Sei sulla pagina 1di 2

Business Statistics Third homework assignment SOLUTION

Spring 2012

Turn in your homework at the beginning of the class. Your homework solution should be well organized. Include graphs that you are asked to view in the problems. You are not limited or restricted to use the statistical packages used/presented by the instructor. You may discuss the problems with your colleagues but in the end you MUST write up your homework individually.

I) The following cross-tabulation shows household educational level of the head of the household.
Educational Level Not HS graduate HS Graduate Some College Bachelors Degree Beyond Bach. Deg. TOTAL Under 25 0.057 0.067 0.038 0.012 0.004 0.178 Household 25-49.9 0.047 0.094 0.071 0.028 0.011 0.251 Income 50.0-74.9 0.019 0.068 0.063 0.039 0.017 0.206 ($1000s) 75.0-99.9 0.007 0.036 0.044 0.035 0.017 0.139

income

by

100 or more 0.005 0.036 0.055 0.073 0.057 0.226

TOTAL 0.135 0.301 0.271 0.187 0.106 1.000

a) What is the probability that the household income being in the range $50,000 and $74,900? Pr(Income = 50,000-74,900)=0.206 b) What is the probability of a head of household having at least some college education? Pr(Education at least some college)= 0.271+0.187+0.106 = 0.564 c) What is the probability of a household headed by someone with a bachelors degree and earning under $25,000? Pr(Income<25,000 AND Education=bachelors degree)=0.012 d) Knowing that a household income is between $50,000 and $99,900, what is the probability that the household head has at most high school degree? Pr(Education at most high school | Income=50,000-99,900)= Pr(Education at most high school AND Income=50,000-99,900)/ Pr(Income=50,000-99,900)= (0.019+0.007+0.068+0.036)/(0.206+0.139)=0.3768116 e) Given that the educational level of household head is Beyond a Bachelors degree, what is the probability the household income is above $50,000?
Pr(Income above 50,000 | Education beyond bachelors degree)= Pr(Income above 50,000 AND Education beyond bachelors degree)/ Pr(Education beyond bachelors degree)= (0.017+0.017+0.057)/0.106=0.8584906

II) Based on existing studies, the probability that an adult person will develop bowel cancer is roughly 2%. Let us call this the condition C and say that C=positive when the person has bowel cancer and C=negative when the person does not have bowel cancer. Therefore, Pr(C=positive)=0.02 and Pr(C=negative)=0.98. A routine test for detecting bowel cancer is the fecal occult blood (FOB) test. Let us call the test T and say that T=positive when the test outcome is positive for bowel cancer and T=negative when the test outcome is negative for bowel cancer. Again based on existing studies, it was shown that Pr(T=positive|C=positive)=0.7 (sensitivity), Pr(T=negative|C=negative)=0.9 (specificity). Sensitivity measures the proportion of actual positives which are correctly identified as such, while specificity measures the proportion of negatives which are correctly identified. From the above probabilities we can compute the joint 2x2 table Pr(T=positive,C=positive)=0.7*0.02=0.014 Pr(T=positive,C=negative)=0.1*0.98=0.098 Pr(T=negative,C=positive)=0.3*0.02=0.006 Pr(T=negative,C=negative)=0.9*0.98=0.882 Pr(T=positive)=0.014+0.098=0.112 a) Compute the probability that a person does not have bowel cancer, given that his/her test turned out positive. This is called the rate of false positives. Pr(C=negative | T=positive)=0.098/0.112=0.875 b) Compute the probability that a person does have bowel cancer, given that his/her test turned out negative. This is called the rate of false negatives. Pr(C=positive | T=negative)=0.006/0.888=0.00676

Potrebbero piacerti anche