Statistics Flashcards
The variance for height (or weight) in a population may be expressed in percentiles or standard deviation scores. One standard deviation below the population mean (-1 SD) approximates which one of the following percentiles?
Values within 1 standard deviation account for 68% of the values
Values within 2 standard deviations account for 95% of the values
Values within 3 standard deviations account for 99.7% of the values
Assuming that height and weight is normally distributed –
Mean - 1SD = (31.7/2) = 15th percentile of a normal distribution
Mean - 2SD = (4.5/2) = 2.5th percentile of a normal distribution
Which randomisation method is used when the number in the population being randomised is small?
Block randomisation is used to keep the numbers in the allocation groups of a randomised controlled trial equal, particularly when the population being randomised is small and there’s a risk of allocating significantly different sized groups. It works by dividing the allocation in blocks which contain an even number of allocations per group. For example, you could use a block size of 4, where every 4 allocations, there are 2 allocations to both groups.
Block randomisation is useful when you have already stratified your population into smaller groups.
Stratification is useful when you want to account for a variable that is likely to influence the risk of the outcome. For example, you can stratify by age group or sex.
A screening program at a large metropolitan hospital is using pulse oximeters to screen for undiagnosed heart disease. Of 50 babies born in one day, two were found to have undiagnosed congenital heart disease.
Of the following options, which parameter of congenital heart disease can best be calculated by the given data?
Prevalence. This is the number of affected cases in a specific population at a given point in time. So in the stem above, this will be 2/50.
Incidence is the number of new cases that develop over a specific period. In this stem, the screening program was conducted in one day.
Relative risk is an estimate of the extent of the association between exposure to a risk factor and a disease. It can be considered as the likelihood of developing a certain disease after exposure to a certain risk factor, or incidence of disease in the exposed group divided by the incidence of disease in the unexposed group. We are not looking at exposures in this current question.
Odds ratio. This is the probability of exposure in those with the disease compared with the probability of exposure in those without the disease.
Which of the following statistical tests should be performed to reduce the risk of a Type 1 error?
Bonferroni adjustment. The usual p value is set at 0.05. This is considered an acceptable false positive rate. However, if multiple outcomes are assessed, then it is much more likely that there will be a false positive. A Bonferroni adjustment to the p value accounts for this.
A test of Power is done to assess the risk for a type 2 error. This is usually done before a trial has begun. It will suggest the recommended number of patients needed to have an adequate sample size so that the risk of a type 2 error is reduced.
