Episode
2026-05-07 – 2026-05-14
66 papers
Covered in this episode
Papers:
Impact of generative AI on studio-based learning in architectural education: insights from a case study in Cairo University
Asesoría en ciencias experimentales: usos de la Inteligencia Artificial por docentes de Centros PILARES en la Ciudad de México
Out-of-Field MAPEH Teaching and Grade 7 Learners’ Attitudes, Engagement, and Skill Development: Evidence from a Philippine Public Junior High School
Elementary Music Teachers’ Implementation of General Music Approaches and Perceptions of Preparation to Teach Elementary Music
+16 more
Transcript 26 lines
Cold Open
Jenny
When a tool helps you make something faster, how do you know it actually made the work better?
Davis
I think you have to look past the shiny part, because faster can mean cleaner thinking, or it can mean you just got to a bad idea sooner.
Jenny
Right, and I'm tempted to call this the AI victory lap, because the students did make more and better work, which is the headline everybody wants.
Davis
But the catch is structure, because in one Cairo University architecture study, guided generative AI use was linked to a 14 percent jump across creativity, design form, visualization, deliverable quantity, and productivity...welcome to This Week In Art Education on paperboy.fm.
Stats Overview
Jenny
This week the funnel is broad but picky: about two thousand search hits, two hundred papers on the semantic shortlist, and sixty-six qualified papers we actually analyze, from one hundred eighty-two authors across twenty-one countries.
Davis
And the qualified count is almost flat: sixty-six this week, sixty-seven last time, down about one and a half percent. With thirty-five qualitative studies and thirteen case studies, this still looks classroom-close rather than scale-chasing.
Jenny
But the search field widened a lot: two thousand thirty-five hits, up from one thousand seven hundred seventy-two, about fifteen percent. So my question is, what got noisier: keywords, venues, or the way art education is now being tagged around creativity, higher education, and sustainability?
Davis
The geography jump is the loud one: twenty-one countries, up from nine, so more than double. The top country codes are the Philippines with four, India and Saudi Arabia with three each, then Mexico, Egypt, China, and Turkey with two.
Jenny
Author mix matters too. Of the one hundred eighty-two authors, forty-six are first-time authors, meaning first-ever paper in the metadata, not just new to us; eighty-six are emerging, and fifty are experienced, so nearly half the week is early-career scholarship.
Davis
That fits the through-line. Creativity shows up six times, higher education five, sustainability four, and architectural education four, which makes art education look less like enrichment and more like infrastructure for access, professional identity, and cultural knowledge.
Paper Walkthrough
Paper 1 Impact of generative AI on studio-based learning in architectural education: insights from a case study in Cairo University
Jenny
Alright, let's get into the papers with a pretty clean AI test case: Impact of generative AI on studio-based learning in architectural education. Mai Ali Abdallah Sinousy, Z. Shafik, and Sherif Raouf Morgan looked at a senior architecture studio at Cairo University Faculty of Engineering in Egypt, where one group used generative AI in a structured way and another group worked conventionally.
Jenny
The plain version is this: the AI group made better work and more work, especially students who had been lower-performing. The authors report a fourteen percent improvement across creativity, design form, visualization, deliverable quantity, and productivity, and the important detail is stage-specific use, meaning AI was tied to moments like form generation and façade design instead of being a free-for-all shortcut.
Davis
How did they actually know the AI improved the design work, rather than just making students produce more renderings and nicer-looking boards?
Jenny
They built it in layers: a literature review, a pilot questionnaire with undergraduate architecture students across all four levels, then a supervised senior-level project comparing an AI-assisted group with a conventional group, plus participant interviews to check what the numbers seemed to show. So the evidence is stronger than a vibes-only classroom story, but it's still one institution and one studio context, which means it supports a structured approach at Cairo University rather than proving AI is automatically good for every design school.
Davis
That feels like the real takeaway for anyone teaching studio: don't just say, use the bot, and hope for brilliance. If AI has a job at a specific design stage, it can become part of the learning infrastructure; if it doesn't, it's just a faster way to hide weak thinking behind a shiny façade.
Paper 2 Asesoría en ciencias experimentales: usos de la Inteligencia Artificial por docentes de Centros PILARES en la Ciudad de México
Davis
That shiny façade problem has a cousin here, because this next paper is also asking when AI becomes help and when it becomes a shortcut. It’s called Asesoría en ciencias experimentales: usos de la Inteligencia Artificial por docentes de Centros PILARES en la Ciudad de México, and it moves us from an architecture studio in Cairo to three hundred twenty-five teachers in Mexico City’s PILARES centers.
Davis
The plain finding is that these teachers mostly liked AI because it made work faster, especially finding information and building activities for students in online high school experimental science modules. But they also warned that generative AI, meaning tools that can produce answers or materials from a prompt, could encourage unethical practices, shallow learning, and weaker critical thinking if students or teachers used it without guardrails.
Jenny
Were these teachers reporting actual classroom outcomes, or mainly their perceptions of AI tools?
Davis
Mainly perceptions and reported uses. López Carmona and her coauthors used a quantitative, descriptive-analytical digital survey, which means they counted patterns in teacher responses rather than testing whether students learned more chemistry or biology after an AI-supported lesson. The sample is substantial at three hundred twenty-five teachers, but it’s still one educational context, PILARES in Mexico City, so it tells us what these teachers see and worry about, not what AI proves across all schools.
Jenny
That makes the training point feel sharper, not weaker. If the people using the tools already see both the speed benefit and the risk of superficial learning, then AI needs structure again: not just prompts and productivity tricks, but rules for ethics, depth, and when a human teacher has to slow the whole thing down.
Paper 3 Out-of-Field MAPEH Teaching and Grade 7 Learners’ Attitudes, Engagement, and Skill Development: Evidence from a Philippine Public Junior High School
Jenny
That line about a human teacher slowing the whole thing down lands differently here, because this paper is about what happens when the teacher may not have the subject background to slow it down well. Mary Fernandez, Leny Agamao, and Priscilla Castro call it Out-of-Field MAPEH Teaching and Grade 7 Learners’ Attitudes, Engagement, and Skill Development, and they studied fifty Grade 7-UNITY learners at a public junior high school in Pangasinan, Philippines.
Jenny
The plain finding is that students felt non-MAPEH teachers hurt their learning most at the skill level, not just the vibe of the class. Academic skills got an average weighted mean of two point six nine, which the authors label High Impact, and students named very concrete losses: forty-two percent cited limited creative expression, twenty-eight percent reduced engagement, motivation, and interest, and twenty-four percent limited hands-on experiences.
Davis
Are we hearing students’ perceptions of harm, or do we have independent evidence that their skills actually declined?
Jenny
Mostly perceptions, with some written student explanations behind them. The design was descriptive correlational mixed-method, meaning the authors counted survey patterns, looked for associations, and paired that with open-ended responses, using frequency counts, percentages, weighted means, and thematic grouping rather than a before-and-after skills test. So the pattern is useful, but it’s fifty students in one school, and we shouldn’t stretch it to every MAPEH classroom in the Philippines.
Davis
Still, the practical warning is pretty sharp for the Teacher Preparation Gaps thread. If a school has to assign a non-specialist to Music, Arts, Physical Education, and Health, the emergency training can’t just be lesson planning; it has to cover demonstration, assessment, and the hands-on creative tasks where students said the damage showed up.
free_promo
Paperboy.fm
This is the free version of the podcast. Subscribe at paperboy.fm to access a dozen different paper review podcasts for five dollars a month.