researcher-made questionnaire, and analysis will be
done using SPSS for descriptive statistics and
“Adoption of Artificial Intelligence in Lesson SmartPLS 4 for structural equation modeling.
Planning: Modeling
Teachers’ Behavioral Intention to Use AI Tools.” 6. What’s your expected outcome?
I expect to find that perceived usefulness and ease of
POSSIBLE QUESTIONS use significantly influence behavioral intention and
intention to use AI. I also expect behavioral intention
2. Why did you choose this topic? to be a strong predictor of actual intention. Overall,
the study should show that preservice teachers are
I chose this topic because lesson planning is at the
motivated to adopt AI, but their readiness depends on
heart of teaching, and AI is now becoming a powerful
support systems such as training, confidence-building,
tool that can support teachers in designing,
and institutional encouragement.
organizing, and enhancing instruction. As a future
educator, I want to understand whether preservice 7. How does it contribute to existing knowledge?
teachers are ready to adopt AI, what motivates them,
and what challenges they face. This is timely because This study contributes by applying the Technology
AI adoption is a growing trend in education, and it Acceptance Model (TAM) in the Philippine teacher
directly connects to our preparation as 21st-century education context. While TAM has been widely
teachers. studied internationally, there is limited research on
Filipino preservice teachers’ adoption of AI in lesson
3. What’s the problem statement? planning. By situating TAM within constructivist
pedagogy and local realities, the study provides new
The problem addressed in this study is the lack of
insights that can guide curriculum design, training
understanding of how preservice teachers perceive
programs, and institutional policies.
and intend to use AI in lesson planning. While AI tools
are available and increasingly promoted, we don’t yet 8. How will you collect data?
know how perceptions of usefulness and ease of use
translate into behavioral intention and actual I’ll collect data using a validated questionnaire
intention to adopt these tools. Without this adapted from TAM-based instruments. The survey
knowledge, institutions may struggle to support will measure perceived usefulness, ease of use,
meaningful AI integration in teacher education. behavioral intention, and intention to use AI, all on a
5-point Likert scale. Before distribution, I’ll seek
4. What are your research objectives? permission from the dean and department heads,
then administer the questionnaire to randomly
To describe the demographic profile of
selected fourth-year BEED and BTLED students.
respondents.
To measure preservice teachers’ perceptions 9. What’s your sample size/population?
of AI in lesson planning based on TAM
constructs: perceived usefulness, perceived The study will involve 95 fourth-year BEED and BTLED
ease of use, behavioral intention, and students enrolled at Southern Leyte State University –
intention to use. Main Campus during the academic year 2025–2026.
To validate the measurement instruments for Respondents will be chosen through simple random
reliability and validity. sampling to ensure fairness and representation.
To analyze the relationships among these
constructs using PLS-SEM. 10. What are the limitations?
5. What’s your methodology? The study is limited to BEED and BTLED students at
SLSU – Main Campus, so results cannot be
I’m using a quantitative, descriptive-correlational generalized to other institutions. It also does not
design. The descriptive part will capture how account for external factors such as faculty influence,
preservice teachers perceive AI, while the infrastructure, or long-term adoption patterns. The
correlational part will test the relationships among focus is strictly on perceptions and intentions within
the constructs. Data will be collected through a the identified population.
11. How will you analyze data? that can benefit the university’s teacher education
programs.
I’ll use descriptive statistics (frequency, percentage,
mean, weighted mean) to describe demographics and 17. Why use the Technology Acceptance Model
perceptions. Then, I’ll apply SmartPLS 4 to test the (TAM)?
measurement model (reliability and validity) and the
structural model (relationships among constructs). I used TAM because it is a well-established framework
This includes checking collinearity (VIF), explanatory for studying technology adoption. Its constructs—
power (R²), effect sizes (f²), and path coefficients with perceived usefulness, ease of use, behavioral
bootstrapping for significance. intention, and intention to use—fit perfectly with my
research focus on AI in lesson planning. TAM also
12. Is this aligned with your field’s current trends? allows me to test relationships among these variables
in a structured way.
