SOLUTIONS TO MIDTERM 1 VERSION 1 1) We have x 1 n xi , which becomes (6+4+2+6+7)/5=25/5=5. Answer is B. n i 1 2) The sorted observations are 2,4,6,6,7. The median is the middle value, 6. Answer is D. 3) The range is the difference between the largest and smallest observations, 72=5. Answer is B. 4) The smallest observation is 2. The middle observation (median) is 4. The third observation x must satisfy (2+4+x)/3=6, so x=186=12. The full data set is {2, 4, 12}. This has a sample mean of 6 and a sample median of 4, as required. The sample standard deviation is 1 n ( xi x ) 2 [( 2 6) 2 (4 6) 2 (12 6) 2 ] / 2 28 5.29. n 1 i 1 Answer is D. s 5) The sample mean is multiplied by 2, yielding (2)(0)=0, so it does not change. The standard deviation doubles. Since not all of the data values are the same, the original standard deviation is positive, and if you double a positive number it changes. So the standard deviation does change. Answer is B. 6) Suppose for example that all observations are equal to 5, except the first observation, which is 1. Since both quartiles will be 5, the IQR will be zero, but since all observations are not equal the sample standard deviation will not be zero. Now, consider a different example where all observations are 5. Here, both the IQR and sample standard deviation will be zero. So if the IQR is zero then the sample standard deviation may or may not be zero. Answer is B. 7) We have z = (x− x )/s = (300000−200000)/50000=2. Answer is C. 8) The sample median (average of the two middle values) will not change unless we are changing one of the two middle values. The interquartile range would not change if we move a point that is below the 25’th percentile further to the left, or if we move a point that is above the 75’th percentile further to the right. The range will not change as long as the point we are changing is not the largest or smallest data point, and we do not change our point enough so that it becomes the largest or smallest data point. But the sample mean must change since it is an average, which will change if a single data point is moved. Answer is A. 9) Let A=Recession, B=Drop in Consumer Confidence. We have P(A)=0.3, P(A∩B)=0.2. We need P(B|A)=P(B∩A)/P(A)=(0.2)/(0.3)=2/3. Answer is B. 10) Using the definition of conditional probability, since P(B∩A)=P(A∩B), the given fact, P(B|A)=P(A|B) reduces to P(A∩B)/P(A)=P(A∩B)/P(B). There are two ways for this to occur. Either both sides are zero (which happens if and only if A, B are mutually exclusive), or else we can divide on both sides by P(A∩B) to find that 1/P(A)=1/P(B), that is, P(A)=P(B). Answer is C. 11) Remember the multiplication rule for independent events, P(A∩B)=P(A)P(B). Remember also the general addition rule for any pair of events, P(AB)=P(A)+P(B)P(AB). But since here A, B are independent, this last equation becomes P(AB)=P(A)+P(B)P(A)P(B). Adding P(A)P(B) to both sides of this we get P(A)+P(B)= P(AB)+P(A)P(B). Answer is D. Since A, B are independent and therefore not mutually exclusive, we cannot say that P(A)+P(B)=P(AB). Answer is D. 12) If the range is zero, all values must be the same. But they need not all be zero, so we cannot conclude that the sample mean is zero. Since all values are the same, the interquartile range will also be zero, as will the sample standard deviation. Answer is B. 13) A=Rain, B=Lightening. We need P( A B ) P( A B) 1 P( A B) . Note that P(AB)=P(A)+P(B)P(AB)=P(A)+P(B) P(A)P(B|A), which from the given information becomes .3+.2(.3)(.4)=.5.12=.38. So the probability of neither rain nor lightening is 1.38=.62. Answer is D. 14) Using the rule of complements, we have 1P(A) > 1P(B), which is the same as P(B) > P(A). Answer is B. 15) Note that P(B)=1/2, while P(B|A)=1, since if we know that the number was a four, we know that it was even. Since P(B) is different from P(B|A), the events are not independent. Since A and B can both happen (note that AB=A), the events are not mutually excusive. Clearly, the events are not complements of each other. Answer is D.