The exam set contains 8 questions. The questions may contain sub-questions. Make sure to indicate which question you are answering. The questions are weighted according to the percentage in brackets. Please remember to report your arguments, calculations etc. Question 1 (16%) A teacher conducted an experiment to test whether new directed reading activities in the classroom will help elementary school pupils improve their reading ability. She arranged for a class of 11 students ( treatment group) to follow these activities for an 8-week period. A control class of 10 students ( control group) followed the same curriculum without the activities. (The two groups of students were selected independently with no attempt to match). The experiment was ended by a test, where the students were asked to read a text. The reading time (in minutes) was measured and registered in the table below. Reading Time [min.] Group Reading Time [min.] Group 5 treatment 8 control 6 treatment 8 control 6 treatment 9 control 7 treatment 9 control 7 treatment 9 control 7 treatment 10 control 8 treatment 10 control 8 treatment 10 control 9 treatment 11 control 9 treatment 11 control 10 treatment Questions: a) What test would you perform to compare reading time in two groups of students? Do the data satisfy the assumptions for this test? b) What null hypothesis do you choose for your test? c) Perform the test. Explain the results. d) What is your conclusion about the effect of new directed reading activities on pupils reading ability?
Question 2 (18%) Description: A group of researchers wanted to investigate whether a new noninvasive method for measuring transmitral volumetric flow (MF) could replace an invasive measure of left ventricular stroke volume (SV). The researchers performed both measurements in 11 healthy subjects as an initial test. The table below shows the results of the two measurements. The researchers defined a clinically acceptable limit of agreement between the two methods of ±16 cm 3. Subject MF (cm 3 ) SV (cm 3 ) 1 47 43 2 68 72 3 70 60 4 73 72 5 79 92 6 85 85 7 87 90 8 104 94 9 112 108 10 120 131 11 132 131 a) The plot below shows a Bland-Altman plot comparing the two measurements. Explain what you can conclude from the plot. b) In addition to Bland-Altman plots, which other statistical methods are appropriate to use to compare the two measurement techniques? c) Perform the appropriate analyses in SPSS. Explain what you have done and what can be concluded from the results.
Question 3 (15%) A group of researchers wanted to investigate the survival probability in a cohort of patients. They formed a logistic regression model with age as a covariate, gender and smoking habit as categorical factors. a) What is the best single predictor for survival in this group of patients? b) Calculate the regression coefficients for the logistic regression model. c) What is the probability that a 45 year old smoking female patent in the study will survive. Assume that the regression constant, 0, is -3.0. Parameter odds ratio P Age 1.14 0.02 Gender, female 1.60 0.18 Smoking, Yes 2.76 0.01 Question 4 (18%) A group of researchers have developed 3 magic pills against stress. To test the effect of the anti-stress pills the recruited 20 stress patients and randomly assigned them into 4 groups. One group received a placebo pill and the 3 other groups received one of the magic pills each. The patients were asked to report how long time lasted before they felt stress again. The time-to-stress is shown in the table below (in months). The researchers wanted to compare the 3 treatment groups with the control group. Placebo Pill 1 Pill 2 Pill 3 0.15 1.34 0.23 2.03 0.02 0.14 0.04 0.27 0.16 0.02 0.34 0.92 0.37 0.08 0.16 1.07 0.22 0.08 0.05 2.38 a) The researchers were considering performing a 1-way ANOVA. Which assumptions must the time-to-stress data fulfill? b) The researcher performed normality tests on the time-to-stress for each group as shown in the table below. What should they conclude? c) Perform the appropriate test for effect of the pills.
dimension1 Tests of Normality treatment Kolmogorov-Smirnov a Shapiro-Wilk Statistic df Sig. Statistic df Sig. outcome control,194 5,200 *,965 5,842 T1,433 5,003,623 5,001 T2,217 5,200 *,925 5,566 T3,221 5,200 *,945 5,703 a. Lilliefors Significance Correction *. This is a lower bound of the true significance. Question 5 (18%) Two randomized, controlled trials comparing Drug1 and Drug2 in the treatment of a disease X were identified from Medline search. In one, there were 77 cases of clinical cure out of 94 patients treated by Drug1 compared with 67 cases of clinical cure in 100 patients treated by Drug2; in the other trial the clinical cure rates were 255/276 for the Drug1 group compared with 241/262 for the Drug2 group. Questions: a) Why did researches choose a randomized controlled trial design for these two studies? Would you use other designs? b) What is the overall odds ratio (for receiving Drug1 therapy among clinically cured patients compared to those who were not clinically cured) across two studies? Explain the result c) What is the confidence interval for this overall odds ratio? What does this confidence interval mean? d) Is there significant heterogeneity between the results in these two studies? Question 6 (5%) In addition to the scientific field, list the key factors for determining where to apply for funds from the Danish research councils and foundations. Question 7 (5%) a) List the ethical issues which should be considered in the trial protocol before its submission to a regional ethical committee. b) At which point should trial subjects be informed of ethical issues?
Question 8 (5%) Describe at least three weaknesses in the following patent claim. A device for determining cardiac contractility from electrocardiography relating to an individual, comprising three Ag-AgCl electrodes for non-invasive recording of electrical signals on the chest of the individual, a data collection device, having means for collecting measurements from the three Ag- AgCl electrodes, a laptop, for receiving data from the data collection device and determining the cardiac contractility. This is the end of the exam set