Tutorial.
Contact.
Tonight was a tutorial on medical statistics by Julie Morris.
The questions she brought were:
Example Statistics Questions
1. For a sample of 500 women giving birth, blood samples were drawn from the umbilical cord artery immediately after fetal delivery, before delivery of the placenta and the ph of cord blood gas measured. Summary statistics of the ph readings were calculated. (Answer yes/no to each statement)-
A The median is the ph value which exactly divides the readings into two halves
B If the readings are found to have a negatively skewed distribution, the mean will be greater than the median.
C The interquartile range will contain 25% of the ph readings
D The mode is the most frequently observed value
E The arithmetic mean is obtained by dividing the sum of all readings by the total number of readings
2. For a cohort of babies, birthweight was considered to be normally distributed with a mean of 3000g and a standard deviation of 500g. It follows that: (Answer yes/no to each statement)-
A All birthweights must be less than 4500g
B The distribution of birthweights is skewed
C About 50% of birthweights are above 3000g
D About 2.5% of birthweights are below 2000g
E About 95% of birthweights are between 2500g and 3500g
3. Theme: Summary statistics
A mean F positive predictive value
B standard deviation G percentage
C range H mode
D median I sensitivity
E interquartile range J standard error
For each type of data description given below choose the SINGLE most appropriate summary statistic. Each option may be used once, more than once, or not at all.
1. The wound infection rate in a cohort of 200 women following a caesarean.
2. The variability of length of hospital stay for 8 mothers following a caesarean, where length of stay has a positively skewed distribution].
3. A measure relating to the probability that a diagnostic test for a fetal disorder carried out in antenatal screening will be positive if a fetal disorder is actually present.
4. To assess the effectiveness of a new antenatal screening test, 100 pregnant women over 45 years of age, were screened. 30 women subsequently had babies with mild congenital problems. 21 of these 30 women had a positive screening test result. 28 of those with normal babies also had a positive test result. (Answer yes/no to each statement).
A The predictive value of a positive test result as a marker for congenital problems is 21/30
B The specificity of the test will depend upon the prevalence of congenital problems in the cohort of women to which it is applied
C The sensitivity the screening test is 21/49
D The positive predictive value of the screening test is 21/30
E The predictive value of the test might be different in another population
5. To compare oxidative stress following spontaneous vaginal delivery with that induced by Oxytocin augmented delivery, mean cord blood GSH was computed and compared. Cord blood GSH was found to be significantly lower in the oxytocin group (p=0.02). (Answer yes/no to each statement)-
A There was a large effect of Oxytocin on GSH
B A p-value of 0.02 for the difference between the two group means is less significant than one of 0.05
C A p-value is always small for small studies
D As the difference between the two groups is significant, it means that the null hypothesis is true
E In general, a p-value can take any value anywhere between -1 and 1
6. In a study looking at the relationship between gestational age at birth and neurological development at one year in 100 babies, the correlation coefficient, was found to equal 0.61 with an associated p-value of p<0.001.
The best interpretation of this result is (choose just one):
A Neurological development increases by 0.61 for each day of gestational age
B Pre-term birth causes an inhibition in neurological development
C Age and neurological development are unrelated because the correlation value is not less than 0.05.
D The small p-value means that the sample size was not large enough and hence there is no clear interpretation of the result.
E There is a significant positive relationship between neurological development and age
7. Days of admission to a neo-natal intensive care unit were recorded for babies born to women with PPROM (preterm premature rupture of membranes). Assuming length of stay in NICU has a highly skewed distribution, what would be the most appropriate method of analysis to compare length of stay between women with PROM < 31 weeks and women with PROM ≥ 31 weeks? (choose just one).
A Chi-square test
B Correlation coefficient
C Mann-Whitney U-test
D Regression analysis
E Two-sample t-test
8. In a small randomized trial comparing birth positions in labour (squatting using a birth cushion vs. conventional management) the number of forceps in the birth cushion group was half that in the conventional management group, but the difference was not statistically significant. What is the best interpretation of this result? (Choose just one statement).
A Reduction in forceps delivery is so great that we should change current practice and be proactive in encouraging the use of squatting cushions
B The result shows that there is little evidence of a beneficial effect of squatting cushions.
C We should carry out a study of much greater size
D The non-significant results could be due to confounding
E We should keep adding women to the trial until the test for the comparison of forceps deliveries shows a significant difference
9. A cohort of women experiencing nausea and vomiting in pregnancy were given acid-reducing medication. The Pregnancy-Unique Quantification of Emesis and nausea (PUQUE) score was recorded before and after treatment. The clinical effect of the medication could be evaluated using: (Answer yes or no to each statement):
A The Student paired t-test
B The Spearman correlation coefficient
C The Wilcoxon matched-pairs test
D The Sign test
E The Mann-Whitney U-test
10. Theme: Statistical significance tests
A Chi-square test F Spearman correlation
B Two-sample t-test G Mann-Whitney U-test
C Friedman test H Wilcoxon test
D Fisher’s Exact test I Paired t-test
E Linear regression J Sign test
For each of the studies described below choose the SINGLE most appropriate statistical test. Each option may be used once, more than once, or not at all.
1. In a cohort of 200 healthy women having their first baby, 24-hour blood pressure readings were taken in early-, mid- and late-pregnancy. The aim of the study is to see whether it is possible to predict the birthweight of the baby from blood pressure at 28 weeks gestation.
2. The numbers of episiotomies in grand multiparas (≥ 5 previous deliveries) and multiparas (1-4 previous deliveries) are compared. There were 1/20 required in grand multiparas and 4/20 required in multiparas.
3. Adrenaline concentrations were measured in 30 mechanically-ventilated babies at 1 hour and 24 hours to assess whether there had been a significant change. The adrenaline concentrations can be assumed to followed a Normal distribution.
Fauzia had also brought a question from a previous exam paper.
Statistics MCQ from Fauzia from past paper. February 2012
In a prospective blind study of a possible new method of AN screening for a fetal disorder, 60,000 consecutive pregnant women were recruited & tested. 100 fetuses were found to be affected. The test had a sensitivity of 90% & specificity of 95%.
The following statements are correct.
1. A women with a +ve test has a 10% chance of having an affected child .
2. The results demonstrate that the test fulfils WHO criteria for screening.
3. 95% of the affected cases have +ve screening test.
4. The false +ve rate can be calculated using Fisher test
If you send me your answers, I'll send Julie's.
