of prediction and individual differences in word-pair semantic - PowerPoint PPT Presentation

An EEG investigation of the role of prediction and individual differences in word-pair semantic priming Xiao Yang Graduate Research Competition March 29, 2016

The role of prediction General question: • Do speakers make predictions about upcoming speech content?  If so, what makes one a better predictor? Linguistic phenomenon of our interest: • Word-pair semantic priming 2

Word-pair semantic priming (prime) (target) CAP hat Related KEY hat Unrelated • After encountering a prime, it is easier to activate a semantically related target, than an unrelated one. 3

Using EEG to study priming effects EEG (electroencephalogram) • records real-time voltage fluctuations in the brain • captures the brain activities triggered by a particular linguistic event (e.g. encountering a related or unrelated word) 4

EEG component of interest: N400 • N400: A negative-going wave peaking at 300-500ms • N400 priming effect: Related word pair elicits N400 reduction in amplitude (Lau et al., 2013) (Negative plotted up ↑ ) unrelated CAP hat KEY hat related 5

Mechanism of N400 priming effect What’s the priming due to? Possible mechanism 1: • The activation of the prime word passively spreads to semantically related words in the mental dictionary. • If so, then the N400 priming effect is based solely on the semantic relatedness between the prime and the target. 6

Mechanism of N400 priming effect Possible mechanism 2: • The priming is in part due to an active process of prediction; the comprehender can actively generate an expectation for specific words that are semantically related to appear. • If so, then the N400 priming effect can be conditioned by some external cues that influence the comprehender’s expectation about whether the word pair will be semantically related or not. 7

Mechanism 1: Passive association between mental lexicon KEY CAP HAT

KEY CAP cap HAT

KEY CAP cap HAT

KEY CAP cap HAT

Mechanism 2: Active prediction KEY CAP HAT

KEY CAP cap HAT

KEY CAP cap HAT

Since “cap” just appeared, I bet “hat” will appear soon KEY CAP cap HAT

Relatedness cue • Relatedness cue has been shown to modulate reaction time in behavioral studies on word-pair priming (Hutchison (2007)) • For a given trial, a relatedness cue is presented before the word pair, indicating how likely it is to encounter a related word pair (proportional cue reflects real proportions) Target (900ms) Cue (1000ms) Prime (500ms) 80% Related hat CAP 20% Related KEY hat 16

Experiment design: Conditions Related Condition Unrelated Condition Relatedness Cue Prime Target Prime Target 80% Related CAP hat KEY hat 20% Related CAP hat KEY hat • 160 prime-target word pairs • 280 filler pairs, used to establish relatedness proportion 17

Experiment design: offline measure • Does each participant exhibit the same level of sensitivity to the cue manipulation? • Previous study on older adults has shown that actively generating predictions could involve similar mechanisms to verbal abilities (DeLong et al. (2012)) • Therefore individual participant’s sensitivity to the cue could be related to their personal verbal fluency 18

Experiment design: offline measure • Individual variation in verbal fluency can be measured by an offline ‘letter and category’ task, as an index of each participant’s verbal fluency (Spreen & Strauss (1998)) Within 1 minute, Paper, tell me all the Stapler, words for office Eraser, supplies that you Sticky note, can think of. Pen, … 19

Experiment procedure • Participants:  22 KU undergraduate students (all native English speakers) • Procedure:  Offline measures (verbal fluency, attentional control, and working memory)  EEG recording while the stimuli are visually presented 20

Research questions 1. What’s the effect of Relatedness on N400 amplitude?  Related pairs will lead to N400 reduction 2. Would participants actively make prediction using the relatedness cue?  If so, ’80% related’ will likely lead to greater N400 reduction (as suggested by Lau et al. (2013 )’s results) 3. Will an individual’s verbal fluency have an effect on whether they show predictive effects?  If verbal fluency correlates with sensitivity to the relatedness cue, then it suggests some relation between verbal fluency and their sensitivity to the cue manipulation 21

Results • The main effect of Relatedness  Related word pairs elicited N400 reduction compared to unrelated pairs • The effect of Relatedness Cue  Marginally significant between 80% and 20% cue 22 All participants, 80% Related All participants, 20% Related

Results • Participants with high verbal fluency shows bigger N400 priming effect than low verbal fluency participants, when the cue is ’80% related’.  The priming is in part due to actively generating predictions by using the cue (for high verbal fluency participants) 23 Low VF participants, 80% related High VF participants, 80% related

Conclusion • Word-pair semantic priming involves actively generating predictions about the upcoming word • Individuals with higher verbal fluency were more sensitive to the relatedness cue  Verbal fluency is an individual difference that modulates comprehender’s ability to make use of the cue and generate lexical predictions  Consistent with previous N400 priming studies using verbal fluency measures among older population (DeLong et al. (2012))  Current study shows for the first time that verbal fluency show a similar correlation among younger adults 24

Thank you! Project team members: Lauren Covey, Caitlin Coughlin, María Teresa Martínez García, Adrienne Johnson, Xiao Yang, Cynthia S. Q. Siew, Travis Major, Robert Fiorentino, and Spring 2014 Neurolinguistics II class members Reference: DeLong, K. A., Groppe, D. M., Urbach, T. P., & Kutas, M. (2012) Thinking ahead or not? Natural aging and anticipation during reading. Brain & Language, 121, 226-239. Hutchison, K. A. (2007). Attentional control and the relatedness proportion effect in semantic priming. Journal of Experimental Psychology: Learning, Memory, and Cognition , 33(4), 645. Lau, E. F., Holcomb, P. J., & Kuperberg, G. R. (2013). Dissociating N400 effects of prediction from association in single-word contexts. Journal of Cognitive Neuroscience , 25(3), 484-502. Spreen O., Strauss E. (1998). A compendium of neuropsychological tests: Administration, norms, and commentary 2nd Edition . Oxford University Press; New York. 25

Results (Cont’d) Participants with high verbal fluency shows smaller N400 priming effect than low verbal fluency participants, when the cue is ’20% related’. 26 Low VF participants, 20% related High VF participants, 20% related