6 24 - Section 6.3 Inference for a Difference in Prop tions Question 17 of19 View Policies Current Attempt in Progress TYPE OF FACE MASK During the 2020 COVlD-19 pandemic, the United States faced a shortage of medical face masks. To save the medical masks for health care workers, people were advised to wear cloth masks in public. Here we look at the only randomized experiment1 (at the time of writing) that investigated the effectiveness of cloth masks at protecting the wearer from respiratory infection. 1 Maclntyre CR, Seale H, Dung TC, et alt (2015). A cluster randomised trial of cloth masks compared with medical masks in health care workers. BMJ Open, 5(4). Health care workers in Vietnam were randomized to wear either medical face masks or cloth face masks during their shifts for four consecutive weeks. The study was designed to see if there is evidence that medical masks are more effective at preventing illness. The researchers collected data on three different response variables: whether the health care worker developed a clinical respiratory illness (CRI), inuenzalike illness (I LI), or laboratory-conrmed virus illness (LCI). The results are given in the table below. -m-Iu LC! Mediwmasks_-l 19 Clem-nas- \"In\" 1 For all three response variables, the proportion of infection was higher for the cloth mask wearers than the medical mask wearers. Recall from Section 1.3 that there are three possible explanations for this: (i) Causal association: medical masks are more effective at preventing infection than cloth masks (ii) Association due to confounding: the groups (cloth mask wearers and medical mask wearers) differed to begin with (iii) No actual association:just random chance Cloth vs Medical Masks and Clinical Respiratory Infection For the clinical respiratory infection (CRI) response variable, (a) Calculate the observed difference in proportions with illness, medical mask - cloth mask. Round to three decimal places. (b) Use technology and the normal distribution to calculate a one-sided p-value. Round to three decimal places. (c) Based on the p-value and the standard signicance level of 00 = 0.05, do we have evidence against alternative explanation (iii)? v (d) Based on your answers to Exercise 6.193 and part (c), does this study provide convincing evidence that medical masks are more effective at protecting health care workers against developing a clinical respiratory infection than clothmasks? v eTextbook and Media Save for Later Attempts: O of 4 used Question List Question 1 Mu tipart Question 2 Mu tipart Question 3 Mu tipart Question 4 Mu tipart Question 5 Ma h Fill in the Blank Question 6 Math Fill in the Blank Question 7 Math Fill in the Blank Question 8 Math Fill in the Blank Question 9 Multipart Question 10 Multipart Question 11 Multipart Question 12 Multipart Question 13 X 0.5 / 1 D Partially correct 0.5 / 1 D Partially correct 1 / 1 ~/ Correct 1 / 1 ~/ Correct / 1 Not started -/ 1 Not started -/1 Unsubmitted work / 1 Not started / 1 Not started / 1 Not started / 1 Not started -/ 1 Not started -/1 6 24 - Section 6.3 Inference for a Difference in Proportions Question180f19 -/1 E View Policies Current Attempt in Progress TYPE OF FACE MASK During the 2020 COVID-l? pandemic, the United States faced a shortage of medical face masks. To save the medical masks for health care workers, people were advised to wear cloth masks in public. Here we look at the only randomized experiment1 (at the time of writing) that investigated the effectiveness of cloth masks at protecting the wearer from respiratory infection. 1 Maclntyre CR, Seale H, Dung TC, et al. (2015). A cluster randomised trial of cloth masks compared with medical masks in health care workers. BMJ Open, 5(4). Health care workers in Vietnam were randomized to wear either medical face masks or cloth face masks during their shifts for four consecutive weeks. The study was designed to see if there is evidence that medical masks are more effective at preventing illness. The researchers collected data on three different response variables: whether the health care worker developed a clinical respiratory illness (CRI), inuenzalike illness (ILI), or laboratory-conrmed virus illness (LCI). The results are given in the table below. Sampte size Clll [LI LU Medical masks 5m 28 1 19 10th IDES 4 l 1 For all three response variables, the proportion of infection was higher for the cloth mask wearers than the medical mask wearers. Recall from Section 1.3 that there are three possible explanations for this: (i) Causal association: medical masks are more effective at preventing infection than cloth masks (ii) Association due to confounding: the groups (cloth mask wearers and medical mask wearers) differed to begin with (iii) No actual association: just random chance Cloth vs Medical Masks and Inuenza-Like Illness For the inuenza-like illness (ILI) response variable, (a) Calculate the observed difference in proportions with illness, medical mask cloth mask. Round to three decimal places. (b) Use technology and the normal distribution to calculate a one-sided p-value. Round to three decimal places. (c) Are the conditions for using a normal distribution satised here? v (d) Use a randomization test to calculate a onesided p-value. Round to three decimal places. (e) Based on the pvalue, do we have evidence against alternative explanation (iii)? v (f) Based on your answers to Exercise 6.193 and part (e), does this study provide convincing evidence that medical masks are more effective at protecting health care workers against developing an influenzalike illness than cloth masks? v eTextbook and Media Save for Later Attempts: O of 4 used Question List X Question 1 0.5 / 1 Mu tipart D Partially correct Question 2 0.5 /1 Mu tipart D Partially correct Question 3 1 / 1 Mu tipart J Correct Question 4 1 / 1 Mu tipart ~/ Correct Question 5 _ / 1 Ma h Fill in the Blank Not started Question 6 _ I 1 Math Fill in the Blank Not started Question 7 _ / 1 :lgilkH\" In the Unsubmitted work Question 8 _ / 1 Math Fill in the Blank Not started Question 9 / 1 Mu tipart Not stared Question 10 -/1 Mu tipart Not stared Question 11 -/1 Mu tipart Not stared Question 12 / 1 Mu tipart Not stared Question 13 / 1 Mu tipart Not stared Question 14 / 1 Mu tipart Not started Question 15 / 1