Omitted variable bias, classical and partial-R2 form

Formula: bias = delta gamma; equivalently |bias| = se(tau_res) sqrt(R2_YZ.DX R2_DZ.X / (1 - R2_DZ.X)) sqrt(df)

Usage

Ovbias(
  delta = NULL,
  gamma = NULL,
  estimate = NULL,
  se = NULL,
  df = NULL,
  r2_yz = NULL,
  r2_dz = NULL
)

Arguments

delta: See Usage.
gamma: See Usage.
estimate: See Usage.
se: See Usage.
df: See Usage.
r2_yz: See Usage.
r2_dz: See Usage.

Value

List with bias, adjusted_estimate, adjusted_se, adjusted_t, relative_bias, bias_factor.

References

Cinelli and Hazlett (2020), Making Sense of Sensitivity: Extending Omitted Variable Bias, JRSS-B 82:39-67. Verified against the author's copy of the paper: bias = delta gamma (Section 4.1), equation (12) for the adjusted standard error, equation (13) for the bias in partial-R2 form, equation (14) for the relative bias.

Examples

Ovbias()
#> $bias
#> [1] NA
#> 
#> $adjusted_estimate
#> [1] NA
#> 
#> $adjusted_se
#> [1] NA
#> 
#> $adjusted_t
#> [1] NA
#> 
#> $relative_bias
#> [1] NA
#> 
#> $bias_factor
#> [1] NA
#> 
#> $method
#> [1] "Omitted variable bias (Cinelli-Hazlett)"
#> 
#> attr(,"class")
#> [1] "morie_rich_result" "list"