Course artifact: Reflection #1, the baseline

I entered STA496 with a strong personal value of “bridging the gap between the rigorous, number-driven side of statistics and the human, interpersonal side” of learning it. I was excited to blend course concepts and this personal value to explore how this gap could be closed in statistics, specifically for life science students given STA496’s conceptual focus in the life sciences. I had an optimistic belief that delivering statistics in a more engaging manner, potentially targeting the human side of learning, could improve learning outcomes for students.

PDF · personal reflection

STA496 Reflection 1

This simple one-page, three paragraph reflection enshrines my original values (making statistical learning more accessible to youth, blending the quantitative and human side of statistics), and beliefs (that making the learning process more engaging could boost statistical takeaways for students), before later literature searches revised certain beliefs and complicated the picture.

Why it matters: reflection 1 establishes the “before” state needed to demonstrate a true learning timeline. By clearly showcasing this “before” state and reflecting on what values, beliefs, and assumptions I started with as a learner, later stages in the ePortfolio can show how my understanding of what effective statistical learning revises with updated evidence. Additionally, this showcases how the final annotated bibliography, one of the last ePortfolio stages, is simply a refinement of this “before” state.

Open Reflection #1 ↗

Mind-map brainstorming, what were my own ideas (even if seemingly disconnected and arbitrary)?

Before beginning the process of narrowing and refining, I mapped the course topic against my own interests and motivations, using the values I had touched on during the Reflection 1 artifact to help me develop new ideas about what I would be interested in exploring. In terms of learning, the goal here was not yet to produce a polished research question; it was to more clearly understand where undergraduate life-science statistics education overlapped with my values, and which strands of overlap could potentially turn into a research focus.

Original handwritten mind map · May 2026

Literature search log, learning to treat searching as data

This STA496 literature search tracking log was relevant to my learning experience as more than just a list of academic papers. Comparing “number of results” and “search terms” columns against “relevance of results” for each distinct literature search, primarily conducted on Google Scholar, helped me learn what an effective literature search comprises (specific research terms, date-filtering, and an interactive trial-and-error approach). Reflecting on my process using the log, earlier searches ranged from 158,000 results for “life science statistics education” to 17,200 results for “undergraduate statistics education restructure”, with one early-stage literature search trial being bluntly recorded as “not great, too broad”. By refining later searches by combining key course-concept related search terms such as “ChatGPT”, “statistics education” and “using”, I was able to obtain a lower result count with each source being more relevant to my potential research interests.

PDF · research-process artifact

Tracking Your STA496 Searches

This artifact is a 13-page log recording search terms, searching platform, date of search, any filters used, and number of searches yielded, as well as a written description of the relevance of the search and potential citations to further explore. This log serves an iterative purpose: each next search can be improved based on reflections made during the prior search.

Why it matters: this artifact demonstrates how my research skills (specifically literature search research skills) developed over the term. When I began using the literature search tracking log, I viewed each entry as a one-time retrieval task. By the end of STA496, I viewed the log as an iterative process, where each result’s relevance directly influenced how the next search term would be conducted by narrowing the search scope and recording specific words that yielded higher quality results. For future research, I am now better equipped in my literature search skills and understand literature search to be a continuous, reflective process instead of many separate searches.

Open full search log ↗

Second mind-map, what patterns were appearing?

As my literature search log expanded, I used another mind-map; this time instead of brainstorming ideas, I used the mind-map to visualised which patterns were occurring across literature pieces, and where connections were present. I noticed course redesign, student engagement, and genAI usage were recurring themes. This encouraged me to question if there was a way to connect all three themes, or pull elements from each to support my course focus of statistics education improvement.

Original handwritten literature-pattern mind map · Late May / early June 2026

Initial topic of interest, improving first-year statistics education

Search-log excerpt · June 1

A first substantive narrowing

This page records the decision to investigate first-year statistics education and to borrow potentially transferable strategies from outside life-science contexts.

Why it matters: it shows an evidence-based reason for narrowing and an emerging willingness to cross disciplinary boundaries while still asking whether the transfer was justified.

Modifying the topic again, bringing GenAI into the problem

Search-log excerpt · June 3

From “improvement” to a specific educational intervention

The juxtaposition on this page is important: a highly narrowed first-year search sits directly above the first explicit ChatGPT search. The research problem was beginning to move from broad curriculum reform toward how a particular tool changes learning.

Why it matters: this is the pivot point where the project becomes recognisably connected to the eventual annotated bibliography.

