Pictures over words : a cross-sectional study reporting short term memory abilities in children [ version 1 ; peer review : 1 approved with reservations , 1 not approved ]

Background: An impressive amount of research has been conducted studying modality-specific Short Term Memory (STM) skills in children differing in the type of tasks used. In the present study, we aimed to profile the STM abilities based on modality, in typically developing children using a story-based assessment. Methods: The study followed a cross-sectional design and comprised of 80 participants between the ages of 6 years to 9 years 11 months. An animated story was shown to the children, following which a word recall task was performed. In this task, children were asked to recall the words mentioned in the story from a pictorial array. Results: One-way analysis of variance revealed a significant difference in the overall recall abilities of children. The recall performance was strongly related to the modality of the presentation of words. A marginal difference was observed for the recall of auditory-visual words in comparison to recall of words in the auditory modality; wherein older children recalled better in comparison to younger children. The findings of the study could be attributed to the "visual superiority effect", "encoding specificity principle of memory" and "multimedia effect." Conclusion: STM abilities were observed to increase with age, with the existence of asynchrony in the auditory-visual and auditory recall scores indicating the firm reliance on the modality of presentation of word. The study implications emphasize on the use of visual stimuli for teaching new vocabularies, skills, and concepts in younger children. These findings also highlight the use of visual stimuli while assessing speech, language, and cognitive skills in younger children.


Invited Reviewers
19 May 2020 report report

Introduction
Short term memory (STM) skills are dependent on the modality of presentation of stimuli. However, the literature on modality has shown contradictory results. Pillai & Yathiraj (2017) reported that auditory recall is superior when compared to visual recall in children aged 7-and 8-years old. On the contrary, Vuontela et al. (2003) observed that visual memory reaches functional maturity earlier than the auditory memory system in children between 6 and 13 years. These studies differed in the type of task used, and neither were context-based nor curriculum-oriented. Academics make children "context-bound" by exposing them to a rich curriculum via textbooks and curricular activities (Walker & Lombrozo, 2017). Given these observations, using stories would be both curriculum-oriented and context-bound in assessing STM skills. The present study aimed to profile modality-specific STM skills in children between the age range of 6 years to 9 years 11 months using a story-based assessment.

Study design
The study followed a cross-sectional design that was approved by the Institutions Ethics Committee (Ethical Reference Number -IEC KMC MLR 11-18/463). The study was conducted between December 2018 and January 2020. The study was conducted within the classroom setting of the approached school. Written informed consent was procured from parents of children who agreed to take part in the study.

Participants
English medium schools affiliated to the Central Board of Secondary Education (CBSE), Mangalore city, were approached, after obtaining authorization from the Block Education officer, to recruit participants for the study. Those schools that provided permission to conduct the study were considered for data collection. Typically developing children who passed the WHO Ten-Question Disability Screening Checklist (Singhi et al., 2007) were recruited for the study. Children with a history of any transfer from more than one school; a history of any shift in the medium of instruction; or a history of academic failures were excluded from the study.

Data collection
A story, "The Wooden Box" (copyright © 2019, Anil and Bhat), was constructed as animated stimuli and a "Word Recall" task (Extended data (Rooha et al., 2020)) was formulated based on the story to assess STM skills. The final modified task included 12 pictorially represented words that had an equal number of words from the story (Gold coins, Cupboard, Keys, Traffic), words thematically related (Hut, Chair, Diamond, Bag), and words unrelated to the story (Apple, Chair, Frock, Flower). Among the four words from the story, two of the words were presented in the auditory-visual modality (Gold coins, Cupboard), while the other two words were presented in auditory modality alone (Traffic, Keys). The task was to identify the pictures representing the words from the story. A score of one was given for each word recall.
The animated story as well as the formulated task was content validated by three speech-language pathologists and three primary school teachers of CBSE. The suggestions provided included using Indian names for the characters of the story, to modify the instructions and to modify certain words in the formulated task. These suggestions were incorporated in the preparation of the final stimuli.
Each child was evaluated individually in a classroom. The animated story was presented on a laptop screen, followed by the administration of the word recall task. The examiner presented the task verbally, and the child's responses were scored simultaneously.

Data analysis
The responses were subjected to statistical analysis using SPSS software version 16.0 and significance was set at the 0.05 level (p<0.05). Descriptive statistics were used to obtain the mean and standard deviation of the data. A frequency measure was done to analyse the percentage of children recalling each of the words from each of the groups. One-way analysis of variance (ANOVA) test was used to test the difference in recall performance across the age groups. Further, the Bonferroni post-hoc test was done to assess the pair-wise differences in performance between the groups.

Results
The results of the descriptive statistics revealed a steady increase in performance across the age groups ( Figure 1). Group IV obtained the highest scores, followed by Group III, Group II, and Group I.
The results of one-way ANOVA revealed a statistically significant difference in the Word Recall task [F(3,76) = 8.387, P=0.000] across the groups. Bonferroni Post-hoc results revealed that only Group I differed significantly from other groups, as depicted in Table 1. Frequency measure of children who recalled each of the words across the four age groups is depicted in Figure 2, which reveals that words 'Gold coins' and 'Cupboard' were recalled by almost all children. However, recall of words 'Key' and 'Traffic' increased with age, with drastic changes in the recall of word 'Traffic'.

