MA Thesis · 2026 · SWPS University Warsaw · Faculty of English Philology
Enhancing Willingness to Communicate through Speaking Activities in the EFL Classroom
A small-scale action research with four Polish Grade-2 learners
Author
Ngoc Yen Nguyen
Supervisor
Sylvia Maciaszczyk, PhD
01 · Background
Young EFL learners often know the language but do not speak. MacIntyre et al. (1998) frame this gap with L2 Willingness to Communicate (WTC) — the readiness to enter discourse, at a particular time, with a specific person, using L2.
Most WTC research focuses on adolescents and adults. Very few studies examine 7-8-year-olds with adapted instruments. This study fills that gap.
02 · Research Questions
- RQ1. What changes can be seen in observed WTC-related behaviour across three AR cycles?
- RQ2. Which activities improved WTC, and how do patterns reflect pre-post SPCC/FLCA?
- RQ3. Which activities do learners like most/least, and how does this relate to behaviour & self-report?
03 · Theoretical Frame
MacIntyre et al.'s pyramid (1998). Top 3 = situational state. Intervention targets layers 2-4.
04 · Method
4
Learners
Grade 2 · A1
8
Weeks
Feb – Apr 2026
3
AR Cycles
Plan → Reflect
5
Activities
PT·TT·IG·RP·FT
Mostly qualitative classroom action research with mixed-methods support. Four data sources: SPCC (Form 1), FLCA (Form 2), Observation (Form 3, 4 indicators), Activity Feedback (Form 4).
05 · Behavioural Trajectory (Figure 6)
Non-linear pattern: growth W1-W4 → W5 dip (language load + absence) → recovery via board game in Cycle 3.
06 · Pre vs Post — Group Mean (Figure 7)
SPCC ↓ (3.29 → 2.92) and FLCA ↑ (1.20 → 1.70) while observed behaviour rose.
07 · Activity Preference (Figure 10)
PT & TT most liked. IG harder but drives strongest peer interaction.
08 · Key Findings
- 1. WTC participation was dynamic, not linear — shaped by task, peers, interruption, language demand.
- 2. Speaking first, asking questions, and peer talk grew clearly; longer answers remained weakest.
- 3. Divergence: observed behaviour ↑ but self-report SPCC ↓ / FLCA ↑ — confirms MacIntyre's layered model.
- 4. Preference ≠ Effectiveness — sequencing matters more than picking one "best" activity.
09 · Recommendations
- 1. Use AI-assisted worksheets for hesitant learners.
- 2. Sequence structured → communicative activities.
- 3. Use playful formats (board games) with explicit turn-taking.
10 · Limitations
n = 4 → transferable, not generalisable · teacher-researcher dual role · 8-week scope · child-adapted Likert scales · no control group (ethically unworkable).