Create an Observed Over Expected Grid Using the work of Katz et al (1978) calculate acceptable limits for the ratio of two binomial proportions. The Type 1 error rate can be specified, and the limits can be one- or two-sided.

Create an Observed Over Expected Grid

Using the work of Katz et al (1978) calculate acceptable limits for the ratio of two binomial proportions. The Type 1 error rate can be specified, and the limits can be one- or two-sided.

Usage

createObservedOverExpectedTable(
  observedData = NULL,
  observedRate = NULL,
  n = NObserved,
  nHistorical,
  nObservedRange = seq(50, 1500, 25),
  expectedRate,
  historicalRate,
  alpha = 0.05,
  sides = c("two", "lower", "upper")
)

Arguments

observedData: a tibble containing the observed data that will be compared to the calculated limit(s). May be NULL.
observedRate: The column in observedData that contains the observed rates. Ignored if observedData is NULL. Uses tidy evaluation.
n: The column in both observedData (if not NULL) and the output tibble that defines the sample size to which the corresponding limit(s) relate(s).
nHistorical: the number of observations on which the historical rate is based
nObservedRange: a sequence of values for which limits should be calculated
expectedRate: the event rate expected in the current study. Usually the same as historicalRate, but need not be so.
historicalRate: the event rate observed in the historical data
alpha: The Type 1 error rate associated with the calculated limits. Default 0.05
sides: the sidedness of alpha. Either two, upper or lower. Default two.

Value

a tibble