The Differences between Suicide Ideators and Suicide Attempters: Simple, Complicated, or Complex?

Abstract

Background: Given that many people think about suicide and relatively few attempt suicide, researchers have proposed that accurate differentiation between ideators and attempters can aid suicide prevention efforts. Ideators and attempters might differ in three possible ways. First, they might differ in a simple way such that one or a small set of factors are both necessary and sufficient to distinguish between the two groups. Second, ideators and attempters might differ in a complicated way such that a specific combination of a large set of factors is necessary and sufficient for the distinction. Third, complex differences might exist: many possible combinations of a large set of factors are sufficient to distinguish the two groups, but no combination is necessary. This study empirically examined the difference between suicide ideators and attempters. To support a simple difference, individual factors need to produce high classification accuracy. To support complicated differences, one optimized algorithm including a large number of factors needs to produce high accuracy, but removal of factors from the algorithm should substantially reduce the accuracy. To support complex differences, multiple algorithms including a large number of factors need to produce similarly high accuracy, and meanwhile not all algorithms should lead to high accuracy.Methods: Across five samples (total N = 3,869), univariate logistic regressions were conducted to test for simple differences. To test for complicated and complex differences, Machine learning (ML) methods were used to identify the optimized algorithm with all available variables. Subsequently, the same methods were repeated after removing the top five most important or discriminative variables, and a randomly selected 10% subset of variables. Multiple logistic regressions were conducted with all variables.Results: Results were highly consistent across samples. Univariate logistic regressions on average yielded chance-level accuracy. ML algorithms with all variables or with a subset of variables removed showed good accuracy, whereas multiple logistic regressions produced poor to fair accuracy.Conclusions: The study found that differences between suicide ideators and attempters are complex rather than simple or complicated. The findings are consistent with broader psychological evidence and bear important implications for research and clinical practice.