STAT 100, Section 1: Statistical Concepts and Reasoning
Spring 2005
Midterm Exam 2, Friday, Mar. 18
The exam
will be worth 100 points (out of 500 for the whole course).
Students are responsible for all material covered in lectures
since the first midterm
and all material in Chapters 10, 11, 12, 13, 16, 17, and 18 of the textbook.
There are some
practice problems
available, along with
answers.
Please remember that on the actual exam, a calculator will not
be necessary. However, you may use one if you wish. (Some of the
sample exam questions would benefit from the use of a calculator, but
on the actual exam these questions will have answers listed in a way
such that a calculator is unnecessary.)
The following brief outline of topics from the book is not an exhaustive
list of things you need to know, but it may help you make sure you
have covered the main ideas.
Chapter 10
Descriptive methods for two measurement variables
Interpretation, properties, and features of correlation
Interpretation of statistical significance
Masking or grouping of variables; its effect on the
slope and intercept of regression lines
Interpretation of the slope and intercept
Using the regression equation for prediction
Link between sign of the slope and sign of the correlation
coefficient
Statistical relationship vs. deterministic relationship
Regression: designation of the response (outcome) variable and
the explanatory (independent or predictor) variable
Statistical significance, practical significance
Effect of sample size on statistical significance and practical
significance
"Two warnings" on pp. 182-183 of the textbook
Issue of extrapolation - when to avoid
Chapter 11
Impact of outliers on correlations
Legitimate outliers, illegitimate correlation
Masking or grouping of variables; its effect on correlation
(Simpson's Paradox)
Correlation does not cause imply causation
Limitation of "cause and effect" statements with statistically
significant correlations
Chapters 12 and 13
Assessing statistical significance with the chi-squared statistic
and
the corresponding p-value:
1. Calculation of expected numbers in a 2x2 contingency table
2. Comparing expected and observed numbers
3. Calculating the chi-squared statistic
4. Determination of statistical significance using the
chi-squared statistic
5. General properties of the chi-squared statistic
Simpson's paradox, and its link to masking or grouping of measurement
variables
Interpreting a two-dimensional bar graph to determine any
relationship between two variables
Calculation and interpretation of risk, relative risk, increased risk,
and odds
Interpretation of 2x2 tables and tables which are not 2x2
using row percentages or two-dimensional bar graphs
Inference for statistically significant
relationship between two variables
Misleading statistics about risk (page 227)
Link between relative risk and the chi-squared statistic
Similarities and differences between correlation and the chi-squared
statistic
Identification of the response and explanatory variable
Chapter 16
Interpretation of personal probability (Section 16.3)
The relative-frequency interpretation of probability (Section 16.2)
Determining the probability of an outcome under the
relative-frequency interpretation
Interpretation: Two different methods (page 299)
Idea and interpretation of long-run relative frequency
Summary of relative-frequency interpretation of probability
(Page 300 textbook)
Application of "Four Simple Probability Rules" (Pages 302 - 304)
Mutually exclusive events
Independent events
When will it happen? The number of attempts needed to achieve success
Accurate prediction of long-term outcomes
Calculation and interpretation of expected value (EV)
Expected value as a mean
Chapter 17
Difference between population and sample
The certainty effect; the attractiveness of reducing risk to zero
The pseudocertainty effect
Distortion of personal probabilities
The availability heuristic
Anchoring
The representativeness heuristic
The conjunction fallacy: Believing
Pr(A and B) is greater than Pr(A)
Forgotten base rates
The dangers of extreme optimism, conservativeness, or overconfidence
Calibrating the personal probabilities of experts
Ues of relative frequency to assess accuracy of personal probabilities
Seven suggestions for improving personal probabilities
Chapter 18
Coincidences, and how often they happen to someone, somewhere, someday
Calculation of birthday probabilities
The Birthday Problem and its relationship to coincidences that seem improbable
General definition of a coincidence
The gambler's fallacy and its link to streaks, sequences, and coincidences
The law of small numbers
Memoryless nature of independent events
Small samples cannot be expected to accurately represent a population
Confusion of the inverse, and which two probabilities are typically reversed
Definition of false positive and false negative
Definition of sensitivity, specificity
Calculation of sensitivity, specificity
Calculation of the probabilities of a false positive or a false negative
Determining the actual probability (Bayes' Rule)
Accuracy of medical tests (base rate size and specificity/sensitivity values resulting in small probability of false positives)
Using expected values for decision-making