Download data. In science, society and everyday life, people use data... world and choose how to act, and statistical methods help... UNIVERSITY OF TORONTO AT SCARBOROUGH STAB22H3

UNIVERSITY OF TORONTO AT SCARBOROUGH
STAB22H3
Fall 2012
STATISTICS I
Course Description: Statistics is the science of collecting, organizing and interpreting
data. In science, society and everyday life, people use data to help them understand the
world and choose how to act, and statistical methods help to separate sense from
nonsense.
In this course, we learn about some of the most important techniques used in statistical
work. The emphasis of this course is on concepts and techniques and will be useful to
students who seek to gain an understanding of the use of statistics in their own field. Our
ultimate goal is to gain understanding from data, going from data collection to analysis to
conclusions.
Content, emphasis, etc. of the course is defined by means of the lecture material - not
only the texts. It is important to attend all lectures, as there is normally no simple way to
make up for missed lectures (perhaps obtain another student’s notes). There will also be
many lecture examples using statistical software, which students will be using.
Important announcements, problem sets, additional examples, and other course info
will be posted on the course web homepage. Check it regularly.
Course Schedule:
Instructors:
LEC01: Sotirios Damouras
E-mail: [email protected]
Office IC344
Office hours: Thu 11-1, Fri 3-5
Lectures: Tue 12:00-13:00 SW319, Fri 12:00-13:00 SW319
Webpage: http://www.utsc.utoronto.ca/~sdamouras/stab22h3.html
LEC02: Mahinda Samarakoon
E-mail: [email protected]
Office: IC442
Office hours: Tue 11-12, Fri 11-12
Lectures: Tue 10:00-11:00 IC130, Fri 10:00-11:00 IC130
Webpage: http://fisher.utstat.utoronto.ca/~mahinda/stab22/stab22.html
Textbook: Stats: Data and Models, DeVeaux, Velleman, Bock, Vukov, Wong
Canadian edition, publ. Pearson Canada
1
Tutorials
Tutorials will begin in the second week of lectures (i.e. the week of Sept 17, 2012). In
preparation for tutorial, you should do the weekly assignment, posted on the web page
(suggested questions). There will be some brief assessment at each tutorial – perhaps a
short quiz based on the assignment
Evaluation
Tutorials (based on quizzes): 20%- quizzes will be held during tutorials and will last
approximately 10 minutes. They will cover material from the previous week of lectures.
Term Test: 30% -a two-hour test on a date to be announced.
Exam: 50% -a three-hour final exam.
The midterm test and the final exam are based on multiple-choice questions.
Missed Tests
There are no make-up tests or quizzes. If the test is missed for a valid reason, you must
submit appropriate documentation to the course instructor within one week of the test.
Print on it your name, student number, course number, and date. If documentation is not
received in time, your test mark will be zero. If a test is missed for a valid reason, its
weight will be shifted to the final exam.
Calculators
Hand calculators are cheap and useful. Any cheap one with a square root and one
memory button will do. Mean, standard deviation, sum, and sum of squares keys may
save you a bit of time on occasion, but we do not recommend the purchase of expensive
calculators to get keys with special statistical calculations. Tests and exams will be
designed so that those calculators give no advantage. We emphasize the use of Minitab
software for doing any tedious or complex calculations.
Computing
Students will be using, StatCrunch for computing. No previous computing experience
is assumed. With this software, you will analyze the data sets used in the text exercises.
The data sets can be found on the CD accompanying the textbook, and on the publisher’s
web site.
