Questions for 800 072
Answers to these questions are available here.
- What is the purpose of a graphical presentation of data?
- Qualitative data is by definition data that is in discrete categories or
classes. Discuss circumstances in which you could group such data into fewer
categories or classes, and why it would be appropriate.
- Discuss the advantages and disadvantages of mean, median, and midrange as
measures of central location.
- Discuss the advantages and disadvantages of percentiles and z-scores as
measures of relative position.
- Give a data set for which P(|x-mu| [greater than or equal to] 2 sigma)
= .25 (i.e., the z-score of 1/4 of the data points is greater than or equal to
2.) (you will need to use sigma rather than s for the standard deviation, i.e.,
divide by n rather than n-1 when calculating the standard deviation).
- Describe an experiment (in the sense of the word "experiment" which we us
in statistics) which you do on a regular basis.
- Give a pair of events which should be independent, and discuss why they
should be independent (you may need to identify the population in which the
- Discuss the extent to which the Iowa poll (or any poll if you are not
familiar with the Iowa Poll) is a binomial experiment.
- Why does the Normal table in the text only go from z values -3.9 to 3.9?
What would you use for the prbability (area) when you encounter z values
outside that range?
- Why would/should one use the normal approximation to the binimial
distribution rather than the exact binomial distribution?
- If an observed x-bar is significant at the 5% level, what can you say
about the significance of that x-bar at the 10% level? 1% level?
- Discuss the relationship between tests of hypothesis and confidence
intervals (cf. problems 9.52 and 9.53, pp. 523, 524).
- Discuss the use of z-scores versus P-values to indicate the likelihood of
an outcome if the null hypothesis is true.
- Discuss the central limit theorem as it applies to binomial experiments.
- Discuss the relationship between the tails of the normal and Chi-squared
distributions, in relation to hypothesis testing.
- What is the relation between the correlation and the slope of the least
squares regression line?
- Discuss the notion of causality in relation to correleation.
- Discuss some variables which you would expect to be negatively correlated.