IB Mathematics AI Internal Assessment (IA) Guide

Understanding the AI IA Rubric: Your Blueprint for Success

Before you even think about a topic, internalize the AI IA rubric. It's not just a grading tool; it's your project plan. The five criteria are: Communication, Mathematical Presentation, Personal Engagement, Reflection, and Use of Mathematics. Each has specific descriptors for different mark bands. For example, 'Communication' isn't just about neatness; it's about clear, coherent, and concise presentation of your ideas and findings.

Crucially, 'Use of Mathematics' demands sophisticated and relevant mathematics applied appropriately, demonstrating full understanding. For AI, this often means robust statistical analysis, advanced modeling techniques, or complex data manipulation. 'Reflection' is where many students falter, simply summarizing results. True reflection involves discussing limitations, extensions, and the implications of your findings, demonstrating critical awareness of your work.

My school's math department provided us with annotated examples of IAs from previous years that scored well across different criteria. If your school offers this, take advantage of it. It's invaluable to see how high-scoring IAs address each criterion in practice, not just in theory. Pay close attention to how they present their data, justify their methods, and reflect on their conclusions.

Topic Selection: Finding Your Niche (and Data)

This is perhaps the most critical initial step. Your topic must allow for sufficient mathematical exploration relevant to the AI syllabus, be personally engaging, and crucially, have readily available, clean data. Avoid overly ambitious or vague topics. 'Analyzing climate change' is too broad; 'Modeling the correlation between average monthly temperature and electricity consumption in Tokyo from 2010-2020' is specific and data-rich.

Think about your other IB subjects or personal interests. Are you passionate about sports? Maybe analyze player performance statistics. Interested in economics? Explore market trends or inflation rates. My own interest in societal inequalities led me to investigate education access. The key is to find a topic where you can genuinely apply AI concepts like regression, hypothesis testing, time series analysis, or network analysis.

Data availability is paramount. Publicly accessible datasets (government statistics, academic research, open data portals like Kaggle, World Bank, or national statistical offices) are your best friends. Avoid trying to collect primary data unless it's very simple and controlled, as it adds unnecessary complexity and potential for bias. Ensure your chosen dataset is large enough to allow for meaningful statistical analysis – typically at least 30 data points for basic regression, and more for complex models.

Structuring Your IA: A Logical Flow

A clear structure guides both you and your examiner. I recommend the following: Title Page, Table of Contents, Introduction, Rationale/Research Question, Methodology, Data Collection/Presentation, Data Analysis, Results, Discussion/Conclusion, Limitations, Extensions, Bibliography, and Appendices. Maintain a consistent and professional formatting throughout.

Your 'Introduction' should set the scene, introducing your topic and its relevance. The 'Rationale/Research Question' section is where you clearly state your focused research question and explain why it's worth investigating mathematically. This question should be specific, measurable, achievable, relevant, and time-bound (SMART). For example, 'To what extent does the number of public transportation stops within a 1km radius predict average housing prices in Shibuya ward, Tokyo, in 2023?'

The 'Methodology' section is where you outline the specific mathematical tools and techniques you plan to use, justifying their relevance to your research question. This is not just a list; explain *why* you chose a particular regression model over another, or *why* a specific statistical test is appropriate for your data. This demonstrates deeper understanding.

Data Collection and Presentation: Clarity and Integrity

Detail your data source (e.g., 'Tokyo Metropolitan Government Statistics, 2020-2023'). Explain any data cleaning or preprocessing steps you undertook (e.g., handling missing values, standardizing units). Present your raw data, or a representative sample, clearly, often in an appendix. In the main body, use appropriate tables and graphs (histograms, scatter plots, box plots) to visualize key aspects of your data. Label everything meticulously.

When presenting graphs, ensure they are clear, concise, and enhance understanding. Avoid cluttered visuals. Each graph should have a clear title, labeled axes with units, and a brief description of what it shows. Don't just dump graphs; explain what insights they provide. For instance, 'Figure 1 shows a strong positive linear correlation between X and Y, suggesting...' This links your presentation directly to your analysis.

