Next: Correlation Up: CHANCE Previous: We're lucky to

Heart study


Partial credit?

Morgen Sullivan is a student at Dartmouth College and answered ``True" to the following True-False question.


True or False:  In a room of 50 people two  have the same birthday.

  1. What credit should Morgen get for this answer?

  2. What would your opinion be if the 50 had been 363? What if it had been 366? What if it had been 500?


The LA Times reported on May 1 the results of a study to compare two drugs designed to save lives of heart attack victims by clearing blocked arteries. One was an old drug called Streptokinase (costs an average of $200a dose) and the other a new drug called t-PA (costs an average of $2,200 a dose. The results of the study are described as follows: The study looked at 41,000 heart attack victims in 15 countries. They looked at deaths and strokes after 30 days with the following treatments:

       t-PA plus intravenous blood thinner
       deaths 6.3 percent 
       strokes .6 percent 

       t-PA, Streptokinase plus intravenous blood thinner
       deaths 7 percent
       strokes .6 percent

        Streptokinase plus blood thinner given by injection
        deaths 7.2 percent
        strokes .5 percent 

       Streptokinase plus intravenous blood thinner
       deaths 7.4 percent
       strokes .5 percent 

       (deaths means death rate after 30 days)

The debate over which drug is more effective has gone on for some time. Previous studies had showed little difference between the two drugs. Newspaper accounts report that researchers `have settled the long-running controversy' Do you feel that this statement is justified? If so why, if not why not?

Journal assignment

There has been a lot of discussion in recent years about the use of SAT scores to predict college performance and in particular as it might differ for men and women. Nationally, on a scale of 200 to 800, male college-bound seniors average 9 points higher on the verbal section and 43 points higher on the math section than female college-bound seniors. It has been claimed that despite the lower SAT scores, women do as well as or better than men in college work. This has led people to claim that SAT scores underpredict the achievement of women in College. There have been a number of studies to investigate this claim. You will find on gopher in the folder `SAT bias' that is in `Chance documents for general use' data desk data sets that illustrate the kind of data that has been used in this investigation. The `ETS evaluation data' is the data for one particular school in a large study done by ETS involving many schools. The Elliott-Strenta data is the data from a study carried out at Dartmouth studying psychology majors at Dartmouth. The course files give the SAT scores and grades of students at Dartmouth in three of the large introductory courses: Math3 (Calculus), English 5(writing), and Economics 1 (Introductory Economics). Using any of this data you want to, see what conclusions you can come to about the possible underprediction of SAT scores for women.

Homework assignment due Thursday 17 April

Read Chapters 8 and 9 of FPPA and do review exercises 1,4,9 at the end of Chapter 8 and 2,6,10,13 at the end of Chapter 9.

Next: Correlation Up: CHANCE Previous: We're lucky to
Tue Jun 28 15:24:59 EDT 1994