Blocked vs. Interleaved: How range contexts modulate time perception and its EEG signatures
Cemre Baykan, Xiuna Zhu, Artyom Zinchenko, Zhuanghua Shi
General and Experimental Psychology, LMU Munich
Background
Human time estimates tend to gravitate towards the center of the stimuli. While previous studies have indicated that the range of a sample distribution can also influence subjective judgments, the neural underpinnings behind this range effect remain elusive. In this study, we employed both behavioral and electroencephalographic (EEG) measures to examine the influence of the range of sample distribution on temporal reproductions. Participants were asked to reproduce target intervals under two temporal context sessions: a blocked-range (BR) session, where intervals, short or long, were presented separately in lower and upper range blocks, respectively, and an interleaved-range (IR) session, where all intervals were randomly presented. The results showed that reproduction biases became less prominent in the BR context compared to the IR context. In the EEG data, the BR context led to a quicker CNV buildup that sustained a higher amplitude and dissolved sooner, coupled with an enhanced offset P2 amplitude. Collectively, our findings suggest that the precision of the prior, impacted by the range, can modulate anticipation uncertainty, subsequently affecting those time-sensitive EEG components.
Experimental design
Each trial started with a randomly timed fixation, ranging between 1.3 and 2.3 s. This was immediately followed by a 0.7 s appearance of an exclamation mark, signaling the imminent tone and the upcoming “encoding phase”. The tone played for a specific duration (elaborated further in the next paragraph) while the exclamation mark remained on the screen. It was then replaced by a question mark, prompting for the upcoming “reproduction phase”. After 1.5 s, the reproduction phase started with another tone played, while the question mark remained visible. Participants were tasked with stopping this tone - by pressing the spacebar - once it matched the initial tone’s duration.
The experiment comprised two sessions: the blocked range (BR) and the interleaved range (IR) contexts. In the BR session, participants encountered subseconds (0.4, 0.57, and 0.8 s) or supra-second intervals (1.2, 1.7, and 2.4 s) in separate blocks. Crucially, to ensure that interval ranges were non-overlapping, we introduced a 0.4 s chasm between the ranges. The order of the types of blocks was counterbalanced among participants. Meanwhile, in the IR session, participants were exposed to the full range of target intervals - extending from sub-seconds to supra-seconds (0.4, 0.57, 0.8, 1.2, 1.7, and 2.4 s). Participants always started with the BR session, ensuring they never encountered the full range before completing the BR. They had at least a 10-min break between sessions.
Data and Analysis Code Folder Structure
- /derivates:
"Behavioral_analysis.Rmd": behavioral data analyses,
"EEG_analysis.Rmd": ERP analyses.
- /data:
"allAverageDat_beh.csv": behavioral data
"allAverageDat.csv": CNV data
"allAverageDat_short.csv": CNV data for plotting
"allAverageDat_resp.csv": response-onset locked data for LPCt analysis
"allAverageDat_p2.csv": target-offset locked data for P2 analysis
"cnv_width_per.csv": CNV dissolving latency data during encoding
"cnv_width_rep.csv": CNV dissolving latency data during reproduction
- /figures: stored figures used in the paper
