Measuring event-order memory with a ranking task

The ranking procedure requires participants to rank the entries on a line-up memory test where there is a single old item and n-1 novel foils; the ranking is from the perceived most likely target (with a rank of 1) to the least likely target (with a rank of n). This assessment procedure results in a critical test for the two-high threshold model (Chechile & Dunn, 2021). Moreover, ranking data can readily be used to construct a hazard function, which can be useful for assessing any model of recognition memory. In the current paper, the ranking procedure is employed to examine the memory of order. After the presentation of a series of items for study, a random triplet of these items is tested to assess the memory of the relative order of the items. Chechile and Pintea (2021) previously developed an Event Order (EO) model for measuring four states of triplet order. They also provided evidence that order knowledge is a separate attribute of memory from item content. Chechile and Pintea (2021) estimated the four-states of order knowledge from data obtained that used a series of forced-choice tests. In the current paper, it is shown that the ranking procedure can be also used to estimate the parameters of the EO model. The ranking test method also provides a way to generate an empirical hazard function for memory order, which can be useful for comparing rival models of memory order.