In the present study we explore the post-learning changes in a novel word’s definition using a cue-induced memory reactivation. Our aim was to evaluate the effect of memory reactivation on the enhancement (Study A) and on the updating (Study B) of the recently learned words. All groups performed the learning task on Day 1. In Study A, reactivated groups (React) received the reactivation phase, while control groups (CT) didn’t. Memory was evaluated with a cued-recall test (Exp A1) or a semantic task (Exp A2). In Study B, 24 h (Exp B1 and Exp B3) or 30 min (Exp B2) after learning, reactivated groups (React) received the reactivation followed by the updating phase for each word, while the control group (CT) only received the updating phase. The new information was congruent (Exp B1 and Exp B2) or incongruent (Exp B3) with the initial definition.
Important note: These are accelerated versions of the experiments as all delays are only 1 minute long (instead of the corresponding delay for each experiment).
Built with Experiment
The aim of this experiment was to evaluate the effect of memory reactivation on long-term memory for the recently acquired definitions.
Materials
Novel words. 20 Spanish words with their corresponding definitions were selected from the EsPal - Spanish Lexical Database (Duchon, Perea, Sebastián-Gallés, Martí, & Carreiras, 2013). As our aim was that participants would learn words that were totally unfamiliar to them, we chose words that had a similar and low frequency (Range = 0.01 - 0.06; word frequency per million words in EsPal corpus). These words were six to eight letters long, pronounceable and they were all nouns. The definitions were simplified in order to consist of a short sentence with no more than five words (Range = 3 - 5 words; 14 - 28 characters).
Procedure
The experiment spanned three sessions, performed on consecutive days. Participants were able to conduct the experiment with a computer, a mobile or a tablet
During the presentation phase, each new word was visually presented for three seconds before its corresponding definition appeared, both remaining on screen for three more seconds. A screen between words where participants had to press a 'Next' button was added to avoid mobile screens from blocking and to check if participants were involved in the task. Participants were instructed to pay attention and try to learn the words´ definitions. The order of presentation was randomized. The practice phase included the presentation of each word and participants were asked to type the definition (or otherwise type 'I don’t know'), with no time limit. After giving an answer, participants received feedback with the correct definition in each case. There were 3 practice blocks and the order of the 20 words was randomized in each case.
The Reactivated group received an intermediate session between the learning and testing sessions, while this phase was absent in the Control group. In this session, the reminder consisted of presenting each of the 20 words without their definitions. In each case, participants were instructed to try to remember the corresponding definition, but without writing it down. Immediately following this, they were asked to rate their subjective feeling of reactivation, that is how much they think they were able to recall the word´s definition, being able to answer 'Nothing', 'Little' or 'Very much' as a subjective measure of the reactivation strength (Van Kesteren et al., 2018).
A cued-recall test was performed 48 h after word learning to evaluate the declarative memory of the words´ definitions. Each word appeared on screen and participants were instructed to type the definition that they remembered, with no time limit and this time without feedback. This test consisted of a single block and the order of the words was randomized.
Creative Commons Attribution (CC BY)
Built with Experiment
Our aim was to establish the effect of memory reactivation on the semantic integration of the recently learned words, by analyzing how they were modulated by related meanings.
Materials
Novel words. The same words as in Experiment A1 were used.
Primes. For the semantic relatedness task, we selected three related words to act as primes for each of the 20 novel and 20 familiar words (one for each of the three blocks of the task). The 120 words were all nouns, with frequency higher than 0.5 in EsPal (Range = 0.61 - 807) and were five to eight letters long. In order to assign three unrelated primes to the 20 novel and 20 familiar words, we used the 120 previously selected primes and pseudo-randomly reassigned them to each of the target words.
Familiar words. The semantic relatedness task included 20 familiar words for comparison with the 20 novel words. These words were all nouns, with frequency higher than 0.5 in EsPal as to consider them frequent words (Range = 0.83 - 82) and were 5 to 8 letters long.
Procedure
The experiment spanned three sessions, performed on consecutive days. Participants were only able to conduct the experiment with a computer.
The same Learning procedure as in Experiment A1 was performed.
The same Reactivation phase as in Experiment A1 was performed.
