Graded and ungraded expectations: Examining prediction dynamics during active comprehension

Lai, Melinh K.

Permalink

https://hdl.handle.net/2142/121520 Copy

Description

Title

Graded and ungraded expectations: Examining prediction dynamics during active comprehension

Author(s)

Lai, Melinh K.

Issue Date

2023-07-13

Director of Research (if dissertation) or Advisor (if thesis)

Federmeier, Kara D.

Doctoral Committee Chair(s)

Federmeier, Kara D.

Committee Member(s)

• Benjamin, Aaron S.
• Dell, Gary S.
• Fisher, Cynthia L.
• Payne, Brennan R.

Department of Study

Psychology

Discipline

Psychology

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• prediction
• language comprehension
• ERPs
• N400
• anterior positivity
• frontal negativity

Abstract

Language comprehension can be facilitated by the accurate prediction of upcoming words, but the precise mechanism of prediction is still largely not understood. A key aspect of prediction and its role within broader comprehension is that prediction is not ubiquitous, and multiple studies demonstrate that it fluctuates with several factors, such as literacy or age. Understanding the circumstances in which prediction can be reduced or promoted is therefore a critical step in uncovering its underlying mechanism and how it ultimately fits within the larger complex of comprehension. This thesis investigates these circumstances in four experiments that use event-related potentials (ERPs) to analyze participants’ electrical brain activity during sentence comprehension. In Experiment 1, we attempt to promote prediction within participants by simply instructing them to do so. ERPs from an active prediction task were compared to a prior block where participants simply read similar items for comprehension. The results suggest that prediction processes are largely stable, but other message-building processes were enhanced. The rest of this thesis continues investigating the prediction mechanism while additionally probing whether predictions are broadly semantic or more lexically focused. In Experiment 2, participants read similar sentences for comprehension with the addition that some sentences ended with words that were synonymous with their predictions. Replicating prior work, the addition of these related endings reduces some prediction patterns while enhancing other comprehension processes. Experiment 3 promotes engagement of prediction processes with a lexical decision task which implicitly promoted prediction as a strategy for more efficient word identification. In Experiment 4, we attempt to explicitly promote prediction by presenting the sentences of Experiment 2 within the paradigm of Experiment 1 and find that message-building processes are once again prioritized without modulating prediction signals.

Graduation Semester

2023-08

Type of Resource

Thesis

Copyright and License Information

Copyright 2023 Melinh Lai

Owning Collections