1st week: Relationship between two variables and regression analysis.

Tuesday  September 2,2003                                        2:30 - 3:55 PM 208 Kupfrian Hall

Methods for obtaining and analyzing data will be discussed in this lecture. First a film clip will present the case of Sera(GHarmon), who’s growth rate follows a linear relationship. However, her growth curve is outside of the normal range for children, and a growth hormone deficiency is the diagnosis. The use of a graph to display data is important.  Linear relationships are the only relationships which are convenient to analyze. These relationships can be characterized by a slope and a y intercept (bond). The concept of using a computer to obtain a scatter plot (STmanatee, Plot) or a linear plot will be discussed. These concepts will be extended to the use of a linear regression (diet), but there are possible pit falls (PlotProb). Exponential growth vs linear growth will be demonstrated in the film clip ChessBd. The concept of changing an exponential plot to a linear plot will be introduced. Then, following a discussion of the propagation of error, the relationship between vapor pressure above a liquid and the temperature of that liquid will be used as an example. A plot of ln(Pv) vs 1/ T using excel will be demonstrated. This graph should be reproduce by all students as a Homework for a maximum of 5 points credit. Finally the statistical error analysis for the ln(Pv) will be discussed in detail.

The following CD ROM movies will be viewed in class: Napgraph(0:38), GHarmon (5:59), Plot(5:04), bond(4:00),  PlotProb(1:53), LeastSq(1:31), Diet(11:56), Chessbd(6:09). (Total time ~38  minutes).

1st week: Written Report and Introduction to Statistics.

Friday September, 5, 2003   11:30 AM - 12:55 PM 208 Kupfrian Hall

The written report preparation, computer graphing of data, analysis of scientific concept behind the results, and an introduction to statistics will be discussed in this lecture. The importance of statistics for Domino’s Pizza (STPizza) will be used to demonstrate the use of statistics to evaluate the potential of a new product. The statistical error, and propagation of error parts of a physical chemistry lab report will be discussed. The use of a histogram (histogram,Livhist,lighning) and the statistical averaging of data (Runmean) to see a trend wll be presented. A lab report with an error analysis, from Physical Chemistry Lab (or senior lab), must be turned in for maximum of 20 points towards the course grade. This grade may be given after the lecture part of the course is over.

CD ROM movies:Vol. 3:Pizza (7:22), Histogram(1:43), lighning (8:36), Height( 0:17 ), Rmean(1:45). (Total time ~20)

Discussion of Report Writing

2nd week: Error in Measurements

Tuesday  Septyember 9, 2003        2:30 - 3:55 PM 208 Kupfrian Hall

The error in measuring the volume of a liquid, the weight of a sample, the temperature of a sample, etc. will be investigated. The accuracy of a volumetric flask will be compared to that of a graduated cylinder. The basic nature of data, how to design the experiment, the relationship between accuracy of the measurements and the accuracy of the results will be discussed. (Homework assignment is to computer the mean and standard deviation of rows and column in the excel file homework on solder tips 5 points).

CD ROM movies: Vol. 3: Weight(2:45), AccPres(4:36) (Total time ~7 1/2 minutes).

Discussion of Measurements

2nd week: Distribution and Standard deviation.

Friday September 12, 2003   11:00 AM - 12:55 PM 208 Kupfrian Hall

First probability will be discussed on the movie coinflip. In the movie, mean, the meadium and the mean of a distribution of salaries will be discussed along with ways of quickly displaying a distribution ( Boxplot). Binomial distributions will be introduced (Normdist, Nomplot, Normrule, baseball). The calculation of the mean and the median, the standard deviation, and representative examples of binomial distribution will all be presented(Samping, Rheight). (Homework assignment is to obtain a histogram of solder tips as a function of thickness 5points)

CD ROM movies: Coinflip(3:51), mean(4:17), Boxplot(5:14), Normdist(5:55), NormPlot(2:05), Normrule( 1:52 ), baseball(6:18), Rheight(3:19),Samping(4:48). Total time (~37 minutes).

Statistical Treatment of Data

Representing Distributions

Normal Distributions

3rd week: Literature and the Library.

Tuesday  September 15, 2003        2:30 - 3:55 AM 208 Kupfrian Hall

The use of the library and Internet to obtain information will be discussed by Bruce Slutsky of the library Staff.

Literature search assignment 10 points.

Discuss the use of the library.

3rd: week: Correlation of data

Friday September 23, 2003   11:30 AM - 12:55 PM 208 Kupfrian Hall

How to derive and interpret the correlation coefficient will be discussed in STCorrel. The correlation between the physical and personal characteristics of twins will be presented. The mathematical formula will be given, along with examples with varying correlation coefficients in rvalue. Then the danger of lurking variables will be discussed in LurkVar. Finally, correlation will be compared to linear regression.

CD ROM movies: Correl(9:51), cake(3:17), Lurkvar(8:14). (Total time ~21 1/2 minutes).

