The World Bank Working for a World Free of Poverty Microdata Library
  • Data Catalog
  • About
  • Collections
  • Citations
  • Terms of use
  • Login
    Login
    Home / Central Data Catalog / MICRODATA_RG / PER_2025_GAITS_V01_M
microdata_rg

Not Faster, but Better: Generative AI and Teacher Skill in Peru 2025
Baseline and Endline Surveys

Peru, 2025
Get Microdata
Reference ID
PER_2025_GAITS_v01_M
DOI
https://doi.org/10.48529/t6x6-2y70
Producer(s)
Carla Z. Glave, Carolina Lopez, Ezequiel Molina, Román Andrés Zárate
Collection(s)
Development Research Microdata
Metadata
Documentation in PDF DDI/XML JSON
Created on
Jul 20, 2026
Last modified
Jul 20, 2026
Page views
4374
Downloads
29
  • Study Description
  • Data Description
  • Documentation
  • Get Microdata
  • Data files
  • ai4t_merged

Data file: ai4t_merged

Merged data

Cases: 1269
Variables: 501

Variables

q86
If you are trying to get the certificate, what motivated you?
q82
Would you like to receive AI trainings?
q87
Who do you thinks benefits more in an AI training program?
q88
Do you think AI use will increase or decrease disparities?
q90
If you had to pay for an AI tool, how much would you be willing to pay?
male
Male
previous_training
Received previous training
has_otherjob
Has another job
add10h_ormore
Works additional 10h or more outside schedule
cellphone24
Cellphone
smartphone24
Smartphone
tablet24
Tablet
laptop24
Laptop
desktop24
Desktop
none24
No devices at home in 2024
tech_familiarity24
Familiarity with tech devices
techprep_p25_24
Least prepared
techprep_p50_24
Less prepared than the average
techprep_avg_24
Average
techprep_p75_24
More prepared than the average
techprep_p99_24
Most prepared
prepared_lessthan_avg
Tech preparedness below average
prepared_morethan_avg
Tech preparedness above average
aiuse_24_onlyheard
Heard of AI but hadn't use
aiuse_24_tried
Tried AI once or twice
aiuse_24_neverheard
Had never heard of AI
aiuse_24_frequent
Used AI frequently
aiuse_24_ocasional
Used AI occasionally
aifreq_24_always
Always
aifreq_24_freq
Frequently
aifreq_24_rarely
Rarely
aifreq_24_never
Never
aifreq_24_dk
Didn't know it
ai24
Used AI in 2024
act1_reg
Last time it took to - lesson plan
act1_wai
Time it would take with AI - lesson plan
act1_dk
Does not know time w/AI - lesson plan
act_missing_1
Did not report time - lesson plan
saved_time_1
Last time it took to minus time it would take with AI - lesson plan
act2_reg
Last time it took to - differentiated materials
act2_wai
Time it would take with AI - differentiated materials
act2_dk
Does not know time w/AI - differentiated materials
act_missing_2
Did not report time - differentiated materials
saved_time_2
Last time it took to minus time it would take with AI - differentiated materials
act3_reg
Last time it took to - designing an evaluation
act3_wai
Time it would take with AI - designing an evaluation
act3_dk
Does not know time w/AI - designing an evaluation
act_missing_3
Did not report time - designing an evaluation
saved_time_3
Last time it took to minus time it would take with AI - designing an evaluation
act4_reg
Last time it took to - grading
act4_wai
Time it would take with AI - grading
act4_dk
Does not know time w/AI - grading
act_missing_4
Did not report time - grading
saved_time_4
Last time it took to minus time it would take with AI - grading
act5_reg
Last time it took to - collaborative work
act5_wai
Time it would take with AI - collaborative work
act5_dk
Does not know time w/AI - collaborative work
act_missing_5
Did not report time - collaborative work
saved_time_5
Last time it took to minus time it would take with AI - collaborative work
act6_reg
Last time it took to - administrative work
act6_wai
Time it would take with AI - administrative work
act6_dk
Does not know time w/AI - administrative work
act_missing_6
Did not report time - administrative work
saved_time_6
Last time it took to minus time it would take with AI - administrative work
act7_reg
Last time it took to - communication with parents
act7_wai
Time it would take with AI - communication with parents
act7_dk