Yes. AI adoption in education is one of the most
important global trends, and preservice teacher 18. How is your study different from previous
readiness is a critical issue in the Philippines. This research?
study aligns with current efforts to promote digital
literacy, constructivist pedagogy, and technology Most previous studies on AI adoption were conducted
acceptance in teacher education. in other countries or with in-service teachers. My
study is unique because it applies TAM to preservice
13. Have you consulted relevant literature? teachers in the Philippine context, specifically
focusing on lesson planning. This localized perspective
Yes. I reviewed both local and international studies adds new knowledge to the global discussion.
from 2021 to 2025, including works on TAM, AI in
lesson planning, preservice teacher perceptions, and 19. What instrument will you use, and how will you
institutional factors. This ensures my study is ensure it is valid?
grounded in current scholarly discussions and builds
on existing knowledge. I will use a validated adapted questionnaire based on
TAM instruments from prior studies. To ensure
14. Timeline for completion? validity, I will have experts review the items for
content relevance, and I will test reliability using
Proposal Defense & Revision: December 2025 Cronbach’s alpha and composite reliability.
– January 2026
Instrument Validation & Data Collection: 20. How will you ensure ethical considerations in
January 2026 your study?
Data Analysis: February 2026
Final Write-up & Defense: February 2026 I will seek permission from the dean and department
heads before data collection. Respondents will be
15. Why focus on preservice teachers instead of in- informed that participation is voluntary, and their
service teachers? responses will be kept confidential. No personal
identifiers will be disclosed in the results.
I focused on preservice teachers because they
represent the future of the teaching profession. Their 21. What challenges do you anticipate in conducting
perceptions and readiness to adopt AI will shape how this study?
classrooms evolve in the coming years. Studying them
now helps institutions prepare training and support Possible challenges include limited availability of
systems before they enter the workforce. respondents, hesitation to answer honestly, or
technical issues in data analysis. To address these, I
16. Why did you choose Southern Leyte State will schedule data collection carefully, assure
University as your setting? respondents of confidentiality, and seek guidance
from my adviser when analyzing data.
I chose SLSU because it is my academic institution,
and I have direct access to the population of BEED 22. How will your findings be useful to teacher
and BTLED students. This makes the study feasible education programs?
and relevant, while also providing localized insights
My findings can guide curriculum designers and I will use a validated adapted questionnaire from
institutions in embedding AI literacy into teacher previous TAM studies. Reliability will be tested using
education. By knowing how preservice teachers Cronbach’s alpha and composite reliability, ensuring
perceive AI, programs can design training modules that the items consistently measure the same
that build confidence, address concerns, and promote constructs.
responsible use of AI in lesson planning.
30. How will you ensure the validity of your
23. What statistical tools will you use aside from PLS- instrument?
SEM?
Validity will be checked through expert review for
I will use SPSS for descriptive statistics such as content relevance and through statistical tests like
frequency, percentage, mean, and weighted mean. Average Variance Extracted (AVE) and discriminant
These tools will help describe the demographic profile validity in SmartPLS. This ensures the items truly
and summarize perceptions before testing represent the constructs.
relationships in SmartPLS.
31. What statistical software will you use and why?
25. What practical recommendations will you give
after the study? I’ll use SPSS for descriptive statistics because it’s
straightforward for summarizing data, and SmartPLS 4
I plan to recommend integrating AI training into for PLS-SEM because it’s powerful for testing complex
teacher education, providing institutional support for relationships among latent variables.
AI adoption, and encouraging reflective use of AI
tools. These recommendations will help preservice 32. Why did you choose PLS-SEM instead of other
teachers use AI responsibly and effectively in lesson statistical methods?
planning.
