Event Schema Stimuli
datasetposted on 24.07.2018 by Christopher Baldassano
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
The stimuli were designed to conform to two naturalistic schematic scripts that we expected to be familiar to all our subjects: eating at a restaurant, or catching a flight at an airport. Each scenario consisted of four events. For the restaurant stories, the events were: entering and being taken to a table, sitting with menus, ordering food and waiting for its arrival, and food arriving and being eaten. For the airport stories, the events were: entering the airport, going through the security checkpoint, walking to and waiting at the gate, and getting onboard the airplane and sitting in a seat.
Each story was approximately 3 minutes long. To identify schema representations that were modality-invariant, we presented 4 audio-visual movies and 4 spoken narratives for each of the two schemas. The stories all involved different characters and spanned multiple genres, sharing only the same high-level schematic script. The movies were sampled from films in which the restaurant schema was depicted (Brazil, Derek, Mr. Bean, Pulp Fiction) or the airport schema was depicted (Due Date, Good Luck Chuck, Knight and Day, Non-stop), and were edited for time and to conform as closely as possible to the four-stage schema script. The audio narratives were adapted from film scripts with a restaurant scene (The Big Bang Theory, The Santa Clause, Shame, My Cousin Vinny) or an airport scene (Friends, How I Met Your Mother, Seinfeld, Up in the Air), also edited for length and to match the schematic script. All narratives were read by the same professional actor.
To generate scrambled versions of the stimuli, each story was divided into its 4 schematic events, and then these clips were concatenated to create 16 new scrambled stimuli (named AA1, etc.). Like the original stories, each of these scrambled stimuli contained all 4 schematic events from one schema and consisted of clips from a single modality (audio or audiovisual). Unlike the original stories, the schematic events in these clips were presented out of order, and were drawn from 4 different stories. 8 different "highly scrambled" permutations of 4 subevents (for which no two neighboring events were in the correct sequence, and at most 1 event was in the correct position in the sequence) were used for the 8 stories within each schema.