In logistic regression, the statistic formed by multiplying the log-likelihood by minus two is commonly denoted by what symbol?

In logistic regression, the statistic formed by multiplying the log-likelihood by minus two is commonly denoted by what symbol?

• A. -2LL
• B. Log-Likelihood
• C. Likelihood ratio
• D. 2LL

Answer: (A)

In logistic regression, the log-likelihood measures how well the model fits the data. If you take that log-likelihood, usually denoted L, and multiply it by minus two, you get the statistic commonly written as -2LL. This form is central to likelihood ratio testing: you compare -2LL from a full model to -2LL from a simpler, nested model. The difference between those values helps determine if adding parameters significantly improves fit, and this difference is approximately chi-square distributed with degrees of freedom equal to the number of parameters added. The other expressions don’t capture this specific statistic. The log-likelihood alone is just L, not the product with -2. A likelihood ratio refers to the ratio of likelihoods between models, not this product. And 2LL omits the crucial minus sign, which changes both interpretation and the reference distribution used for testing.

In logistic regression, the log-likelihood measures how well the model fits the data. If you take that log-likelihood, usually denoted L, and multiply it by minus two, you get the statistic commonly written as -2LL. This form is central to likelihood ratio testing: you compare -2LL from a full model to -2LL from a simpler, nested model. The difference between those values helps determine if adding parameters significantly improves fit, and this difference is approximately chi-square distributed with degrees of freedom equal to the number of parameters added.

The other expressions don’t capture this specific statistic. The log-likelihood alone is just L, not the product with -2. A likelihood ratio refers to the ratio of likelihoods between models, not this product. And 2LL omits the crucial minus sign, which changes both interpretation and the reference distribution used for testing.