Does not know time w/AI - communication with parents
act_missing_7
Did not report time - communication with parents
saved_time_7
Last time it took to minus time it would take with AI - communication with paren
diff1
Difficulties in current work - Overload of admin work and planning
diff2
Difficulties in current work - Difficulty to properly evaluate learning
diff3
Difficulties in current work - Insufficient time for personalized attention
diff4
Difficulties in current work - Inefficient communication with families
diff5
Difficulties in current work - Difficulty to adapt materials to diff. styles
diff6
Difficulties in current work - Inadequate support for students with difficulties
diff7
Difficulties in current work - Challenges to stay up-to-date
diff8
Difficulties in current work - Limits to incorporate technology
diff9
Difficulties in current work - Managing large or diverse groups
diff10
Difficulties in current work - Other
aifreq_25_always
AI frequency 2025 - always
aifreq_25_freq
AI frequency 2025 - frequently
aifreq_25_rarely
AI frequency 2025 - rarely
aifreq_25_never
AI frequency 2025 - never
ai25
AI use 2025 - frequently or always
aitasks1
AI use for lesson plans
aitasks2
AI use for creating didactic materials
aitasks3
AI use for adapting materials to diff. learning levels
aitasks4
AI use for grading
aitasks5
AI use for designing evaluations
aitasks6
AI use for communication with parents
aitasks7
AI use for administrative tasks
aitasks8
AI use for information search
aitasks9
AI use for other tasks
safety_ai
How secure in using AI
safe_ai
Feels secure using AI
unsafe_ai
Feels insecure using AI
copilot_most_used
Copilot is most used AI
ai_use_2weeks
Frequency of AI use in past 2 weeks
recent_use
Used AI 6 or more times in past 2 weeks
recent_use_always
Used AI always in past 2 weeks
recent_use_none
Used AI never in past 2 weeks
learnedai_selftaught
Learned AI - selftaught
learnedai_ourtraning
Learned AI - program training
learnedai_othertraining
Learned AI - other training
learnedai_family
Learned AI - family member
learnedai_colleague
Learned AI - colleague
learnedai_videos
Learned AI - videos
ai_iteration
What do you do with AI response
modification
Use AI response as base and modify on my own
iteration
Usually iterate with AI to get a better response
ai_use_3months
Frequency of AI use past 3 months
use_3m_regular
Use AI regularly or has become indispensable past 3 months
barrera_dispositivo
Lack of proper devices
barrera_internet
Lack of proper internet
barrera_tecnologia
Hard to understand technology
barrera_colegas
Co-workers don't use AI
barrera_apoyo
Need constant support
personal_use
Frequency of personal AI use
personal_use_yes
Uses AI for personal purposes
personal_use_always
Uses AI for personal purposes - always
pertask1
Personal task with AI - writing letters or documents
pertask2
Personal task with AI - searching health information
pertask3
Personal task with AI - planning trips and activities
pertask4
Personal task with AI - help my kids with homework
pertask5
Personal task with AI - translating texts
pertask6
Personal task with AI - social networks
pertask7
Personal task with AI - solving general concerns
pertask8
Personal task with AI - entertainment
pertask9
Personal task with AI - cooking and recipes
pertask10
Personal task with AI - personal finance
pertask11
Personal task with AI - other
ai_for_teaching
Frequency of AI for teaching
ai_for_personal
Frequency of AI for personal use
personal_gt_teaching
Personal use greater than teaching use
personal_goe_teaching
Personal use equal or greater than teaching use
q72_1_num
Numeric response - AI tools can help me save time
q72_2_num
Numeric response - AI tools can help improve the quality of my classes
q72_3_num
Numeric response - AI tools can help personalize learning
q72_4_num
Numeric response - AI tools can help evaluate more effectively
q72_5_num
Numeric response - AI tools can help be more creative
q72_6_num
Numeric response - AI tools can help me go home earlier
q72_7_num
Numeric response - AI tools can help give me more time to dedicate to my student
q72_8_num
