SPSS Log-Likelihood Analysis

I have already extracted all modal-verb tokens from two learner corpora—one tagged “Intermediate,” the other “Advanced”—and calculated their normalized frequencies (per 10,000 words). What I still need is a statistically sound comparison of those frequencies.

Please run log-likelihood (LL) tests in SPSS to determine whether the differences between the two proficiency levels are significant. If the LL results indicate that additional inferential statistics would clarify the picture—say, a follow-up effect-size measure or confidence interval—you may include them, but keep the focus on the LL output itself.

Deliverables
• SPSS syntax file and/or detailed steps so I can replicate the procedure
• The SPSS output tables and a brief interpretation in plain English (no jargon—just what is significant and why)
• A short note on any extra tests you considered necessary and the rationale behind them

Everything else—the raw counts, corpus sizes, and my existing frequency tables—will be shared as soon as the contract starts, so you can plug them straight into SPSS and get to work quickly.