The first article that I decided summarize is labeled: The Production Effect: Costs and Benefits in Free Recall. The Research report was written by Angela C. Jones of John Carroll University and Mary A. Pyc of Washington University in St. Louis. It is found in the 2014 edition of the Journal of Experimental Psychology: Learning, Memory, and Cognition.
This experiment was aimed at examining the costs and benefits of production, through use of free recall paradigm. Paradigm is defined as a typical example or pattern of doing something. Free recall is defined as the process in which participants study a list of items, and then are prompted to recall the items in any order. The production effect is the memorial benefit of reading aloud compared to reading silently. Some studies have shown the production effect as a simple memory improvement method. “The production effect is additive to the benefits of generation and semantic processing, extends to a delayed retention interval, and has been demonstrated with nonwords, word pairs, and sentences” (Jones and Pyc 300). As we can see the production effect does have certain advantages, but does it actually augment the ability of our memory? Is the effect due to increased memory for items read aloud, or is it something else? Even though at the advent of this experiment statistical tests had not been reported, Jones and Pyc hypothesized that the benefit of production was possibly instead due to a memory reduction for silent items, and thus the goal of their experiment was to prove this. What causes the production effect to alter memory ability? Jones and Pyc decided it had to do with the way in which information is organized when read silently or aloud. “The increases in recognition accuracy for items read aloud may be the result of item-specific gains associated with production, and the costs to silent items may be the result of minimal relational encoding afforded by the typical production effect paradigm” (Jones and Pyc 300).
The authors addressed this issue by splitting the study into two experiments. The goal of Experiment 1 was to discover the benefits and costs underlying the production effect. Thus, the study included one mixed list (silent and aloud items) and two pure lists (one silent, one aloud). After this the participants completed a free recall final test. The study included 48 undergraduate students from John Carroll University. First they underwent the encoding phase. The students were shown 30 items. Fifteen of the items were in blue font, and the other 15 were in red font. The words were split into two different colors because it allowed for relational processing, which increases recall when added to items that naturally elicited item-specific processing (the random non-associated words that the students were to memorize). They did this because, based on prior experiments, they were led to believe that, “the increases in recognition accuracy for items read aloud may be the result of the item-specific gains associated with production, and the costs to silent items may be the result of minimal relational encoding afforded by the typical production effect paradigm” (Jones and Pyc 300). 17 of the students were assigned to read words of one color aloud and the words in the other color silently. This group was labeled the mixed group. 16 of the students read every word silently, while the remaining 15 read all words aloud. These two groups were the pure groups. Thus, there were four variables in the experiment: silent pure, silent mixed, aloud pure, and aloud mixed. The pure list was used to allow the experimenters to assess the costs and benefits of production. After the encoding phase the students were directed to type every word that they remembered from the phase.
The results showed that there was no effect of list type, or basically that recall data was not influenced by mixed or pure list reading. Production showed greater recall from students who read aloud than those who read silently. The most notable and interesting result of the experiment was the interaction of list type and production. Production only played a benefit on the mixed list group. The most significant jump in data was between the mixed silent group (around 8% recall), and the mixed aloud group (around 24% recall). All results considered, the experimenters concluded that the production effect for the mixed list group was most likely driven predominantly by the costs to silent items. Basically, the significant variation between silent-mixed and aloud-mixed groups was less due to the benefit of reading the mixed group aloud, and more so due to the negative cost of reading the mixed group silently.
The second experiment replicated the first experiment mostly, however there was one change. Now 30 five letter words were represented, half of which were high frequency words (words that are more common in the english language), and the other half were low frequency words (words that are less common). They decided to do this because almost all previous experiments on the production effect used high frequency words, and therefore they wanted to see if the production effect extended to low-frequency words. 23 students read words from the mixed list, 23 of the students read from the pure silent list, and 23 read from the pure loud list.
The recall percent for the high frequency words correlated very closely with the results from experiment 1, as predicted. The low frequency words had higher recall percents across the board for each category, and the rise in word recall for each category was proportional to the trends in the higher frequency words. In other words, the relationship between the categories was the same, with the difference being that each category was higher in word recall in low frequency than its high frequency counterpart.
The general results of this experiment gives us good insight on the ability of memory. “We demonstrated that the production effect is not simply the result of enhanced memory for items read aloud but instead results from a cost to memory for items read silently” (Jones and Pyc 300). Both experiments reflected that the benefits of production were less than the costs of silent items. Thus, this experiment discredits the belief that the production effect is a memory tool, as memory is rather decreased by reading silent items, not increased by reading aloud.
