Fig 1.
The metal pipe for administration of food reinforcement extended out of the arena on the leftmost side. Right: The door opening mechanism of the restrainer; a lever can be pushed down to tip open a door attached to a counterweight.
Table 1.
Overview of experimental conditions across sessions.
Fig 2.
From left to right: Overview of the arena, day 1, day 4 and the last day (7) of shaping.
Rats were given food reinforcement for spending time in the quadrant (top left corner) of the box containing the opening mechanism, door, and food box. Red/bright blue areas indicate areas where the rats spent the most time during the trials.
Fig 3.
Door opening (boxplot) and percentage of openings (solid line) across the experimental conditions (left and right ordinate, respectively).
Door opening latency based on mean of three trials was shorter when the restrainer contained a trapped cagemate than when it was empty but longer than when the restrainer contained food. The box whiskers extend to the most extreme data points lying within 1.5 interquartile range. The percentage of openings were calculated by number of subjects that opened the restrainer door at least once during the condition, divided by total number of subjects.
Table 2.
Nemenyi’s comparisons of condition.
Fig 4.
Cumulative incident plot showing percentage of not opening across time.
All conditions with a trapped cagemate (CM1 through DL) show a decline with a steepness in-between the Food and Empty conditions. Each step down represents rat(s) opening the restrainer in the different conditions and are based on subject mean latency of the three trials in each condition.
Fig 5.
Average data illustrated in Figs 3 and 4 broken down into individual trials.
The Food condition had the lowest and most stable latency, the Empty condition had a rising latency across trials, and the conditions where the restrainer contained a cagemate had a semi-stable pattern with an intermediate latency compared to Food and Empty.