This is the course webpage for Statistical Methods 1 for
B.Stat(hons) first year students (2023). I shall post various
things (code snippets, data sets etc) here from time to time.
Please see the projects page.
Google classroom joining link
To learn more about them, lookup their helps by typing a
question mark(?) followed by the name (enclose the name in quotes
if it is some symbol like ":" or "*"). In particular, there is a
wealth of information that is worth knowing
for plot, barplot and lines.
Oct 05, 2023:
We have covered the hard part. Now we shall cruise through the
remainder of the course. I have created a
Youtube playlist
containing my lecture videos (the order may differ slightly from the order
I follow in class). Also we shall now adhere to the textbook by
Witte and Witte more closely. Read chapter 1 and 2. Skim through
chapter 8, and sections 9.1 and 9.2. These roughly cover the
topics taught so far. The rest of the course will primarily deal
with chapter 3, 4 and 5.
Nov 02, 2023: An astrology fallacy regarding pooled
correlation [summary]
Nov 06, 2023: Simpson's paradox. Concept of confounding,
blocking and randomisation: [R code
| data]
Nov 09, 2023:
Please skim through Part III (Correlation and
Regression) from Freedman, Pisani and Purves. It is about 100
pages, most of which you will find trivially easy. Pay special
attention to Section 5 of chapter 9 (Association is not
causation) and Section 4 or chapter 10 (The regression fallacy).
Also skim through chapter 1 of the same book to develop
intuition about controlled experiments.
The lecture notes are the main reference materials. You should also
take a look at the following books. The lectures
will not follow the books.
Statistics by Witte and Witte
The book that is closest to being our main textbook. It is an
easy-going book, not too ambitious.
Statistics by Freedman, Pisani and Purves
This book is a thought-provoking one. It has ideas and open
questions. Not a regular textbook, but a welcome change from the
world of classroom lectures.
How to Lie with Statistics by Huff
A fun book, as its name suggests. Contains cartoons, and some
valuable insights. A must read, when you get bored!
We shall
use the R environment (language plus libraries) for all our computation. If you know Python,
that`s great, but please also learn R. We shall need only only a small
subset of features of R, that we shall discuss and demonstrate
in class. The following books are for optional self-reading.