What This Book Actually Does
Mendenhall's Introduction to Probability And Statistics Mendenhall is a standard undergraduate textbook that covers the core material you need before moving into more advanced theory. It works well for two-semester introductory sequences or one intensive semester if you already have some calculus background. The book is heavy on examples and has detailed walk-throughs of how to get from raw data to a finished statistical conclusion. I ran into a real problem with this text when teaching my first probability course. The section on conditional probability and independence has a handful of problems where the answers in the back of the book are wrong. I caught it because my students were reporting consistently off results when they worked through example three in chapter four. The workaround was straightforward: I stopped having them check answers against the back and instead had them verify by running simulations in R. That took about twenty minutes extra per problem set but saved me hours of explaining why their correct work didn't match the key. The errors are in the later editions; the 13th edition cleaned up most of these issues. The book opens with descriptive statistics and works its way into probability, random variables, sampling distributions, and then inference. That's the standard sequence and it makes sense for students who haven't seen any of this before. The probability chapters are the strongest part. Mendenhall explains sample spaces, event operations, and the three axioms without being overly formal about it. Most other books either go too abstract too quickly or never quite get there. This one lands somewhere usable.
Where the book gets thin is in the coverage of nonparametric methods. The Mann-Whitney U test gets two pages. If your course needs any serious treatment of distribution-free procedures, you'll need a supplementary source. The bootstrap section exists but it's more of an acknowledgment than a proper explanation. That's a structural weakness in how the whole thing is organized, not something you can fix by reading harder. The simulation approach Mendenhall uses for sampling distributions is actually one of the more useful features in the book. Instead of just presenting the central limit theorem as a fact to memorize, the text shows you how to approximate a sampling distribution by generating data and computing statistics repeatedly. I found this especially useful when explaining why the t-distribution has heavier tails than the normal. Running a quick simulation where students generate their own samples and watch the distribution of means converge tends to stick better than any proof will. It takes about ten minutes in class instead of a full lecture that most students tune out of anyway. One thing beginners miss is the treatment of continuous random variables. The textbook covers the uniform, normal, exponential, and gamma distributions, but it doesn't spend enough time on transformations of random variables. You'll find yourself needing to look elsewhere for techniques like the change-of-variable method or the moment-generating function approach to deriving distributions. These come up in later courses and the gap here shows. I usually assign supplementary notes from online lecture materials to cover the transformation material that Mendenhall skips over too quickly.
The regression chapters are solid for what they cover. Simple linear regression gets proper attention to least squares estimation, residual analysis, and the assumptions behind the model. Multiple regression follows and the matrix algebra is introduced at a level that won't intimidate someone who has only had one semester of linear algebra. That said, the coverage of diagnostic plots is surface level. You'd want to pair this with a resource that goes deeper into influence measures and leverage diagnostics if your course requires it. If you are looking to get a copy, the textbook is widely available through major retailers and academic sellers. The digital version tends to be cheaper and the searchability helps when you are working through problem sets. Physical copies hold up better if you plan to annotate them heavily, which most students doing the problem sets end up doing anyway. The companion website has datasets and occasionally updated solution supplements that aren't always aligned with the print edition. The main bottleneck with this book is the pacing. The early chapters move slowly and deliberately, which helps beginners but can feel tedious if you already understand basic statistics from another source. Students who have taken AP statistics often breeze through the first third before the book really gets going. Skipping ahead isn't always a bad move here since the later material builds on concepts that are introduced early but revisited in more depth. I've had students use the book primarily as a reference for the probability sections and spend most of their time on the inference chapters where the actual learning happens.
Get the Full Details
For anyone working through the exercises, start with the odd-numbered problems first. The answers are in the back and the even-numbered ones are left for assignment. The odd-numbered set gives you a clean feedback loop. If you're stuck on a particular concept, the examples at the start of each section are usually sufficient before you move to the practice problems. The worked examples are where the text does its best work.