The Core: Data Analysis and Mathematical Application

This is where you demonstrate your mathematical prowess. Apply the techniques you outlined in your methodology. For AI, this often involves: descriptive statistics, correlation analysis, linear/non-linear regression, hypothesis testing (t-tests, chi-squared tests), time series analysis, or even network analysis. Show your working clearly, but don't just copy-paste calculator outputs. Explain the steps, the formulas, and the interpretation of your results.

For example, if performing regression, explain the R-squared value, the significance of your p-values, and the meaning of your regression coefficients in the context of your research question. If conducting a hypothesis test, state your null and alternative hypotheses, choose an appropriate significance level, calculate the test statistic and p-value, and make a clear conclusion based on the evidence.

Leverage technology effectively. Graphing calculators (TI-84, Nspire), spreadsheets (Excel, Google Sheets), or statistical software (R, Python, GeoGebra) are powerful tools. However, *you* must interpret the output. Don't just present a p-value; explain what it means for your hypothesis. This is where many students lose marks – they show the math but fail to explain its significance in relation to their specific problem.

Discussion, Conclusion, Limitations, and Extensions: Critical Thinking

Your 'Discussion' should interpret your results in the context of your research question. Did you answer it? What do your findings mean? Compare your results with existing knowledge or expectations. This is not just a summary; it's an analytical interpretation. The 'Conclusion' should succinctly state your main findings and answer your research question.

The 'Limitations' section is crucial for demonstrating reflection. No study is perfect. Discuss the weaknesses of your data (e.g., small sample size, potential biases, limited scope) and your methodology (e.g., assumptions made, choice of model). Be honest and specific. For instance, 'The linear regression model assumed a constant relationship, but a more complex non-linear model might have captured nuances not addressed here.'

Finally, 'Extensions' demonstrate foresight. What could be done next? How could your study be improved or expanded? This shows a genuine interest in the topic and a deeper understanding of the research process. Could you use a different dataset, a more advanced statistical technique, or investigate related variables? This is where you really show that you've thought beyond the immediate scope of your IA.

Personal Engagement and Reflection: Beyond the Numbers

Personal Engagement isn't just about choosing a topic you like. It's evident in your genuine curiosity, the depth of your investigation, and the independent choices you make. This can be shown through exploring unexpected results, dealing with challenges, or going beyond the basic requirements of the syllabus. For example, if your initial data didn't yield clear results, how did you adapt your approach? Did you research alternative models?

Reflection, as mentioned, is more than summarizing. It's about critically evaluating your entire process. What did you learn? What challenges did you face and how did you overcome them? What are the implications of your findings? What are the limitations of your study and how might they be addressed in future research? This metacognitive aspect is highly valued.

Consider asking yourself throughout the process: 'Why am I doing this step?' 'What does this result truly mean?' 'What are the weaknesses of my approach?' 'What could I do differently next time?' Documenting these thoughts as you go can make writing the reflection section much easier and more authentic.

Final Polish: Presentation and Academic Integrity

Proofread meticulously for mathematical errors, grammatical mistakes, and typos. Ensure all graphs, tables, and figures are correctly labeled and referenced in the text. Maintain a professional tone throughout. Your IA should be coherent, concise, and easy for an examiner to follow. Adhere strictly to the word count (typically 2000-2250 words for HL, 1800-2000 for SL, excluding appendices and bibliography).

Cite all sources using a consistent referencing style (e.g., MLA, APA, Chicago). This includes data sources, academic articles, textbooks, and any online resources. Plagiarism is taken extremely seriously by the IB. Ensure every idea that isn't your own is properly attributed. Your bibliography should be comprehensive.

Finally, give yourself ample time. The IA is not a last-minute project. Start early, work consistently, and seek feedback from your teacher at various stages (topic approval, methodology review, draft review). Iteration is key to refining your work and achieving that top score.