On Under-Socialization

Hallie Williams, CEI Research Assistant

It’s 11 AM, but your energy levels already feel like it’s 3 PM. Your class cannot seem to quiet down no matter what you try— these are not the 2nd graders you’ve taught in the past. 

You are not alone. Recent reports reveal that post-pandemic educators of young children designate their students as behaving below grade level norms both academically and emotionally (Contreras, Intagilata, & Chang, 2023; DeArmond et al., 2022; Mann et al., 2021). If an educator usually gives 20 minutes of academic instruction, but their current students can only handle about 12 — what happened?  

The aftereffects of the pandemic’s lockdown have significantly impacted young children, and we need to reevaluate our expectations of their academic engagement. Otherwise, if left unchecked, the pressure to “return to normal” may deny young learners the opportunity to recuperate learning and may exacerbate exisiting academic inequities. 

Read more.

AI and Education – Promise or Peril?

Meghan Wenzel, CEI Researcher & Co-Editor

We often look to technology to solve big problems. Sometimes it pans out, while others we are arguably left worse off. Artificial Intelligence (AI) has received significant attention lately, particularly with the buzzy release of ChatGPT-4, an AI-powered chatbot that captured people’s attention and curiosity around the world. It became the fastest-growing app in history, boasting 100 million users just two months after its launch (Dickson, 2023).

What is ChatGPT?

ChatGPT is powered by a deep neural network, which is a software architecture inspired by the human brain. Deep neural networks learn to perform tasks by processing vast datasets and tuning their parameters to approximate the common patterns they spot across those examples.

The model takes a block of text and tries to predict the next item in the sequence. It then compares its output with the actual text in the training set, and then “learns” from its mistakes. By doing this at a huge scale, it can develop a language model that can create coherent text when given a prompt (Dickson, 2023).

ChatGPT, like all other AI models, are trained based on historical data available online. However, historical data has inherent flaws – “bias, inaccuracies, misinformation, delusions, limited representation, and more” (Dance, 2023). Because it’s trained on flawed data, it continues to perpetuate and spread bias and misinformation if left unchecked. Thoughtful regulations as well as an ethical and inclusive lens will be crucial as we develop and use AI more. 

Read more.

Curiosity – Motivation, Exploration, and Learning

Hannah Puttre, CEI Research Assistant, Alison Sumski, CEI Research Assistant, and Christine Mason, CEI Executive Director

Curiosity is a marvel of early childhood. As children grow up, they are constantly asking questions and exploring the world around them. This natural, often inherent, curiosity helps children learn (e.g., Shah et al., 2018) and is viewed as an integral strategy in early education to engage children’s natural wonder. 

What exactly is curiosity? Is it an individual trait? A drive similar to hunger that motivates a person to learn? Or perhaps, a desire to gain missing information? Perhaps all of these. Curiosity also helps to prepare our brains for learning as well as making learning more rewarding (Stenger, 2014). A particularly useful way to think about curiosity in early learning settings is the information-gap theory (Lowenstein, 1994). Lowenstein’s information-gap theory suggests that curiosity arises from realizing there is something you do not know.

Read more.