ALL ODDS VIDEOS
All Odds” is a set of videotapes that cover a variety of introductory topics
in statistics. The College of
Business purchased these tapes and they are available for student viewing in the
Media Center of the Library. “Against
All Odds” is an excellent review of material in both MGT 216 and MGT 217.
The following topics are covered in the series.
An overview of the nature and impact of statistics using historical
anecdotes and short views of contemporary applications.
2: (lectures 2 and 3--216)
Presenting and interpreting the distribution of a single variable. Techniques
taught include stem plots, frequency tables and histograms.
(lectures 1 and 2--216)
Numerical measures of specific
aspects of a distribution: center (mean, median, mode), spread (percentiles,
five number summary, boxplots, and standard deviation). Resistance and its
Density curves as smoothed histograms: mean, median, percentiles for density
curves: the normal distributions (general shape, locating the mean and standard
deviation the 68-95-99.7 rule).
Standardization and calculation of normal relative frequencies from tables;
assessing normality by normal quartile plots.
(lectures 11, 12, 13--217)
Series From the
distribution of a single variable we move to an examination of change over time.
Topics: Statistical control, inspecting time series for trend, seasonal
variation, cycles; smoothing by averaging either over many units per time or
over time by running medians.
for Growth Mathematical
Models for the overall pattern of simple lands of growth over time. Topics:
Linear growth with review of the geometry of straight lines and an introduction
to the least squares idea; exponential growth, and straightening an exponential
growth curve by logarithms: prediction and extrapolation.
(lectures 5 and 6--217)
scatterplots and their variations; smoothing scatterplots of response vs.
explanatory variable by median trace; linear relationships, least squares
regression lines and comment on outliers and influential observations.
Correlation and it properties: the relation between
correlation and regression.
Data Analysis The
impact of computing technology on statistics especially graphics for displaying
multidimensional data. A case study in data analysis will employ techniques
discussed in previous program.
Question of Causation
Association between categorical variables displayed in a two-way table;
Simpson’s paradox: the varied relations among variables that can underlie an
observed association; how evidence For causation is obtained.
Design Advantages of planned data collection over anecdotal evidence
or available data. The idea of an experiment. Basic principles of
design: comparison, randomization, replication.
Experiments and samples further principles of design: two or more factors and
blocking. Introduction to sample surveys, the danger of bias, random sampling
Samples and Surveys: Sampling and Sampling
Distributions More elaborate sample
designs: stratified and multistage designs. The practical difficulties of
sampling human populations. The idea of a sampling distribution.
Probability as a model for long term relative frequencies or personal assessment
of chance. Sample space, basic rules of assigning probability: addition rule for
independence and the multiplication rule for independent events. Discrete and
continuous random variables. Mean and variance of a random variable.
law of large numbers. Addition rules for means and variances of random
variables. The binomial distributions for sample counts. Norman approximation to
Sample Mean and Control Charts
The sampling distribution of x-bar, the central limit theorem, x-bar
control charts and statistical process control.
The reasoning behind confidence intervals. z-intervals for the mean of a
normal distribution. Behavior of confidence intervals.
Tests The reasoning
behind significance tests illustrated by the sample case of tests on a normal
mean with known standard deviation. Null and alternative hypotheses and p-values
and cautions on the limited information provided by tests.
(lectures 13 and 16--216)
for One Mean
Inference about the mean of a single distribution with emphasis on paired
samples as the most important practical use of these procedures. The t
confidence interval and test.
(lecture 14 and 15-216)
The two-sample t confidence intervals and tests for comparing means;
brief mention of the sensitivity of the corresponding procedures for variances
to non-normality and their consequent impracticality.
12 and 16--216)
Confidence intervals and tests for a single proportion and for comparing
proportions based on paired and independent samples.
for Two-Way Tables
Chi-square test for independence/equal distributions in two-way tables.
Inference for simple linear regression, emphasizing slope and prediction.
A case study that illustrates the major aspects of statistical thinking;
planning data collection, analysis by graphs and informal inference, more data
collection in response to partial success.