PLS-SEM is suitable for exploratory studies and for
26. How did you come up with your title? models with multiple constructs like TAM. It allows
me to test reliability, validity, and causal relationships
I crafted the title to highlight both the focus on lesson even with a relatively small sample size, which fits my
planning and the modeling of teachers’ behavioral study.
intention. It reflects the main framework I’m using,
the Technology Acceptance Model, and makes clear 33. What is the significance of using TAM in your
that the study is about AI adoption in education. study?
27. Why focus on lesson planning instead of other TAM provides a clear framework to understand how
teaching tasks? perceptions of usefulness and ease of use influence
behavioral intention and actual use. It’s widely
Lesson planning is the foundation of teaching. If recognized, so applying it in the Philippine teacher
teachers can adopt AI effectively in planning, it will education context strengthens the credibility of my
naturally influence other areas like instruction, study.
assessment, and classroom management. Starting
with lesson planning makes the study more practical 34. How will you interpret your SEM results?
and relevant.
I’ll interpret them by looking at path coefficients,
28. What makes AI different from other educational effect sizes, and R² values. Significant positive
technologies? relationships will confirm that perceptions influence
behavioral intention and actual use. This will show
Unlike traditional tools, AI can generate, organize, which factors are most important for AI adoption.
and adapt content in real time. It doesn’t just store
information—it can assist teachers in brainstorming, 35. What practical benefits will your study give to
sequencing lessons, and even suggesting students?
differentiated activities. That makes it more dynamic
compared to static tools.
29. How will you ensure the reliability of your
instrument?
Even though students are not the direct respondents, age distribution. These tools don’t manipulate the
they benefit indirectly. If preservice teachers learn to data—they just present it in a way that makes the
use AI in lesson planning, students will experience profile of the respondents easy to understand and
clearer, more engaging, and more personalized interpret.
lessons.
42. Why use weighted mean for perceptions?
36. What practical benefits will your study give to
institutions? Weighted mean is the most appropriate tool for
Likert-scale data because it considers both the
Institutions can use the findings to design training frequency of responses and the weight assigned to
programs, policies, and support systems that each option (from Strongly Disagree to Strongly
encourage responsible AI adoption. This helps them Agree). Instead of just counting responses, it gives a
prepare future teachers for modern classrooms. balanced picture of how strongly respondents agree
or disagree with each statement. This allows me to
37. How will you handle ethical concerns about AI interpret whether preservice teachers find AI “very
use in education? effective,” “moderately effective,” or “less effective”
in lesson planning.
I will emphasize that AI should be used as a
supportive tool, not a replacement for teacher 43. How will you interpret Cronbach’s alpha results?
judgment. My study also highlights the importance of
critical evaluation, ethical awareness, and institutional Cronbach’s alpha measures how consistently the
guidance in AI adoption. items in a questionnaire reflect the same concept. For
example, if I have five items measuring perceived
38. What gap in the literature does your study usefulness, Cronbach’s alpha tells me whether those
address? items are reliable together. A value of 0.70 or higher
means the items are consistent and reliable. If the
The gap is the lack of localized research on preservice
value is lower, it suggests that some items may not fit
teachers’ adoption of AI in lesson planning in the
well, and I may need to revise or remove them. This
Philippines. Most studies are international or focus on
ensures my instrument is trustworthy.
in-service teachers, so my study fills that gap.
44. Why use composite reliability in addition to
39. How will you ensure your respondents answer
Cronbach’s alpha?
honestly?
Composite reliability is more precise because it
I will assure them of confidentiality and anonymity. I’ll
doesn’t assume all items contribute equally—it
also explain that their honest responses will help
considers the actual strength of each item’s loading.
improve teacher education programs, which
In PLS-SEM, this is important because some items
encourages sincerity.
may be stronger indicators than others. By using
40. What will you do if your findings don’t support composite reliability, I can confirm that the construct
your hypotheses? is measured accurately, even if some items are
weaker. Values between 0.70 and 0.90 are considered
If the results don’t support the hypotheses, I’ll still good, showing that the construct is reliable.
report them honestly. Unexpected findings can reveal
new insights, and they may suggest that other factors 45. How will you check convergent validity?
beyond TAM influence AI adoption.
