1 Kontiokari et al. (2001) contains results from an experiment on possible treatments to prevent recurrence of urinary tract infections (UTIs) in women. (Full article available at http://www.bmj.com/content/322/7302/1571 .) Women who already had a UTI were randomly assigned one of three treatments: drinking cranberry-lingonberry juice; drinking a lactobacillus GG drink; or drinking neither (the control group). The following table summarizes how many women did/did not have at least one recurrence of UTI in the next year. No At least one Tota Recurrence recurrence l Cranberry38 12 50 Lingonberry Lactobacillus GG 28 21 49 Neither (Control) 31 19 50 Total 97 52 149 Which of the following are the correct hypotheses for the chisquare test of independence in this experiment? Question 1 options: 1)Null Hypothesis: There is a relationship between the treatment (drink regularly consumed) a recurrence. Alternative Hypothesis: There is not a relationship between the treatment and UTI-recurrence. 2)Null Hypothesis: There is not a relationship between the treatment (drink regularly consume recurrence. Alternative Hypothesis: There is a relationship between the treatment and UTI-recurrence. 3)Null Hypothesis: All women with UTI are equally likely to use each of the three treatments. Alternative Hypothesis: All women with UTI are not equally likely to use each of the three treat 4)Null Hypothesis: The proportion of women that have UTI-recurrence is the same as the prop women that don't have UTI-recurrence. Alternative Hypothesis: The proportion of women that have UTI-recurrence is different than th of women that don't have UTI-recurrence. Question 2 The next series of questions asks about the expected counts, one cell at a time. Report the answer, which needs to be accurate to at least two decimal places. No UTI UTI Total Recurrence Recurrence C-L this expected Juice value? Lact. GG Neithe r Total 97 50 49 50 52 149 Answer: Question 3 Report the expected value for the indicated table cell below. The number you enter should be accurate to at least two decimal places. No UTI UTI Total Recurrence Recurrence C-L this one? 50 Juice Lact. 49 GG Neithe 50 r Total 97 52 149 Answer: Question 4 Report the expected value for the indicated table cell below. The number you enter should be accurate to at least two decimal places. No UTI UTI Total Recurrence Recurrence C-L 50 Juice Lact. this one? 49 GG Neithe 50 r Total 97 52 149 Your Answer: Question 5 Report the expected value for the indicated table cell below. The number you enter should be accurate to at least two decimal places. No UTI UTI Total Recurrence Recurrence C-L 50 Juice Lact. this one? 49 GG Neithe 50 r Total 97 52 149 Answer: Question 6 Report the expected value for the indicated table cell below. The number you enter should be accurate to at least two decimal places. No UTI UTI Total Recurrence Recurrence C-L 50 Juice Lact. 49 GG Neithe this one? 50 r Total 97 52 149 Answer: Question 7 Report the expected value for the indicated table cell below. The number you enter should be accurate to at least two decimal places. No UTI UTI Total Recurrence Recurrence C-L 50 Juice Lact. 49 GG Neithe this one? 50 r Total 97 52 149 Answer: Question 8 The chi-square test has several conditions in order for the results to be considered valid: 1) representative data from the population(s) of interest 2) what happens for one individual/unit does not affect what happens for any other individual/unit (i.e., independence of results from person-to-person) 3) Each expected count is at least 1 4) At least 80% of all of the expected counts are at least 5. Which, if any, of these conditions is definitely violated in this UTI experiment? Question 8 options: 1)Condition 1 2)Condition 2 3)Condition 3 4)Condition 4 5)None of the conditions are definitely violated; they all seem reasonable in th Answer: Question 9 Compute the value of the test statistic and report it. Your number should be accurate to at least two-decimal places, which means you will need to keep a considerable degree of precision in intermediate calculations. 2= Question 10 In order to determine the p-value we need the area to the right of our test statistic for a chi-square distribution with how many degrees of freedom? (That is, the p-value is Pr(X2 > test statistic), where X2 is a chisquare random variable with df=?) Question 10 options: a)1 b)2 c)3 d)6 e)148 answer: Question 11 The Excel function chisq.dist.rt(test statistic, df) can be used to calculate the p function to determine the p-value. p-value= Question 11 options: a)0.041 b)0.123 c)0.683 d)0.297 answer: Question 12 Using the p-value from the previous question, and supposing that we have alpha=0.05, we would reject the null hypothesis. Question 12 options: a)True b)False Question 13 What is the conclusion in context? Question 13 options: a)We have enough evidence to say that there is a relationship betwee or not there is a recurrence of urinary tract infections within a year fo b)We do not have enough evidence to say that there is a relationship whether or not there is a recurrence of urinary tract infections within c)We have proven that there is a relationship between the treatment recurrence of urinary tract infections within a year for women. d)We have proven that there is no relationship between the treatmen recurrence of urinary tract infections within a year for women. Question 14 (The remaining problems are not a continuation of the previous 13.) If we have a 4 by 4 table with one of the expected counts less than 1, wh Question 14 options: a)Proceed anyway as is; whether or not the conditions are met really b)Give up--there is no hope. c)Consider if it would make sense to combine rows and/or columns ins expected counts in the revised table are less than 1. If so, do it. d)Start off calculating the test statistic as you normally would, but the determining the associated p-value. Question 15 Under what circumstances will a chi-square test of independence give on difference in proportions? Question 15 options: a)If the contingency table has two rows and two columns, and the p1-p2 is that p1-p2 is not equal to 0. b)If the contingency table has two rows and two columns, and the p1-p2 is that p1-p2 is greater than 0. c)If the contingency table has two rows and two columns, and the p1-p2 is that p1-p2 is less than 0. d)All of the above. Question 16 Suppose we are comparing two treatments (an investigational drug to how prone they are to cause dizziness. Let population 1 be all those that conceivably would take the inves those that would take the widely used drug. If 8% of those taking the investigational drug would experience diz widely used drug would experience dizziness, then which of the fol Question 16 options: a)The risk of dizziness is 100% higher for those taking the inves the widely used drug. b)The relative risk of dizziness is 2 for those taking the investiga widely used drug. c)The odds ratio for dizziness would be greater than 1 (when co drug to those taking the widely used drug). d)All of the above