The second article I selected is titled: Parametric Effects of Word Frequency in Memory for Mixed Frequency Lists. This research report was written by Lynn J. Lohnas and Michael J. Kahana of the University of Pennsylvania. It was published on July 8, 2013, in the Journal of Experimental Psychology: Learning, Memory, and Cognition.
An important concept to consider, as the article is built around this concept, is word frequency paradox. As defined in the abstract of the article, word frequency paradox is the finding that low frequency words are better recognized than high frequency words yet high frequency words are better recalled than low frequency words. However, based on prior experiments, this view is partially challenged, as the types of word that are recalled better can vary between high and low frequency. Thus an important question in the article is brought up. Why is item recognition consistently favorable towards low frequency words in mixed lists, but during superior recall of mixed lists there can be variations in which word frequency type is superior? Previous experiments showed instability in recall results. The authors believe that the instability is due to the substantial difference in the range of word frequencies between the high and low frequency groups. The main goal of this experiment was “to quantify the functional relation between word frequency and memory performance across the broad range of frequencies typically used in episodic memory experiments.” (Lohnas and Kahana 1).
The authors address their questions concerning relations between high and low frequencies by conducting an experiment aimed at collecting data on both recognition memory and free recall. For the free recall portion of the experiment, instead of just collecting data on results from high frequency words and low frequency words, the authors decided to use mixed frequency lists that included all the frequencies in between the high and low as well. 132 participants were used in the overall experiment. For each session of the experiment there were 16 lists of 16 words. One list containing sixteen words would be presented on a computer screen, one at a time. Each word would be accompanied by between 0 and 2 encoding tasks (these tasks included a size judgment and an animacy judgment. The number of encoding tasks changes not by each photo, but by each list. Following each list was an immediate free recall test.
The results showed that participants recalled higher proportions of both low and high frequency words than words of intermediate frequency, forming a sort of U shape. This U shape held true for both items without an encoding task, and those with an encoding task. However, when no task was presented, the recall probability for each frequency was higher by about .05 to .08.
At the end of the 16 lists presented in the session, participants would be presented with a recognition test. For half of the sessions (randomly selected) students would be given a final cumulative free recall test, in between the recall test from the 16th list and the recognition test. During this free recall test participants were asked to recall all possible items from all the lists in the section. For the recognition test, 320 words were presented one at a time on a computer screen, and participants had to select which words had showed up in the lists, and which one’s hadn’t.
The results from the recognition tests show us that with increasing word frequency, participants were more likely to incorrectly accept lures and less likely to correctly recognize targets. Thus the lower the frequency, the more likely participants were to select them in recognition tests. When no encoding tasks were presented, participants were just a little more likely to have a higher hit rate in the recognition test.
The final article that I decided to summarize is: Learning to Remember by Learning to Speak. The article was written by Marc Ettlinger of the Veterans Affairs Northern California Health Care System, Jennifer Lanter of the University of Wisconsin–Green Bay, and Craig K. Van Pay of the University of Houston. This article is found the 2014 edition of Developmental Psychology.
The goal of this experiment was to test if a child’s memory can be impacted by language. Many psychological studies regarding language had been conducted before, however none had ever had directly connected memory and language, and thus these authors were interested in digging into this topic. The authors predicted that the children’s ability to recall the plurality of different items depended on the phonology of the word, which is the sounds associated with a certain word. The authors saw it best to use three different categories of plural words. “We also considered the correlation between children’s ability to recall the plurality of sibilant-final words and their ability to articulate the plural for sibilant-final words, their recall and articulation of plosive-final words, and their recall and articulation of vowel-final words” (Ettlinger 432).
For the experiment the authors selected monolingual children that were ages 3–5 years old. In total there were 50 participants. Once they started to undergo the tests, children were show pictures of 36 objects, either shown as a singular object, or the same object four times. The child is later tested on 18 of the photos seen earlier by moving the picture he or she saw into the middle, lower box in the center of a board. If it was one of the photos with four objects, and the child selected, it means that he or she most likely understands the phonology of the name of the object selected. A certain production task, called the wug test was used to test their ability to produce the plural. In this test, the experimenter took a photo of a novel item that the child had not yet seen yet, and told him the name of the object, which was a nonce word. He then shows the child a photo of multiple units of the same object, and asks the child to tell him what it he or she is seeing in the photo, in a complete sentence. In the data collected, the researchers found an interesting correlation between plosive final words and sibilant final words. There was no connection with vowel-final words. As stated in the article, “This suggests that memory mirrors the development of plural production, where children first develop mastery of the pluralization of vowel-final words but still struggle with sibilant final words, with plosives somewhere in the middle” (Ettlinger 436). As a result of their studies, these psychologists were able to accurately prove a connection between language and memory.