Convergent validity ensures that items measuring the
41. Why did you choose descriptive statistics for same construct actually share variance. I’ll check this
your demographic data? using Average Variance Extracted (AVE) and outer
loadings. An AVE of 0.50 or higher means the
I chose descriptive statistics because they provide a construct explains at least half of the variance in its
simple yet clear way to summarize the characteristics items. Outer loadings of 0.708 or higher mean each
of my respondents. For example, frequency counts item strongly represents the construct. If items fall
and percentages will show how many are male or below this threshold, I’ll evaluate whether removing
female, or how many belong to BEED versus BTLED. them improves the model.
The mean and standard deviation will help describe
46. How will you check discriminant validity? mean a positive effect—for example, higher perceived
usefulness increases behavioral intention. Negative
Discriminant validity ensures that each construct is values mean the opposite. I’ll use bootstrapping with
distinct from the others. I’ll use two tests: 5,000 resamples to test significance. If p < 0.05, the
relationship is statistically significant. This confirms
Fornell-Larcker Criterion: The square root of
whether my hypotheses are supported.
AVE for each construct should be greater than
its correlation with other constructs. 51. What if some indicators have low loadings?
HTMT Ratio: Values below 0.85 mean
constructs are sufficiently distinct. If an indicator has a loading below 0.40, I’ll remove it
because it doesn’t represent the construct well. If it’s
Together, these tests confirm that perceived between 0.40 and 0.70, I’ll check whether removing it
usefulness, ease of use, behavioral intention, and improves AVE and reliability. This ensures that only
intention to use are not overlapping too much. strong and meaningful items remain in the model,
keeping the constructs valid and reliable.
47. Why use Variance Inflation Factor (VIF)?
52. How will you handle missing or incomplete data?
VIF checks whether predictor variables are too highly
correlated, which can distort regression results. For I’ll minimize missing data by checking responses
example, if perceived usefulness and ease of use are during collection. If some questionnaires are
almost identical statistically, it could cause collinearity incomplete, I’ll exclude them from analysis to
problems. Acceptable VIF values are below 3, maintain accuracy. It’s better to work with fewer but
meaning predictors are independent enough to be complete responses than to include incomplete ones
reliable. This ensures my path coefficients are not that could distort the results.
biased.
53. Why combine SPSS and SmartPLS?
48. How will you interpret R² values?
SPSS is excellent for descriptive statistics—it’s simple
R² tells me how much of the variance in a dependent and effective for summarizing demographic data and
variable is explained by its predictors. For example, if perceptions. SmartPLS, on the other hand, is powerful
behavioral intention has an R² of 0.50, it means half of for testing complex models like TAM. By combining
its variance is explained by perceived usefulness and both, I can capture the basic profile of respondents
ease of use. I’ll interpret values as: and also test the deeper relationships among
constructs. This makes my analysis both
0.75 = substantial
comprehensive and rigorous.
0.50 = moderate
0.25 = weak 54. How will you present your data analysis results?
This shows how strong the model is in predicting I’ll present results using tables and figures. For
outcomes. example, frequency tables for demographics,
weighted mean tables for perceptions, and SEM
49. Why include effect size (f²) in your analysis?
diagrams for relationships. This makes the findings
Effect size tells me how much each predictor clear, organized, and easy to interpret. Visual
contributes to explaining the dependent variable. For presentation also helps panelists and readers quickly
example, if perceived usefulness has a large f², it grasp the results without getting lost in numbers.
means it strongly influences behavioral intention.
Even if a path is statistically significant, f² shows
whether its impact is small, medium, or large. This
helps identify which factors are most important in AI
adoption.
50. How will you interpret path coefficients?
Path coefficients (β) show the strength and direction
of relationships between constructs. Positive values