Empirical validation of the diffusion model for recognition memory and a comparison of parameter-estimation methods

Abstract

The diffusion model introduced by Ratcliff (Psychol Rev 85:59-108, 1978) has been applied to many binary decision tasks including recognition memory. It describes dynamic evidence accumulation unfolding over time and models choice accuracy as well as response-time distributions. Various parameters describe aspects of decision quality and response bias. In three recognition-memory experiments, the validity of the model was tested experimentally and analyzed with three different programs: fast-dm, EZ, and DMAT. Each of three central model parameters was targeted via specific experimental manipulations. All manipulations affected mainly the corresponding parameters, thus supporting the convergent validity of the measures. There were, however, smaller effects on other parameters, showing some limitations in discriminant validity.

Fig. 1

Schematic illustration of the diffusion model. The process starts at the starting point z and accumulates information over time until one of two thresholds is reached. The speed of information accumulation is indicated by the drift rate v. Due to random influences the process is not linear, but fluctuates between the thresholds. The upper threshold a is associated with the old response, the lower threshold 0 is associated with the new response. As soon as a threshold is reached the corresponding response is initiated. Adapted from “An illustration of the random walk and diffusion process, together with relatedness distributions that drive the diffusion process” by Ratcliff (1978), A theory of memory retrieval. Psychological Review, 85, p. 64, and “Schematic illustration of the diffusion model” by Voss et al. (2004), Interpreting the parameters of the diffusion model: an empirical validation. Memory & Cognition, 32, p. 1207

Fig. 2

Mean fast-dm parameter estimates for new-bias and old-bias conditions in Experiment 1. Bars represent standard deviation. We show the absolute values of the drift rates. z/a represents the bias parameter, a the threshold parameter, v _old the drift rate for old items, v _new the drift rate for new items, and t ₀ the response-time constant

Fig. 3

Mean DMAT parameter estimates for new-bias and old-bias conditions in Experiment 1. Bars represent standard deviation. We show the absolute values of the drift rates. z/a represents the bias parameter, a the threshold parameter, v _old the drift rate for old items, v _new the drift rate for new items, and t ₀ the response-time constant

Fig. 4

Mean fast-dm parameter estimates for speed and accuracy conditions in Experiment 2. Bars represent standard deviation. We show the absolute values of the drift rates. z/a represents the bias parameter, a the threshold parameter, v _old the drift rate for old items, v _new the drift rate for new items, and t ₀ the response-time constant

Fig. 5

Mean DMAT parameter estimates for speed and accuracy conditions in Experiment 2. Bars represent standard deviation. We show the absolute values of the drift rates. z/a represents the bias parameter, a the threshold parameter, v _old the drift rate for old items, v _new the drift rate for new items, and t ₀ the response-time constant

Fig. 6

Mean EZ parameter estimates for speed and accuracy conditions in Experiment 2. Bars represent standard deviation. a represents the threshold parameter, v the drift rate for correct answers, and t ₀ the response-time constant

Fig. 7

Mean fast-dm parameter estimates for not presented items, items presented once and items presented twice conditions in Experiment 3. Bars represent standard deviation. We show the absolute values of the drift rates. z/a represents the bias parameter, a the threshold parameter, v the drift rate, and t ₀ the response-time constant

Fig. 8

Mean DMAT parameter estimates for not presented items, items presented once and items presented twice conditions in Experiment 3. Bars represent standard deviation. We show the absolute values of the drift rates. z/a represents the bias parameter, a the threshold parameter, v the drift rate, and t ₀ the response-time constant

Fig. 9

Mean EZ parameter estimates for not presented items, items presented once and items presented twice conditions in Experiment 3. Bars represent standard deviation. We show the absolute values of the drift rates. a represents the threshold parameter, v the drift rate, and t ₀ the response-time constant