Chapter 2-3

Where in your RRL is the basis for AI usage across job positions stated?

In the opening literature review sections discussing AI adaptation across judicial and public sector administrative positions.

Where in the RRL does it state that there is a research gap?

In the first paragraph of the literature review, which highlights the lack of local empirical research in Philippine court.

Does your study suggest employees should fully depend on AI?

No. The literature shows AI acts as a force multiplier for efficiency, but human legal reasoning, fact verification, and critical thinking remain mandatory.

Do all court employees use AI?

No. Preliminary interviews showed AI usage varies depending on specific job duties.

What is your research design and why did you choose it?

Our study uses a quantitative descriptive-correlational research design. We chose this for two specific reasons:

Descriptive: To measure baseline levels of AI tool usage and job performance.

Correlational: To test if a statistical relationship exists between them without disrupting daily work processes.

Why quantitative instead of qualitative?

We chose a quantitative approach because our research requires statistical proof rather than subjective opinions.

Why not descriptive design only?

A purely descriptive design cannot test relationships or answer whether AI usage links to job performance.

Why conduct preliminary interviews if this is a quantitative study?

Conducted as due diligence to confirm that RTC employees were actively using AI tools before distributing the survey.

How did you select your sample and determine sample size?

We selected our respondents and determined our sample size using purposive sampling targeting 70 accessible RTC employees actively across participating RTC branches.

How were the study respondents computed?

Using Slovin’s Formula with a 5% margin of error, we surveyed 70 accessible court staff out of a total population of 85.

How did you gather your data?

Distributed structured 5-point Likert scale questionnaires via Google Forms.

How did you process your data? / What statistical tools were used to analyze the data?

We processed our quantitative data using SPSS through four main statistical treatments:

Frequency & Percentage (profiles)

Weighted Mean & Standard Deviation (AI usage and job performance levels)

One-Way ANOVA (group differences)

Pearson Correlation r (relationships).

Why is your sample size (N = 70) appropriate?

It is appropriate because it captures 100% of the active court staff who directly handle daily legal documentation.

The survey achieved Cronbach’s Alpha scores of 0.86–0.95, confirming reliability

Were pilot test respondents included in the final 70 sample?

No, the 30 pilot test respondents were strictly excluded from the final dataset.

Why did you use a fully modified questionnaire?

We used a fully modified questionnaire because the existing instruments we found did not completely match the specific context, variables, and respondents of our study.

What does Cronbach's Alpha measure and what was your result?

It measures questionnaire internal consistency. All constructs scored between 0.86 and 0.95, well above the 0.70 reliability threshold.

Why choose a 0.70 Cronbach’s Alpha benchmark?

It is the standard social science threshold for acceptable internal consistency.

How were Likert-scale questionnaire responses interpreted?

Mean scores were categorized into descriptive interpretation levels, such as Agree.

What is the basis for analyzing your research data?

Standard quantitative statistical tools matched to the research objectives and data scale.

Was data processing manual or computerized?

Computerized, using Google Forms, Google Sheets, and SPSS software.

How did you generate the raw tally?

Survey responses were collected via Google Forms, compiled into an Excel sheet, and totaled per item before running analysis in SPSS.

What level of significance did you use?

0.05 level of significance (95% confidence level).

Why did Task Performance (0.95) score higher reliability than Facilitating Conditions (0.86)?

It was due to work standardization versus infrastructure variance.

Task performance measures standardized legal duties, while facilitating conditions evaluate hardware, internet, and tech support.

Why use One-Way ANOVA instead of an Independent T-Test?

All demographic variables contain 3 or more independent subgroups to compare simultaneously.

Why use Pearson r correlation instead of Linear Regression?

Pearson r checks if two things are connected, while linear regression builds a formula to predict exact scores.

Since our research objective was strictly to discover whether a statistical link exists between our variables, Pearson r was the exact tool we needed.

What does the high number of Court Interpreters in Table 2.2 mean?

It shows that many respondents perform heavy information processing and communication tasks, which affects how AI tools are used.

Why do Court Interpreters represent nearly a quarter (24.3%) of respondents?

They form a major segment of frontline court staff heavily engaged in document and language processing.

Why is length of employment the ONLY factor creating significant AI usage differences?

Tenure reflects practical legal experience which drives effective AI adaptation far more than age or job title alone. Also, experienced staff know exactly which court routines can be safely offloaded to AI.

Why do senior staff rate ease of use higher than younger staff?

Senior staff know legal workflows thoroughly, making it easier to guide AI tools to produce correct outputs.

Why is job performance uniform across all demographic groups?

Job performance is uniform because court operations are strictly governed by Supreme Court rules. Every employee—regardless of age, position, or tenure—must meet the exact same legal standards, templates, and procedural deadlines.

Why did Social Influence score lower than Performance Expectancy and Facilitating Conditions?

Social Influence scored lower because AI adaptation in courts is functional, not social. Staff use AI because it cuts down their paperwork load (PE) and because they have the necessary tech setup (FC), not to follow peer pressure or trends.

Why is Task Performance rated higher than Adaptive and Contextual Performance?

Task performance scores higher because core legal duties are mandatory, while adaptive and contextual behaviors are optional.

Task performance involves non-negotiable core duties (such as case processing) governed by mandatory court deadlines.

Adaptive behaviors (handling unexpected changes) and contextual behaviors (helping peers) are voluntary, so staff naturally prioritize their required workload first.

Why do staff adjust well to new tech (3.96) but rarely seek tech training independently (3.80)?

Staff adjust quickly out of operational necessity to handle high caseloads and meet deadlines. However, because their daily schedules are already packed, they lack the extra time and incentive to seek out voluntary tech training on their own.

You state your study is located in Manila, but Chapter 1 mentions CALABARZON. Which location is correct?

Our actual locale is in Regional Trial Court, Manila City Hall. The mention of CALABARZON was a drafting error, and we will correct it in the final manuscript.

Chapter 2 lists an independent t-test, but there are no t-test results in your data analysis. Did you actually run this test?

No, we did not. The t-test was part of our early proposal but was not needed for our final data analysis. Its inclusion in Chapter 2 is a leftover draft error, and we will remove it.

Your paper uses the word 'adoption' throughout, but you claim to focus on 'adaptation.' Which process are you actually studying?

Our study strictly focuses on adaptation—how staff adjust their work habits when using AI. Using adoption was a drafting oversight, and we will replace it with 'adaptation' throughout the paper.

Your paper states a population size of 100 in Sample and Sampling Technique, but your sample size calculation uses N = 85. What is your actual total population?

Writing 100 was a drafting error, but our sample size of 70 was correctly calculated using 85. We will update all references to 85 in the final paper.