2
STABB22 (Fall 2012) - TENTATIVE LECTURE GUIDE
Do many of the odd problems in the text for practice (answers are in the back)
Week 1: Introduction to course, overview. Data, Variables, units (Ch 2). Displaying and describing
categorical data, frequency tables (p21), bar charts, pie charts (p23), Contingency tables (p24), Conditional
distributions, (p27)
Week 2: Displaying and summarizing quantitative data (Ch 4, p49). Histograms (p49), Stem-and-leaf
displays (p51), The shape of a distribution (p54), The centre of a distribution, means (p57) and median
(p59), skewness. Spread of a distribution (p61), Range, interquartile range (p61), and standard deviation
(p63). The five number summary (p66) . Understanding and comparing distributions (Ch 5), boxplots,
(p89), 1.5IQR rule for outliers(p90) comparing groups with histograms (p91), comparing groups with
bosplots (p92)
Week 3: The standard deviation as a ruler and the Normal distribution ((ch 6, p121). Standardized values ,
z-scores (p123), Shifting data (p124), rescaling data (p126), Linear and non-linear transformations of data
(p127). Density curves and the normal model (p129),The 68-95-99.7 rule for Normal models (p133), Finding
normal percentiles (p135). Normal probability (quantile) plot. Form percentiles to scores (p139).
Week 4: Scatterplots, association and correlation 9Ch 7, p168), describing scatterplots (p170), role of
variables (p172), Correlation (p173), Correlation conditions (p176), correlation properties (p178), Linear
regression(Ch 8, p198), The least-squares line (p201), predicted values and residuals (p202), residual plots
(p206), regression assumptions and conditions (p213), R-square, the variation accounted for (p209), , ,
what can go wrong (p216)
Week 5: Regression wisdom (ch p, p231), Residuals, nonlinear relations (p231), subsets of data (p234),
outliers and influence (p234), lurking variables and causation (p238), Extrapolation (p239), , working with
summary values (p242), restricted range (242)
Re-expressing data (Ch 10, p263), goals of re-expression (p266), the ladder of powers (p269).
Week 6: Understanding randomness (ch 11, p300) sample surveys (ch 12, p314), population. Sample,
bias (p215), randomization (p316), sample size (p317), census (p318), Populations and parameters,
samples and statistics (p318), simple random samples, (p319), stratified samples (p321), cluster and
multistage sampling (p322), systematic samples (p325), what can go wrong p330.
Week 7: Experiments and observational studies (ch 13, p341), Observational studies (p342), Experiments
(p343), principals of experimental design (p345), does the difference make a difference (p349), Experiments
and samples (p350), Control groups, blinding (p351), placebos (p352), blocking (p353), more factors
(p355),c onfounding (p356).
Probability (ch 14, p376), The law of large numbers, empirical probability. (p378), Theoretical probability
(p380), personal (subjective) probability (p381).
Week 8: Probability rules (p382) , Conditional probability ((p395), independence (p398), General
multiplication rule (p399), multiplication rule for independent events (p400), independence and disjointness
(p404)
Week 9: Random variables (ch 16, p422), discrete random variables, probability distributions (models) ,
expectation (means) of a random variable (p423) , standard deviation of a random variable (p425), linear
transformations (p427), Two or more random variables (p428) , continuous random variables (p433),
combining random variables (p433)
Probability models (p445), Geometric model, (p447), Binomial model (p449), Binomial tables, Normal
approximation to the Binomial distribution.
Week 10: Sampling distribution models (ch 18, p473). Sampling distribution for sample proportions, CLT for
sample proportions (p474), Sampling distribution for sample mean (p482), CLT (p484), Confidence
intervals for proportions (ch 19, p504), margin of error (p508), sample size (p514),
Testing hypotheses about proportions (p530), null, alternative hypotheses (p531), p-values (p533), onesided and two-sided tests (p538),
3
Week 11: More bout tests (ch 21, p554), alpha levels (p561), confidence intervals and hypothesis tests
(p565), type I, type II errors (p 567), power of a test (p569) Power and sample size (p573),
Comparing two proportions (ch 22, p585) The standard deviation of the difference between two proportions
(p587), Confidence interval for the difference between two independent proportions (p589), .
Week 12: Inference about means (ch 23, p617), One sample t-interval for the mean (p621), Tests for the
mean (p628), Comparing two means (Ch 24, p654), Two sample t-interval for the difference between two
means (p659), a test for the difference between two means (p663). Paired sample (Ch 25, p688), the paired
t-test p692.
4