Experiment A2 evaluated the integration of the new lexical items by the degree to which they can be semantically primed by related words. A semantic relatedness task (adapted from Bakker, Takashima, Van Hell, Janzen, and McQueen, 2015 and Poort & Rodd, 2019) was performed 48 h after the word learning. Participants were instructed to decide as quickly and accurately as possible whether a pair of sequentially presented words (i.e., prime-target) were related or not. Recently learned words (e.g. 'Citola', a musical instrument) were used as targets with related words (e.g. 'melody') or unrelated words (e.g. 'bottle') as primes. In addition, a set of familiar words were included as a baseline condition. A practice block with feedback of 12 prime-target pairs was followed by three blocks of 80 experimental pairs each. The order of the pairs within blocks was randomised for each participant, as was the order of the blocks. The experimental pairs included 20 novel words and 20 familiar words, each one associated with three semantically related primes and three unrelated primes (thus, no prime– target pairs were repeated). A trial started with a 1000 ms fixation screen. The prime was presented for 250 ms, followed by a blank screen for 50 ms, and the target remained for 2500 ms, where a 'yes' or 'no' response should be made. If participants didn't respond after 2000 ms passed, a warning was presented on screen indicating that response was being slow. The responses were made with a button press from the keyboard, using the 'j' key for the 'yes' responses and the 'k' key for the 'no' responses.
Creative Commons Attribution (CC BY)
Built with Experiment
The aim of this experiment was to evaluate the effect of memory reactivation on the incorporation of new information to novel words.
Materials
Novel words. The same words as in Experiment A1 were used.
New info. The new information for the words’ definitions was constructed using additional data from the original definition of each word. For some words, new information was adapted or invented in a way that it was congruent with the original definition. The new information was no longer than 8 words (Range = 3-8 words; 29-42 characters). See Supplementary Materials for the full list of novel words together with their corresponding new congruent and incongruent information.
Procedure
The experiment spanned three sessions, performed on consecutive days. Participants were only able to conduct the experiment with a computer.
The same Learning procedure as in Experiment A1 was performed.
The Reactivated group received an intermediate session between the learning and testing sessions where the reactivation and the updating phases took place. The session started with the same reactivation protocol as Study A, where a reminder consisting of the 20 words without their definitions was presented and, in each case, participants were instructed to try to remember the corresponding definition but without writing it down. Immediately following this, they were asked to rate how much they remembered the word´s definition, being able to answer 'Nothing', 'Little' or 'Very much' as a subjective measure of the reactivation strength. The updating phase for each word’s definition was performed after its reactivation. It consisted of the presentation of the new information together with its corresponding word (e.g. 'Citole' next to 'with four chords, similar to a violin'), with no time limit. This new information was then practiced once, as the word appeared again on screen and participants were asked to type the new information that they had just seen, also with no time limit. A single practice trial was included in order to avoid ceiling effects. The Control group also received this intermediate session but it only included the updating phase, thus starting the session with the presentation of the new information for each word.
Participants received a cued-recall test 48 h after the word learning to evaluate the updated definition of each word (i.e., the initial definition plus the new information). Each word appeared on screen and participants were instructed to type all the information that they remembered, including the one shown in the first and in the second session of the experiment. There was no time limit and they did not receive any feedback. This test consisted of a single block and the order of the words was randomized. .
Creative Commons Attribution (CC BY)
Built with Experiment
In this experiment we were interested in determining the time course of the reactivation effect that we obtained in the previous experiment.
Materials were the same as those used in Experiment B1.
Procedure was mostly identical to Experiment B1, but in this case reactivation was performed 30 min after the end of Session 1 (instead of 24 h), addressing whether a long-term consolidation was necessary for obtaining the boosting effect of reactivation.
Creative Commons Attribution (CC BY)
Built with Experiment
In this experiment we aimed to manipulate the congruence between the initial definitions and the new information, taking into account previous findings that demonstrated the importance of schemas in the integration of new information.
Materials were the same as those used in Experiment B1.
Procedure was mostly identical to Experiment B1, but in this case, the new information included for the updating was not congruent with the word’s definition previously learned. In order to assign the incongruent new information to the 20 novel words, we used the previously selected congruent new information and pseudo-randomly reassigned it to each of the words.
Fully open! Access by URL and searchable from the Open Materials search page
