on Sturges’ rule and on nominal, ordinal and interval data.

• Unlikely. Past tests haven’t asked for an accurate pie chart.
• If you ever needed a protractor, your professor would tell you to bring one.
• At most, you may have to work out an angle: relative frequency x 360.

• Interval: real measurements or counts (income, marks, time). The gaps between values mean something, so you can do any math with them: add, subtract, average.
• Ordinal: categories with a natural order (ratings, rankings). They can look like numbers, such as 1st, 2nd and 3rd in a race, but the gaps aren’t equal: 1st and 2nd might be one second apart, 2nd and 3rd a full minute. So you can rank them, but you can’t average them.
• Nominal: categories with no order (country, colour, degree type). Numbers used as labels, like jersey numbers, are still nominal.

This is a simplification, and worked examples are the quickest way to get it.

▶ Types of data · Calculations for each type · Hierarchy of data

• Number of classes (Sturges): 1 + 3.3 log(n).
• Class width: (largest value minus smallest value) divided by the number of classes. Round it to a convenient value.

Does class width round up? Round to whatever is convenient:

• (99 − 16) / 9 = 9.22, rounded up to 10.
• (94 − 48) / 8 = 5.75 could be rounded to 6, but a width of 5 makes a histogram that’s easier to read, so 5 works too.
• There’s no strict rule. Sturges’ rule and the class width formula are guidelines only; what matters most is choosing classes that are easy to interpret. The one firm rule: the first class must contain the smallest value.

▶ Sturges’ formula · Class width

• A graph of the cumulative total. Cumulative frequency or cumulative relative frequency are both correct.
• On the test, it’s rarely worth more than a point. Don’t lose sleep over it.

The test tells you. On past tests, the question names the measure it wants: the variance, the standard deviation or the mean absolute deviation (MAD). The coefficient of variation is asked for as a relative measure of variation between two groups.

Cautions:

• Samples and populations have different formulas for finding the variance.
• MAD shows up frequently on past tests, so learn it.
• They won’t always use the word “variance”. They may say “variation”, “consistency” or sometimes “risk”.

▶ Worked MAD example (free past test, Q2b) · Variance · Standard deviation

• Sample and population variance use different formulas. Put both on your cheat sheet.
• Use the shortcut formula whenever you can. It’s faster. If you’re only given the sums of x and x², you must use it.
• Not sure if it’s a sample? On Term Test 1 it usually is.

Tip: always take the fastest route to the variance.

• Average of past rates of return: geometric mean. Predicting next year’s rate: the regular (arithmetic) mean.
• Make sure to convert percentages to decimals first when working with the geometric mean.
• You need the x-root key (ˣ√y) on your calculator.
• It has been on many past tests, so expect it.

▶ The geometric mean

• Mean: the usual average. Pulled toward extreme values.
• Median: the middle value. Best when the data is skewed or has outliers.
• Mode: the most common value. If nothing repeats, there is no mode.
• Put the data in order for the median and percentiles. Order doesn’t matter for the mean or variance.

▶ Measures of central tendency

1. Put the data in order.
2. Find the location of the percentile (where in the list to look).
3. The percentile is the value you find there.

• Quartiles are just the 25th, 50th and 75th percentiles.
• IQR = Q3 minus Q1. It can never give you a negative result.

▶ Location of a percentile · Percentiles · A past test IQR question

• They give the same answer. Mine plugs in the variances directly, because tests usually give you variances.
• Given standard deviations instead? Square them first.
• You will get full marks for using the version on the cheat sheet.

▶ Covariance · Coefficient of correlation

• Bell-shaped data: empirical rule. About 68%, 95% and 99.7% fall within 1, 2 and 3 standard deviations.
• Not bell-shaped, or you can’t tell: Chebysheff.
• The 68% is a rule to remember, not something you calculate.
• This is the only test where Chebysheff appears, so if it’s tested, it’s now.

▶ Empirical rule · Chebysheff’s theorem

• Joint: two outcomes in the same observation, e.g. P(male and black hair).
• Marginal: the probability of one kind of outcome in general. It can often be found by adding up joint probabilities, e.g. P(male) = P(male and black hair) + P(male and blonde hair) + … and so on for every hair colour.
• Conditional: a probability that applies only when a certain outcome is already known to have happened.

Example: given that a person has black hair, what is the probability that they are male? We know they have black hair; it’s not left to chance. What is left to chance is whether they are male, so it’s a conditional probability: P(male | black hair).

Tip: write the conditional correctly from the wording, and the rest of the question basically solves itself.

▶ Conditional probability (free, past test) · Joint probability (free, past test) · Focus on conditional probabilities (Term Test 1 Prep) · Joint and marginal tables (Term Test 1 Prep)

• Yes, and it’s important.
• Spot it: you’re given P(A | B) and asked for P(B | A).

Very important: you don’t need the Bayes formula. You can solve it with a probability tree or table:

1. Draw the tree, with the “given” event on the first branches.
2. Multiply along each path to get the joint probabilities.
3. Put them in a joint probability table.
4. Divide: P(B | A) = P(A and B) / P(A).

Good news: you WILL get full marks using this method instead of Bayes’ formula, because this method really IS Bayes’ formula, just broken up into easy pieces.

▶ Bayes’ theorem (Term Test 1 Prep) · More Bayes’ theorem (Term Test 1 Prep)

• Check if P(A | B) = P(A). If yes, independent. (Or check P(A and B) = P(A) x P(B).)
• One check is enough.
• Off by a tiny bit? It may be rounding. Test questions are usually set up so the answer is clear.

▶ Test for independent events (free, past test)

• No. Mutually exclusive means if one happens, the other can’t: P(A and B) = 0.
• If P(A and B) is more than 0, they are not mutually exclusive.
• Mutually exclusive doesn’t mean “covers every outcome” either. A 70% chance of rain doesn’t make snow 30%.

▶ Mutually exclusive probabilities (free, past test)

• It’s the hardest chapter, and it’s always on the test.
• Do questions from scratch, without looking at the answer.
• Many people think they understand probability until they try a question on their own.

▶ Chapter 6 lessons (Term Test 1 Prep)