BCPS ACCP Study Materials QUESTIONS
& CORRECT 100%
Types of random variables - ANSWER discrete and continuous
discrete - values are limited (nominal and ordinal)
continuous - values are not (interval and ratio)
Discrete Variable - ANSWER Can only be limited number of values within a given
range
Nominal and ordinal data
Discrete Variable Subtypes - ANSWER Nominal - classification into groups that are
not ordered or ranked (yes/no; gender; race; etc.)
Ordinal - classification into groups that are ranked or tied to severity
Nominal data - ANSWER type of discrete data that is classified into groups that are
not ordered or ranked (yes/no; gender; race; etc.)
Ordinal data - ANSWER type of discrete data that is classified into groups that are
ranked or tied to severity (pain scales, NYHA functional classifications, etc.)
, Means and standard deviations should not be used for this data (but often is)
Continuous variables - ANSWER can be any value within a given range and
numbers are ranked
subtypes: interval and ratio
Interval data - ANSWER type of continuous variable - numbers are ranked
Zero is arbitrary (in ratio zero is significant)
Percentile Interpretation - ANSWER 75th percentile = 75% of all values in set are
smaller than that data point
it does NOT assume normal distribution
Distribution of percetiles - ANSWER can be any; normal is not assumed
Interquartile range - ANSWER Middle 50% of the data set (25-75% percentiles)
Kolmogorov-Smirnov test - ANSWER a statistical test of normal distribution
this is the formal way to decide but it should be easy to see when data is shown
on histogram
, Standard error of the mean - ANSWER Quantifies uncertainty in the estimate of
the mean (does not quantify variability of the sample)
SD/square root of sample size
95% CI = +/- 2x SEM
95% CI Calculation - ANSWER Plus or minus 2x the SEM
SEM = SD / square root of sample size
What do p-values tell us? - ANSWER If there is a difference between groups but
not how big that difference is (CI's tell us more about magnitude of difference)
Do you need to show both p-value and CI? - ANSWER Nope
Rejecting and not rejecting null hypothesis meaning - ANSWER Rejecting: there is
a statistical difference between groups
Not rejected: no statistical difference; but not saying that the groups are equal
Hypothesis Test for non-directional difference between groups - ANSWER 2-sided
t-test or confidence intervals
& CORRECT 100%
Types of random variables - ANSWER discrete and continuous
discrete - values are limited (nominal and ordinal)
continuous - values are not (interval and ratio)
Discrete Variable - ANSWER Can only be limited number of values within a given
range
Nominal and ordinal data
Discrete Variable Subtypes - ANSWER Nominal - classification into groups that are
not ordered or ranked (yes/no; gender; race; etc.)
Ordinal - classification into groups that are ranked or tied to severity
Nominal data - ANSWER type of discrete data that is classified into groups that are
not ordered or ranked (yes/no; gender; race; etc.)
Ordinal data - ANSWER type of discrete data that is classified into groups that are
ranked or tied to severity (pain scales, NYHA functional classifications, etc.)
, Means and standard deviations should not be used for this data (but often is)
Continuous variables - ANSWER can be any value within a given range and
numbers are ranked
subtypes: interval and ratio
Interval data - ANSWER type of continuous variable - numbers are ranked
Zero is arbitrary (in ratio zero is significant)
Percentile Interpretation - ANSWER 75th percentile = 75% of all values in set are
smaller than that data point
it does NOT assume normal distribution
Distribution of percetiles - ANSWER can be any; normal is not assumed
Interquartile range - ANSWER Middle 50% of the data set (25-75% percentiles)
Kolmogorov-Smirnov test - ANSWER a statistical test of normal distribution
this is the formal way to decide but it should be easy to see when data is shown
on histogram
, Standard error of the mean - ANSWER Quantifies uncertainty in the estimate of
the mean (does not quantify variability of the sample)
SD/square root of sample size
95% CI = +/- 2x SEM
95% CI Calculation - ANSWER Plus or minus 2x the SEM
SEM = SD / square root of sample size
What do p-values tell us? - ANSWER If there is a difference between groups but
not how big that difference is (CI's tell us more about magnitude of difference)
Do you need to show both p-value and CI? - ANSWER Nope
Rejecting and not rejecting null hypothesis meaning - ANSWER Rejecting: there is
a statistical difference between groups
Not rejected: no statistical difference; but not saying that the groups are equal
Hypothesis Test for non-directional difference between groups - ANSWER 2-sided
t-test or confidence intervals