Numeric response - AI tools can help reduce stress due to work overload
q72_9_num
Numeric response - AI tools can help me have free weekends
q72_10_num
Numeric response - It is a waste of time to use AI
q72_11_num
Numeric response - The time invested in learning AI is recovered quickly
acuerdo_creatividad
Agree - AI tools can help me be more creative
acuerdo_mastemprano
Agree - AI tools can help me go home earlier
acuerdo_tiempoestudiantes
Agree - AI tools can help me give me more time to dedicate to my students
acuerdo_estres
Agree - AI tools can help reduce stress due to work overload
acuerdo_findesemana
Agree - AI tools can help me have free weekends
acuerdo_perdidatiempo
Agree - learning AI is a waste of time
acuerdo_inversiontiempo
Agree - Time invested in learning AI is recovered quickly
benefits1
AI benefits particular group - young teachers with tech expertise
benefits2
AI benefits particular group - older teachers who need more support
benefits3
AI benefits particular group - teachers with better technology at home
benefits4
AI benefits particular group - teachers motivated from the get go
benefits5
AI benefits particular group - teachers with less family responsibilites
benefits6
AI benefits particular group - teachers from schools with better infrastructure
benefits7
AI benefits everybody to the same degree
benefits8
AI benefits nobody really
benefits_particular_group
Believes AI benefits one group in particular
benefits_mostmotivated
Believes AI benefits most motivated teachers
benefits_everyone
Believes AI benefits everybody
disparities
Response to disparities - how do you think AI impacts existing disparities
increase_disparities
Believes AI will increase disparities
willingness_to_pay
Ranges of prices teachers are willing to pay for AI
wouldnt_pay
Would not pay for AI
would_pay_marketprice_ormore
Would pay marketprice or more for AI
lessonplan_useai
Would use AI to create lesson plan
feedback_useai
Would use AI to draft feedback to student
prompt_eval_incorrect
Incorrectly evaluates a prompt (gives good grade to bad prompt)
considers_goodresponse
Considers an incomplete AI response as good
considers_badcontent
Considers an incomplete AI response as bad due to content
considers_badformat
Considers an incomplete AI response as bad due to format
iteration_useai
Would use AI to improve given AI response
iteration_useai_ideal
Would use AI to improve given AI response and make adjustments after
q201
When you use AI for teaching work, how useful do you find the responses?
q202
How easy is it to use AI for teaching work?
endline_online
Responded to endline, online version
calidad_general
Overall prompt quality - training based rubric
contexto_y_rol
Context and role - training based rubric
claridad_tarea
Clarity of task - training based rubric
formato_estructura
Format and structure - training based rubric
alineacion_pedagogica
Pedagogical alignment - training based rubric
broad_calidad_general
Overall prompt quality - general rubric
broad_claridad
Clarity - general rubric
broad_especificidad
Specificity - general rubric
broad_contexto
Context - general rubric
broad_componentes
Clear components - general rubric
endline_match
Merge result from endline
endline
Responded to endline, online or in person
estudiantes_total
School level - number of students
clases_total
School level - Number of classrooms
network_size
School level - size of network
mean_mate4
School level - math in previous standardized test
mean_cl4
School level - language in previous standardized test
sh_equipos
Share of devices (% of 13 possible school devices)
anon_school_id
Anonymous school identifier
anon_ugel
Anonymous regional office code
anon_teacher_id
Anonymous teacher identifier
Total: 501
<12
Back to Catalog
The World Bank Working for a World Free of Poverty
  • IBRD IDA IFC MIGA ICSID

© The World Bank Group, All Rights Reserved.

This site uses cookies to optimize functionality and give you the best possible experience. If you continue to navigate this website beyond this page, cookies will be placed on your browser. To learn more about cookies, click here.