## Step 5: Formula Conclusions

(m) Summarize your conclusions in context from the matched-pairs study.

• Significance: Was the sample mean difference in melting time significantly different from zero? (support your answer)
• Estimation/Confidence: How large is the average difference in melting time in the population? (support your answer)
• Causation: Are you able to draw a cause and effect conclusion based on the type of chip? Explain. Hint: Think about what type of randomness is important for the matched-pairs design.
• Generalizability: Are you willing to consider our class as representative of all college students on this issue? Explain briefly.

## Step 6: Look back and ahead

One of the benefits of designing a paired study is to reduce variability, so that we can make a more powerful comparison of the two treatments we are interested in.

(n) Explain why it made sense in advance to use a paired design for this particular study. That is, why might it be beneficial to have everyone melt both chips instead of some people melting the %chip1% chips and everyone else melting the %chip2% chips and comparing those two distributions?

(o) Did the pairing actually appear to be effective in this study? That is, was there less variability in the differences than in the individual melting times? Justify your answer.