Do sequential lineups impair underlying discriminability?

Cognitive Research: Principles and Implications

Table 6 Likelihood ratio tests comparing fits of unconstrained models to a series of constrained models where equality for each parameter is imposed across the simultaneous and sequential conditions

Significant p values indicate that model fit significantly worsened when a parameter was constrained to be equal across the simultaneous and sequential conditions. For each unconstrained model, we fit SDT-SEQ to the sequential data and SDT-MAX to the simultaneous data. The unconstrained models had 16 degrees of freedom, fixing one parameter increases the degrees of freedom to 17, χ²(17) - χ²(16) = χ²(1), thus the χ² tests above have one degree of freedom