Confidence interval for an odds ratio

Source: R/statpsych3.R

Computes a confidence interval for an odds ratio with .5 added to each cell frequency. This function requires the frequency counts from a 2 x 2 contingency table for two dichotomous variables.

ci.oddsratio(alpha, f00, f01, f10, f11)

Arguments

alpha

alpha level for 1-alpha confidence

f00

number of participants with y = 0 and x = 0

f01

number of participants with y = 0 and x = 1

f10

number of participants with y = 1 and x = 0

f11

number of participants with y = 1 and x = 1

Value

Returns a 1-row matrix. The columns are:

  • Estimate - estimate of odds ratio

  • SE - standard error

  • LL - lower limit of the confidence interval

  • UL - upper limit of the confidence interval

References

Fleiss JL, Paik MC (2003). Statistical Methods for Rates and Proportions, 3rd edition. Wiley.

Examples

ci.oddsratio(.05, 229, 28, 96, 24)
#>      Estimate        SE       LL       UL
#> [1,] 2.044451 0.6154578 1.133267 3.688254

# Should return:
#      Estimate        SE       LL       UL
# [1,] 2.044451 0.6154578 1.133267 3.688254