Discussion
The results revealed that the recall ability increases significantly with age. The findings agree with the study done by Belacchi et al. (2017), where they concluded that the word recall scores increased significantly from 6 to 12 years. The study also observed a difference in recall of the words 'Gold coins' and 'Cupboard' in comparison to the words 'Keys' and 'Traffic'. The differences in the recall can be attrib uted to the inherent characteristic of the words, as the words 'Gold coins' and 'Cupboard' were presented in auditory-visual modalities in the story. In contrast, the words 'Key' and 'Traffic' were presented in auditory modality alone. Thus, it can be Table 1. Post hoc results of word recall task in primary schoolaged children. Group I -6-6.11 years, Group II -7-7.11 years, Group III -8-8.11 years, and Group IV -9-9.11 years. I -III  I -IV  II-III  II -IV III -IV p -value 0.039* 0.004* 0.000* 1.000 0.271 1.000 *Significance at the 0.05 level Figure 2. Percentage of primary school-aged children who recalled each of the words correctly in a word recall task. Group I -6-6.11 years, Group II -7-7.11 years, Group III -8-8.11 years, and Group IV -9-9.11 years. observed that auditory-visual recall was superior to auditory recall in these children. The findings of the present study can be attributed to various reasons which are discussed ahead.

Groups I -II
Visual superiority effect Younger children are fascinated by the illustrations of the story, and focus more on visual animations in comparison to the auditory narration of the story. Attention thereby forms a critical prerequisite to encode, store, and subsequently recall information. Reduced attention to auditory information may have contributed to poorer recall of words presented in the auditory modality. Hayes & Birnbaum (1980) observed similar behaviour and termed it as the "visual superiority effect," i.e., younger children are more inclined to "look and not listen." Further, there exists a difference in processing the two types of sensory information. For the items to be stored in the STM, the brain has to cognitively create 'mental images' of these items, which are pictorial representations of words inside one's mind. When processing visual stimuli, the brain functions to discover a 'mental image,' but when processing auditory stimuli, the brain has to create a mental image of the heard word for correct recall (Hilton, 2001). These brain functions are mediated by higher cognitive skills, which develop only with age. This could have contributed to better recall of auditory-visual words when compared to auditory words.
Encoding specificity principle of memory This principle states that recall of memory is optimal when the retrieval conditions replicate the conditions present when memory was created (Tulving & Thomson, 1973). In the current study, recall of auditory-visual memory was superior because the retrieval condition duplicated the conditions when the memory was formed, i.e. children had to identify the same pictures as seen in the story. However, the retrieval of auditory stimuli did not duplicate the conditions when the memory was formed, i.e. children had to identify pictures of words heard in the story, which could have contributed to poorer performance.

Multimedia effect
It is claimed that presenting multimedia information, i.e. presentations of material using words and pictures (Mayer, 2002) results in deeper comprehension (Boerma et al., 2016), subsequently improving recall. This can be considered as a contributing factor for better recall of auditory-visual words.
Lastly, Ferrara et al. (2017) reported that detailed visual memory capacity is present in children as young as six years of age, as a result of faster maturation of visual memory than the auditory memory. This can be considered as a contributing factor for observing no differences in auditory-visual recall performance across the age group.
These evidences supports the findings of the present study; an increase in STM skills with age, the existence of asynchrony in the auditory-visual and auditory recall scores, and recall performance strongly relating to the modality of presentation of information. These findings provide implications for the use of visual stimuli while teaching new vocabularies, skills, and concepts in younger children. These findings also highlight the use of visual stimuli while assessing speech, language, and cognitive skills in younger children as it will serve as a framework for maintaining their attention while evaluating various communicative skills.

INTRODUCTION:
The major issue with this article is that the background is poorly developed. There is abundant evidence in the literature on the effect of visual, auditory or cross-modal effects in STM

INTRODUCTION:
The paper was designed considering a word limit of 1000 words, hence an extensive literature review was not included in the Introduction. However, the authors have built the need by citing relevant literature to peruse the study in the introduction.

DISCUSSION:
Also, while the authors associate the findings to the aging effect, it could well be that the effects are merely the effect of learning since children become more experienced learners with age.

DISCUSSION:
While the 'aging effect' and 'more learning experience with age' can be associated with memory performance, the authors have discussed majorly on the difference in recall between the modalities. The discussion revolves around specific concepts of modality difference.

DISCUSSION:
The findings are not discussed adequately in the context of the existing literature. I suggest addressing these issues before the article is accepted for indexing.

DISCUSSION:
The findings are related to the concepts of existing literature. The review of the existing literature has not been added to the manuscript as the paper was designed considering a word limit of 1000 words. However, we have tried to discuss in light of major contributing studies.
○ how long each of the stimuli appeared on the screen. How long were the participants allowed to wait before providing a response? Was there a penalty if the participants waited for an extended period of time before they provided a response?

METHOD:
As the study focused only on the accuracy of the recall and not the reaction time measures, time-related aspects were not considered in the methodology section.

DISCUSSION:
The authors make a case that words such as "gold coins" and "cupboard" were recalled by more number of participants. I wonder is it because the participants are exposed to these stimuli items more than the other items that were presented to them? I wish the authors had performed a familiarity rating on each of the stimuli items. Currently, it is unknown what played a role in better identification of specific words over others? This is a major confound of the current study,

DISCUSSION:
The story included words that are familiar to children, ascertained through a textbook analysis of vocabulary.
○ Content validation was also done by primary school English teachers who approved the familiarity of the words considered in the story by all students.

○
The words considered were all nouns, having only a single frequency of occurrence in the whole story which was also content validated by primary school English teachers and Speech-Language Pathologist. ○ Competing Interests: No competing interests were disclosed.