Discussion of Correlation

4th: week: Confidence intervals and Significance testing.

Tuesday  September 23, 2003        2:30 - 3:55 PM 208 Kupfrian Hall

The relationship between the number of standard deviations and relative number of samples in a normal distribution (confide Intervals) will be shown in the movies STBatter and examples given. For example in the movie, Bloodpr, the author sample her blood pressure for seven days and statistically analyzes the results. In the movie.  Then in the movie sample the use of statistics to maintain the quality control at Frido Lay’s of beginning materials through sampling.

CD ROM movies:    Batter(12:30), Blood(6:33), comut(1:53),chips(4:35)    (Total time ~26 minutes).

Discussion of Confidence Interval

4th: week: Control Charts.

Friday September 26, 2003  11:30 AM - 12:55 PM 208 Kupfrian Hall

The game of roulette will be used to demonstrate the central limit theorem and the reduction in the standard deviation with repeated sampling in roulett. The use of sampling and control charts to provide quality control will be discussed in the movie STControl. Then in Tstat the concept of t-tests, the statistical analysis of confidence if the standard deviation is not available, is discussed. Examples are given with PCB in Ttest and with foams in DoubTtst.

CD ROM movies:  Roulet( 8:38),  Ttest(3:35), Control(6:41),Tstat(9:29), DoubTtst(5:50)  (Total time ~32  minutes).

Use roulette to understand law of averages

Discuss how to control a process

5th week: Oral Reports

Tuesday  September 26, 2003        2:30 - 3:55 AM 208 Kupfrian Hall

Methods for oral presentations using powerpoint will be given. The use of excel and powerpoint together will be discussed. Examples of good and poor presentation of results will be given. (Homework assignment is to set up a power point lecture with at least 10 slides and one graph on the solder tips (5 points))

Discussion on Power Point Presentations

There are additional student sample reports on a CD-rom. (pptsam1-4) (78 slides)

Note: There are errors in these presentations, do not take all of these examples as goals to strive for.

Against All Odds: Inside Statistics

This primer on statistical method and inference introduces the practical applications of statistics. Most of the videos used in this course come from this series. Produced by the Consortium for Mathematics and Its Applications and Chedd-Angier. Released: 1989.

Most of the film clips used in this course are taken from this series. It is recommended that these VHS tape be checked out of the library and viewed.

1: What is Statistics? Discover how this complex discipline has evolved.

2: Picking Distributions Construct stemplots and histograms, and understand the importance of pattern deviation.

3: Describing Distributions Examine the difference between mean and median and learn of quartiles, box-plots, interquartile range, and standard deviation.

4: Normal Distributions Shows the progression from histogram to a single normal curve for standard measurement.

5: Normal Calculations Emission standards and cholesterol studies give examples of normal calculations at work.

6: Time Series Statistics identify patterns over time, answering questions about stability and change, as seen in the stock market.

7: Models for Growth Topics include linear growth, least squares, exponential growth, and straightening an exponential growth curve by logic.

8: Describing Relationships Scatterplots, smoothing scatterplots of response versus explanatory variables by median trace, and least squares regression lines are covered.

9: Correlation How to derive and interpret the correlation coefficient using the relationship between a baseball player’s salary and his home run statistics.

10: Multidimensional Data Analysis This program recaps the data analysis by showing computing technology at Bell Communications Research.

11: The Question of Causation The relationship between smoking and lung cancer is examined, and a study of admissions data illustrates Simpson’s paradox.

12: Experimental Design Distinguish between observational studies and experiments, and learn the basic principles of design including comparison, randomization, and replication.

13: Blocking and Sampling Understand random sampling and the difference between single-factor and multifactor experiments.

14: Samples and Surveys Stratified random sampling is explained. A 1936 Gallop election poll yields information about undercoverage.

15: What is Probability? Distinguishes between deterministic phenomena and random phenomena.

16: Random Variables Topics covered include independence, the multiplication rule for independent events, and discrete and continuous random variables.

17: Binomial Distributions Calculate the mean and standard deviation of binomial distributions, and see a representative example of binomial distribution.

18: The Sample Mean and Control Charts Roulette and business demonstrate the use of the central limit theorem, control chart monitoring of random variation, creation of x-bar charts, and definitions of control limits.

19: Confidence Intervals Explains the confidence interval using population surveys to show how margin of error and confidence levels are interpret

20: Significance Tests A hiring discrimination case illustrates the basic reasoning behind tests of significance.

21: Inference for One Mean Study inference about the mean of a single distribution, with emphasis on paired samples and the t confidence interval and test.

22: Comparing Two Means Learn to recognize a two sample problem and to distinguish it from one- and paired-sample situations.

23: Inference for Proportions See confidence intervals and tests for comparing proportion applied in government estimates on unemployment.

24: Inference for a Two-Way Tables The chi-square test and the relationship between two categorical variables are covered.

25: Inference for Relationships Understand inference for simple linear regression, emphasizing slope, and prediction.