📄 Development of Basic English Grammar Teaching Materials Based on Situational Approach in English Language Education Study Program of Universitas Muslim Indonesia

Summary

The paper looks at how the first-semester Basic English Grammar course is taught in the English Language Education Study Program at Universitas Muslim Indonesia. The course relied on lectures and on grammar books written for native speakers, which left new students passive and struggling. The authors (the PDF also lists Andi Hudriati) tested a situational, Oral Approach style of teaching. In this approach, grammar is first met in speech, practised inside situations, and ordered from simple to complex forms, with reading and writing added later. They ran a quasi-experiment with 50 students: 25 were taught with the situational method and 25 with conventional lectures. Both groups took a pre-test and a post-test. The experimental group's mean rose from about 69 to about 88. The control group's mean rose from about 71 to about 84. The authors credit the bigger gain to more active student participation, with the lecturer acting as a facilitator. The study matters as local evidence for activity-based grammar teaching. However, its statistical support is thin: it reports only descriptive statistics and gives no significance test.

Key points

• Design: nonrandomized pre-test → treatment → post-test with a control group, n = 50 (25 + 25), at the UMI Faculty of Literature.
• Scoring: number of correct answers × 3.3, so 30 items give a maximum of 99.
• Experimental group: mean 69.036 → 88.308, minimum 26.4 → 72.6. In the top band (75.6–99.6), 8 students before, 23 after.
• Control group (lecture / Teacher Center Learning): mean 71.148 → 84.216, a gain of 13.07. It started from a slightly higher baseline.
• The claim of a "significant difference" rests on comparing mean gains only. No t-test, effect size or p-value is reported, and the text contains small arithmetic slips (for example, the experimental gain is given as 19.00).

Interesting ideas

• I1. Teaching grammar items strictly from simple to complex forms gives a principled curriculum order, not a random one.
• I2. New language units are introduced inside concrete situations rather than as isolated rules, so form and use are learned together.
• I3. Running a control arm alongside the treatment shows that both groups improve, so what counts is the extra gain over the control, not the raw post-test score.
• I4. The two groups did not start at the same level, which weakens a comparison based only on means.
• I5. Banded frequency tables reveal how the spread of results shifts, for example how many low scorers move up, which a single mean can hide.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

• I1 (learn): order training and self-play material by structural complexity, for example clause depth or the number of dependents, so the parser and its repair rules are learned on simple constructions before harder ones.
• I2 (world): attach each frame or grammatical pattern to a small situational scene (entities and states) in tests, so that a construction counts as understood only if the world it builds is correct.
• I3 (learn): always report improvement relative to a baseline run on the same data and the same schedule, as a negative control, before attributing gains to a new component.
• I4 (answer): when comparing readers on school-course tests, report the starting accuracy and the gain on the same item set, and add a significance test, rather than citing post-test averages alone.
• I5 (answer): report score distributions per question type or per text, so that improvements concentrated among the weakest cases are visible.