Efficiently establishing concepts of inferential statistics and hypothesis decision making through contextually controlled equivalence classes.

The team built three short computer lessons. Each lesson used matching-to-sample to link stats terms, formulas, and decisions.

Ten college students with no stats background took the lessons. The goal was to see if the kids could pick the right test and decide if results were significant.

After only three lessons every student passed the final test. They picked t-tests, ANOVA, or chi-square without hints.

Kids also answered questions that were never taught. For example, they matched a new p-value to “reject” or “fail to reject” even though that pair was not drilled.

Melchiori (2000) taught kids to read new words the same way. Both studies show that once you build equivalence classes, untaught skills pop out.

Sigurðardóttir et al. (2012) used the same matching game with Icelandic nouns. Adults then used plural forms they had never seen, just like these students picked correct stats rules.

Branch (1999) tells us to ignore p-values. This paper does the opposite—it trains students to use them. The clash is only on paper. N worries about over-trust; M et al. simply teach the tool correctly.

You can pack an entire stats unit into three 15-minute drills. Use matching-to-sample software before your next research-methods class. Students leave knowing which test to run and how to read the p-value, even if you never lectured on it.