How to Track and Analyze Your Own English Learning Data for Better Results

Data mining requires data, yet learners often lack the hardware and software to capture it automatically, so they must collect their own learning data.

Inspired by a roommate studying English, the author considers various classroom‑level metrics: the proportion of English versus Chinese used by the teacher, a student’s comprehension of each language, the correlation between Chinese and English content, and whether misunderstandings stem from vocabulary limits or auditory sensitivity.

Potential indicators are illustrated below:

These personalized metrics are hard for teachers to observe directly; only the learner may know them, and even the learner may be uncertain. While the author does not provide exact values, similar indicators appear in English‑learning research.

Without interviews, the learner can assign provisional values to their own experience—e.g., estimating that English occupies 80% of class time and that comprehension is 50%—and compare across multiple sessions to gauge progress.

Recognizing that a single‑class view is narrow, the author expands the scope to track weekly study time, content, and outcomes, as shown here:

Memory time reflects word‑memorization effort; expression time covers speaking or writing practice, and learning‑effect measurement remains challenging due to limited data sources.

Extending the observation period, the author seeks to map these metrics to final learning outcomes, a more complex task.

To assess personal learning ability improvement, the author proposes three new indicators—reception (listening/reading), thinking (processing stored knowledge), and expression (output)—illustrated below:

By comparing data over a couple of weeks, one can see what activities were performed and gain insights for adjusting study habits.

The author then applies this approach to vocabulary memorization, using a strategy of reading texts, looking up unknown words, and employing information‑processing techniques such as repetition, imagination, and spelling, followed by testing.

The specific metrics used are:

‘See’ counts how many times a word is viewed; ‘Hear’ counts listening to its pronunciation; ‘Think’ counts associating the word with a Chinese or visual representation; ‘Speak’ counts silent rehearsal; ‘Spell’ counts imagined or written spelling. ‘Single session’ measures the time from first exposure to memorization completion; ‘Test’ measures the interval between memorization and recall testing; ‘Effect’ rates recall as 1 (no impression), 2 (partial), or 3 (complete).

Collected sample data (shown below) reveal that the strategy yields acceptable memorization results, with reduced operation time as proficiency improves, allowing calculation of averages and trend analysis.

Further reflection on the data leads the author to hypothesize a memory schema for word learning, illustrated here:

While prior accumulation is needed to retrieve related words, memorizing a word is not the sole goal; the willingness and proficiency to use new words also matter, aspects not captured by data alone.

Collecting data initially feels cumbersome, but with repeated practice it becomes a valuable tool for informed decision‑making and continuous improvement.