Two sides of the same coin? Comparing structural priming between production and comprehension in choice data and in reaction times

ABSTRACT

Although structural priming seems to rely on the same mechanisms in production and comprehension, effects are not always consistent between modalities. Methodological differences often result in different data types, namely choice data in production and reaction time data in comprehension. In a structural priming experiment with English ditransitives, we collected choice data and reaction time data in both modalities. The choice data showed priming of the DO and PO dative. The reaction times revealed priming of the PO dative. In production, PO targets were chosen faster after a PO prime than after a baseline prime. In comprehension, DO targets were read slower after a PO prime than after a baseline prime. This result can be explained from competition between alternatives during structure selection. Priming leads to facilitation of the primed structure or inhibition of the opposite structure depending on the relative frequency of structures, which may differ across modalities.

KEYWORDS:

• Structural priming
• maze task paradigm
• language production
• language comprehension

Disclosure statement

No potential conflict of interest was reported by the author(s).

Materials, data, and analyses are available online

https://osf.io/c6uyz/?view_only=62c937654b1b4133ba20619652ede266

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Notes

1 It may be possible that this choice between two plausible continuations of the sentence draws the attention of participants and reveals the purpose of the experiment. Therefore only half of the participants were assigned to the free-choice version, and we inserted the version (no-choice vs. free-choice) as a fixed effect in our analyses. In none of the analyses there was an interaction between the priming effect and the version. In addition, after the experiment participants were questioned on their thoughts on the purpose of the experiment. The majority of participants thought the experiment was on lexical effects, such as measuring their vocabulary size. Therefore it seems that the choice was not very obvious for participants.

2 If structural priming between production and comprehension is the same, we may expect a similar effect size across modalities. For production, Mahowald et al. (2016) assume a true effect size of .51 (log-odds), and an interaction with lexical overlap with a coefficient of 1. However, we did not a priori exclude the possibility that the effect size is smaller in comprehension. The reported effect size of priming relative to a neutral baseline condition in Traxler (2008) was d = 0.44 in the verb-overlap condition and d = 0.49 in the condition without lexical overlap (for first-pass regressions in an eye-tracking experiment). Note that Traxler did not observe a significant lexical boost effect, and the effect sizes in for example Arai et al. (2007) were lower (reporting a partial η²= 0.20 in the same-verb condition, and partial η²= 0.0015 in the different-verb condition). For the comprehension experiment, we therefore based ourselves on the more conservative recommendations provided in the Supplementary Materials of Mahowald et al. with an assumed true effect size of .34 and a coefficient of .5 for the interaction with lexical overlap. The estimated power to detect abstract structural priming with 384 participants and 24 critical items (excluding the 12 baseline items) is then between 87% and 94%, and the power to detect an interaction with lexical overlap is between 85% and 88%.

3 The tested group has a relatively wide age range, which may have led to more variability in the data than in other experiments that are otherwise comparable. We added random intercepts for participants in our statistical models to minimize the impact of interspeaker variability on our statistical results.

4 The estimated power for the production experiment to detect an interaction between the prime structure and the verb overlap condition is, based on the meta-analysis by Mahowald et al. (2016), over 97% (192 participants with 24 critical items, excluding the 12 baseline items).

Additional information

Funding

This work was supported by Fonds Wetenschappelijk Onderzoek [grant number G027319N].