Trust, Readiness & AI Competence in Education
Teacher Readiness for AI: Competence, Trust and Concerns
An interactive look at how K-12 teachers' AI teaching competence, trust in AI and concerns about AI compared before and after a two-day professional development program. Choose a question, examine the evidence, then build your own version through vibe coding.
Keep exploring
How to use this page
Each one answers with real study results, as group summaries.
Browse questions ↑Copy the ready prompt, download practice data, and build your own version in Google AI Studio (free).
Open Google AI Studio ↗Save first and improved prompts in the Prompt Lab and see what changed.
Open Prompt Lab → 4Ask TRACEAsk questions in plain English, answered only from the study's summaries.
Open Ask TRACE →Findings
Key findings
Higher skills and trust, lasting concerns.
Teachers moved toward confident adoption
Group averages with 95% confidence intervals. Hollow = before, filled = after.
Skill and worry are different things
Correlation runs from −1 to 1; 0 means unrelated. Hollow = before, filled = after.
Some concerns eased; others did not
% of teachers agreeing with each concern. Hollow = before, filled = after.
How we know the survey works
The researchers used structural equation modelling (SEM), specifically confirmatory factor analysis, to test whether teachers' answers match the survey's design.
What is SEM?
A way to check that groups of survey statements really measure the skill or belief they were written for, like checking that all the "fractions" questions on a test behave like a fractions test.
What is model fit?
How closely teachers' real answers match the survey's design. CFI and TLI near 1 are good; RMSEA and SRMR near 0 are good (under .08 acceptable).
What is a factor loading?
How strongly one statement reflects its skill or belief, from 0 to 1. Above .70 means the statement is a strong signal.
Why remove statements?
Statements that repeat another, overlap with a different skill, or don't fit the overall pattern make a survey longer without making it better. 21 of 52 were removed.
Concerns form their own factor
What earlier research says
Notes from the study's literature review. The study builds on these ideas and tests them together for U.S. K-12 teachers.
References
What this means
For PD designers
Pair hands-on practice with time for sensemaking about transparency, autonomy and privacy, and measure trust and concerns as well as skills.
For school leaders
Clear policies on data use and human oversight help turn new skills into confident, responsible practice.
For EdTech builders
Explain how decisions are made, keep teachers in control and be open about data. Teachers trust tools they can understand.
Before and after are different groups of teachers and there was no comparison group, so differences cannot be attributed to the workshop alone.
Prompt Lab
Your prompt journal
Keep your first prompt, write an improved one, and note what changed. Entries are saved only in this browser.
Choose an entry or start a new one. Tip: in Explore, use "Vibe-code this question", then "Save to Prompt Lab".
Ask TRACE
Ask the data in plain English
Chat about the study's findings, no sign-in needed. Attach a document or spreadsheet to ask how it relates.
About
About the study and this dashboard
Read with care
Privacy
Downloads
Synthetic practice data · Dimensions · Subscales · Items · Groups
Build your own Ask TRACE app
Open Google AI Studio → Build, attach the synthetic data and the item summary,
and paste the prompt from the TRACE Studio Kit (ask_trace_app_kit.md). AI Studio builds a chat app you can share with colleagues.