When the method is used in fitting

When the method is used in fitting logistic models in datasets giving rise to separation, theaffected estimate is typically approaching a boundary condition. As a result, the likelihood profileis often asymmetric under these conditions; Wald tests and confidence intervals are liable to beinaccurate. In these circumstances, Heinze and coworkers recommend using likelihood ratio testsand profile likelihood confidence intervals in lieu of Wald-based statistics. Calculation oflikelihood ratio test statistics with the method is done differently by Heinze and coworkers fromwhat is conventionally done: instead of omitting the variable of interest and refitting the reducedmodel, the coefficient of interest is constrained to zero and left in the model in order to allow itscontributing to the penalization. The test statistic is then computed as twice the difference inpenalized log likelihood values of the unconstrained and constrained models by lrtest in a mannerdirectly analogous to that of conventional likelihood ratio tests.

結果 (繁體中文) 1:

When the method is used in fitting logistic models in datasets giving rise to separation, the
affected estimate is typically approaching a boundary condition. As a result, the likelihood profile
is often asymmetric under these conditions; Wald tests and confidence intervals are liable to be
inaccurate. In these circumstances, Heinze and coworkers recommend using likelihood ratio tests
and profile likelihood confidence intervals in lieu of Wald-based statistics. Calculation of
likelihood ratio test statistics with the method is done differently by Heinze and coworkers from
what is conventionally done: instead of omitting the variable of interest and refitting the reduced
模型中，感興趣的係數被限制到零，並留在模型中，以便允許其
有助於懲罰。然後，將檢驗統計量計算為兩倍的差異中
的無約束和約束模型的懲罰數似然值由lrtest的方式
直接類似於常規似然比檢驗。

結果 (繁體中文) 2:

當該方法用於在導致分離的資料集中擬合邏輯模型時，
受影響的估計值通常接近邊界條件。因此，可能性設定檔
在這些條件下，經常是不對稱的;瓦爾德測試和置信區間是
錯誤。在這種情況下，Heinze 和同事建議使用可能性比率測試
和剖面可能性置信區間，以代替基於 Wald 的統計資訊。計算
海因策和同事對該方法的測試統計不同。
是常規的做：而不是省略利息的變數和重新安裝減少
模型，利息係數被約束為零，並留在模型中，以允許其
有助於懲罰。然後，測試統計量的計算值是
懲罰無約束和約束模型的日誌可能性值
直接類似于常規的可能性比測試。

結果 (繁體中文) 3:

當該方法用於對導致分離的數据集進行logistic模型擬合時
受影響的估計通常接近邊界條件。囙此，似然分佈
在這些條件下通常是不對稱的；Wald檢驗和置信區間可能是
不準確的。在這種情況下，海因策和同事建議使用似然比檢驗
用輪廓似然置信區間代替基於Wald的統計。計算
海因策和他的同事用不同的方法進行似然比檢驗統計
傳統做法是：不要忽略興趣變數，而要重新調整
模型中，興趣係數被限制為零，並留在模型中，以便
有助於懲罰。然後，將測試統計值計算為
基於lrtest的無約束約束約束模型的懲罰對數似然值
直接類似於傳統的似然比檢驗。