Question

Part of the question highlighted in Blue. Thanks!

1 If babies are born prematurely, there is a risk that they wll be under-developed and suffer a variety of health problems. There is good evidence that mothers who give birth prematurely in one pregnancy will also do so in later pregnancies, so in the following studies, the subjects were mothers who had given birth prematurely siniz) an a Cn-1)d in their first pregnancy () Clinicians designed and conducted a study to assess whether a drug extended pregnancy, by comparing duration of the second pregnancy for two independent groups each of 40 mothers. One group was given the active drug and the other a placebo version. The results of the trial (in CSC (x) weeks) were as follows: Group mean s.d Drug Placebo 38.2 1.60 Sinh (x (a) Verify that the trial had an adequate sample size to test for a differ- ence of 1 week, when results expected on the placebo were a mean 2 2 ec Cx) duration of 38 weeks, with a standard deviation of 1.5 weeks (4 marks) Did the trial provide evidence that the drug is beneficial in terms of (4 marks) (b) extending pregnancy? (ii) When writing up their study, as well as assessing the effect on duration of pregnancy, the clinicians decide to examine whether the drug has any effect on the birthweights of the babies. Differences in mean birthweight between 2 Ta rallelogrant bh the two groups give a p-value for the appropriate t-test of p-0.03. (a) (b) (c) Give 2 reasons why including this additional investigation is prob- (2 marks) lematic What should they conclude about the effect of the drug on birth (3 marks) csch (2)-cos (iz) weights? Explain your reasoning. Does this change the way you would evaluate the effect of the drug (1 mark) on pregnancy duration in (i)(b)'? in,(-x):-sin (x) P(x) csch(x)- (ex-e ) (iii In fact, it is subsequently revealed that one of the clinicians involved in the trial did not wait until the end of the trial for the formal test of effect of the drug on duration. Instead he also looked at data so far the available at 3 interim time points, after roughly each quarter of the patients had had Q(x) their babies. A summary of the t-tests he conducted is: Group Number on Drug Number on Placebo p-value Time point 1 Time point 2 ime point 3 20 0.065 0.081 0.056 (a) (b) Give 2 reasons why this might have been sensible. Explain why there might be a problem in adopting this approach. (2 marks) Download fron Dreamstime. 70138800 (2 marks) Somjai Jaithieng Dreamstime.com

Answer 1

Smpk SR'n'nuind ngịnh 눴. @h)<E|(m11xr Sundittin-the values n>M29 Stme the Sampe Ssze s 4029. Sompes as dled o louo: 2 1.

Clasm, nan Y a Sample t2362 Calculalions bar 38.2 Crl 664625 - va lye