Statistics for Business: Decision Making and Analysis, 3rd Edition

You Do It

1. 39. Emerald diamonds

   1. The two categories have similar average weight. Weight is unlikely to confound this analysis.

      Level	Number	Mean	Std Dev
      VS1	90	0.4136	0.0544
      VVS1	54	0.4081	0.0537

   2. The two−sample t−test finds a statistically significant difference: VVS1 diamonds cost on average about $112 more than VS1 diamonds.

      VS1–VVS1, allowing unequal variances

      Difference	− 112.30 t Ratio	− 2.88504
      Std Err	38.93 DF	103.4548
      95% CI	− 189.50 to − 35.11 Prob > t	0.0048

   3. Because the interaction is not statistically significant, remove it and refit the model. Without the interaction, the estimated slope for clarity is significant.

      R2	0.4974
      se	161.8489
      n	144

      Term	Estimate	Std Error	t−statistic	p−value
      Intercept ...