Course artifact: Reflection #2, search became iterative rather than mechanical

By Reflection #2, I could articulate a methodological change in my own behaviour: I was no longer keeping one fixed search query. I was judging relevance, changing terms, comparing natural-language and keyword searches, and allowing the search process to reshape the research focus. I also explicitly identified the emerging intersection of first-year learning, ChatGPT, and life-science curriculum needs.

PDF · reflective artifact

STA496 Reflection 2

The reflection records both a technical research skill, better keyword refinement, and a conceptual shift toward using literature search as an ongoing feedback loop.

Why it matters: it provides direct evidence of learning progression. The change was not simply that I knew more papers; I had changed how I approached finding and evaluating them.

Open Reflection #2 ↗

Brainstorming the preliminary ePortfolio, making the medium match the goal

I wanted the ePortfolio itself to communicate the bridge I cared about: rigorous statistical research on one side, and the human process of learning on the other. That led to a timeline rather than a static repository, and to including informal notes beside formal documents.

Original handwritten ePortfolio planning mind map · June 2026

Preliminary ePortfolio, a clear format, but an unfinished intellectual destination

My preliminary plan already named the four phases used here, collection, selection, reflection, and connection, and committed to a permanent personal website. At the same time, the plan was stronger on how I would show the process than on the precise research question the process would ultimately answer. That gap became useful evidence of where my learning still needed to develop.

Feedback mind-map, turning critique into a narrower question

Feedback highlighted three especially useful directions: the scope could be narrower; relevant evidence might sit beyond statistics-education journals; and emerging concerns such as over-reliance, responsible GenAI use, and cognitive laziness were worth considering. Rather than treating that feedback as a list of edits, I mapped it against the existing project and used it to choose what the next literature search needed to resolve.

Original handwritten feedback-response mind map · July 2026

Course artifact: Reflection #3, breadth stopped being automatically synonymous with strength

Reflection #3 is where the self-critique becomes explicit. I identified creativity, synthesis, and relevant paper selection as strengths, but also recorded the need for clearer descriptions, cleaner citations, and a narrower research focus. I specifically named course-specific chatbots and cognitive laziness as possible next directions.

PDF · reflective artifact

STA496 Reflection 3

This reflection documents both confidence and limitation: I could synthesize a large body of relevant papers, but I needed to become more selective about language, scope, and what evidence could genuinely support.

Why it matters: it captures critical self-awareness rather than a success-only narrative. I revised an implicit belief that “more papers and more connections” necessarily meant a stronger project; disciplined narrowing became part of the skill I was learning.

Open Reflection #3 ↗

Working on the annotated bibliography, a useful binary that I eventually had to reject

I refined the topic to the tension between GenAI as a personalised learning support and GenAI as a source of cognitive laziness. For the working bibliography, I deliberately searched both sides: studies of personalised feedback, pacing, and conceptual support, and studies of over-reliance, cognitive offloading, reduced critical thinking, and self-regulation.

Working research snapshot · August

Personalised support vs. cognitive laziness

The two-lens structure gave the bibliography a tractable way to collect opposing findings. It was analytically useful, but I later recognised that “pro-GenAI” and “anti-GenAI” were too blunt: studies were often measuring different outcomes or different modes of use.

Why it matters: this is where refinement becomes conceptual rather than merely narrower. The final insight came from questioning the categories I had created for the literature.

Open working snapshot ↗

Final annotated bibliography, from “which side is right?” to “what mechanism explains both?”

The final synthesis across twelve core sources did not support a binary conclusion. Studies reporting personalised feedback, guided problem solving, pacing, and repeated examples found stronger engagement, reasoning, or learning gains; studies centred on over-reliance found weaker independent problem-solving, critical thinking, metacognitive effort, and self-regulation. The pattern connecting both groups was how GenAI was used: as scaffolding for a learner's own cognitive process, or as a substitute for it.

PDF · final research artifact

Conceptual Growth or Cognitive Shortcut?

The 13-page final bibliography evaluates source methodology and relevance, then synthesises evidence across personalised learning, statistical reasoning, cognitive offloading, self-regulation, and instructional design.

Why it matters: this artifact is the endpoint of the timeline and the strongest evidence of growth in statistical-education knowledge, source evaluation, and synthesis. It does not merely collect research; it reconciles apparently conflicting findings by identifying a mechanism that explains when each is plausible.

Read final bibliography ↗

Synthesis excerpt · page 10

The distinction that resolved the tension

The synthesis separates learning outcomes that initially looked interchangeable. Motivation, confidence, task performance, statistical reasoning, and knowledge gain are different measures. A student can become faster or higher-scoring without developing an equivalent increase in conceptual understanding.

Why it matters: this distinction changed my own definition of successful statistical education. Statistical literacy cannot be inferred from a polished output alone; the learner's reasoning process matters.