Statistic Ⅰ. The defendant is guilty or no

Statistics
    Good morning, ladies and gentlemen. Today, we are going to have a lecture on statistics. Firstly, let’s see the effect of statistics in the courtroom.
    After the November 2000 election, Americans waited while court challenges debated who won Florida’s electoral votes. In Palm Beach County, third-party candidate Pat Buchanan got a higher percentage of votes than he did elsewhere. Was that because the country’s "butterfly ballot" caused many Albert Gore supporters to punch the wrong hole? Lawyers also argued about other claimed irregularities in the balloting.
    The Supreme Court finally stopped all vote recounts in early December. But if Gore’s statistical arguments had convinced the judges, he would have become president instead of George W. Bush. More than ever, plaintiffs must often prove their case with numbers. Let’s see’ how statistics is taking center stage in some other courtroom cases.
    Let’s see the first case whether the defendant is guilty or not.
    Imagine you’re on the jury in a murder case. An expert testifies about DNA evidence. She says that a sample from the crime scene matches a defendant’s. She also gives the odds that someone else would randomly match the tested fragments. If the odds are one in a million, the makes it sound very likely that the defendant is, in fact, guilty. The defense lawyer may try to counter that by saying that in a city of three million people, at least two others would also probably match. Of course, the defendant was not arrested at random. Almost always, police have some other evidence linking a person to a crime. But the statistics supporting DNA evidence may be just the proof needed to find someone guilty beyond a reasonable doubt.
    Now, I would like you to look at the cases against unjust discrimination.
    In a U.S. Supreme Court case earlier this year, lawyers argued over whether a state university’s admission plan unlawfully added points for students from certain minority groups. Statistically, that made it easier for those students to get in. Statistics factors into other discrimination cases, too. Suppose Annie claims that unlawful discrimination kept her from getting a promotion. Her lawyers may use statistics to show that workers with the same qualifications were significantly more likely to get promotions if they were male or not in a minority group. If the employer can’t show that Annie didn’t do her job well, she could then win her case.
    Age, family history, exercise habits, diet, weight, and other factors affect the likelihood of developing cancer, heart disease, and other illnesses. For years, tobacco companies said that smoking was not the cause of plaintiffs getting sick. They won most cases against them. Then, judges and juries listened to statistical evidence that even when other factors were equal, smokers had much higher disease risks. Finally, some cases started to hold tobacco companies liable, or legally responsible.
    Of course, the plaintiffs had other evidence, too. Tobacco companies own documents showed that they knew about disease risks. Yet their ads still targeted young people. In other words, the statistics did not stand alone. When using statistics, it is important to understand that statistical evidence complements other forms of evidence. Statistical evidence should be evaluated along with other evidence and not alone.
    And then how can the court work out the compensation the victims should gain? Statistics help add up damages actually. Statistics help decide how much people or companies must pay if they are liable. Suppose a defect in a car caused an accident. As a result, Bert could no longer work. Statistics could show how long Bert would other wise have worked and how much he probably would have made. When two variables correlate with each other statisticians can often predict one value from another with regression analysis. If someone plotted all the data points on a scatter plot, the analysis would find the line with the best fit through them.
    But suppose that people in case claimed that nearby pollution lowered property values for a whole neighborhood. It may be unfair just to match sale prices with distance form the pollution, or to compare average prices with another town. After all, many factors affect property values: style of house, size, age, number of bathrooms, and so forth. That’s where multiple regression analysis can help. It finds the "best fit" for all the sample data when multiple independent variables are at work. It nets out the effects of all these things that are different, so that you are comparing apples to apples.
    You will also know that statistics have the power on the witness stand. Good statistical experts make numbers "user--friendly" for the judge and jury. Many use high-tech graphics and other tools to present their conclusions. But statistics can be and have been misused, typically when people have interpreted that statistics to mean more than they really do. A good statistician is careful to explain just how reliable the statistics really are. Cross-examination lets each side attack the other side’s analysis is flawed. Were data accurate, or may they have been biased? What was the margin of error? Did onevunfairly affect the outcome?
    Finally, the jury weighs statistical evidence along with all the other evidence. The verdict makes a real difference in the lives of parties to a case and to our justice system.
    Today, we’ve talked about the magic of statistics working in various fields and events, like in the court, in the issues against discrimination, in calculating damages and on the testimony. I hope you would have had a good time.