Monitoring of memory performance using implicit and explicit measures in five to 11 year olds

• In Event: Similarities and differences of metacognitive monitoring development in the context of cognitive control and memory

Abstract

The ability to monitor one’s memory performance is a crucial skill that is said to undergo gradual development. Monitoring is a component of metacognition which, in the context of memory, can enable children to be aware of whether they have accurate memory. Monitoring of memory accuracy is argued to be present in seven-to-eight-year-olds and has improved in nine- to 11-year-olds, but others suggest that this skill is acquired earlier in development (5-6 years; Smortchkova & Shea, 2020). Our study investigated the development of monitoring for episodic memory using implicit (e.g. gestures) and explicit (e.g. confidence) measures.

74 participants aged five to 11 years completed a modified cued recall memory task. First, they watched video-clips of neutral episodic events (e.g. person making breakfast). Next, participants were asked to recall in response to memory questions which cued an item or action in the episodic events (e.g. “where did you see the fizzy drink bottle?”). Monitoring ability was measured by asking participants to rate their confidence after making a memory response using a 5-point picture cup-scale (Confidence scale). Each participant was video-recorded for gestures (behavioural & vocal) made during recall. If monitoring has developed, we expected lower confidence ratings and more uncertainty gestures for inaccurate vs accurate responses. In addition, we assessed if children could control their behaviour in accordance to their monitoring (Nelson & Narens, 1990). Participants chose to ‘show’ or ‘hide’ each response from the experimenter (Box Sorting task). If control has developed, we predicted more ‘hide’ box responses for inaccurate vs accurate responses.

The results showed that children aged five to 11 years have developed monitoring and control. Linear mixed modelling showed that Confidence (χ2(1)=237.23, p<.001; b=.66), Age (χ2(1)=33.78, p<.001; b=.51) and Box Sorting (χ2(1)=174.19, p<.001; b=.29) predicted memory accuracy in five- to 11-year-olds. A Confidence x Age interaction (χ2(1)=6.56, p=.010; b=.15; see Figure 1) suggested that nine- to 11-year-olds were better at using confidence to inform about their memory accuracy, relative to five- to six-year-olds and seven- to eight-year-olds. Preliminary results suggest that children displayed greater head tilts, fillers (e.g. “um”) and speech delays for inaccurate responses compared to accurate responses. Gesture coding is ongoing by three trained raters who are blind to the study’s purpose and will be available upon presentation. For control, no age-difference was found in Box Sorting (χ2(1)=.12, p=.727; b=.02) which suggests that children aged five to 11 years were able to control their behaviour, according to their monitored memory performance, to a similar extent.

The findings suggest that monitoring has developed in early-childhood and gradually improves with age, mirroring improvements in memory accuracy. Children demonstrated a confidence-accuracy relationship; with nine- to 11-year-olds being better at using confidence ratings to reflect their memory accuracy, relative to younger children. Our Box Sorting findings suggest that control behaviour, based on monitored information, has a relatively stable development between ages five to 11 years. Our findings have both theoretical and practical implications, which can, for example, prove useful for assisting legal decision-makers in evaluating forensic child memory reports.

Authors

• Rumandeep Kaur Hayre, University of Birmingham
  Presenting Author

• Madeleine Ingham, University of Birmingham

• Shona Smith, University of Birmingham

• Brooke Findel, University of Birmingham

• Chloe Sargent, University of Birmingham

• Melissa Colloff, University of Birmingham