Computer software is commonly used to translate text from one language to another. As part of his Ph. D. thesis, Philipp Koehn developed a phrase-based translation program called Pharaoh. The quality of the translation can vary. A good translation system should match a professional human translation. It is important to be able to quantify how good the translations produced by Pharaoh are. The IBM T. J. Watson Research Center developed methods to measure the quality of a translation from one language to another. One of these is the BiLingual Evaluation Understudy (BLEU). BLEU is a score ranging from 0 to 1 that indicates how well a computer translation matches a professional human translation of the same text. Higher scores indicate a better match. BLEU helps companies who develop translation software "to monitor the effect of daily changes to their systems in order to weed out bad ideas from good ideas." To compare Pharaoh's ability to translate with similar computer translation software, Koehn took a random sample of 100 blocks of Spanish text, each of which contained 300 sentences, and used Pharaoh to translate each of these to English. The BLEU score was calculated for each of the 100 blocks. He wants to use this data to see if it differs from the mean BLEU score of another leading translation software which has a population mean score of 0.295. Open the data file BLEU-Scores.

1. . Assuming the requirements are satisfied, calculate a 95% confidence interval for the mean of the BLEU test scores.

2. Calculate the degrees of freedom and the test statistic for a test of H0:μ=0.295H0:μ=0.295 against Ha:μ≠0.295Ha:μ≠0.295. Assume the requirements are satisfied.

3. Calculate the `P`-value for a test of H0:μ=0.295H0:μ=0.295 against Ha:μ≠0.295Ha:μ≠0.295. Assume the requirements are satisfied.

4. Based on the results of this test, what would you conclude? Use a level of significance of α=0.05α=0.05

A. We have sufficient evidence to say that the true mean is equal to 0.295.
B. We have insufficient evidence to say that the true mean is equal to 0.295.
C. We have insufficient evidence to say that the true mean is different than 0.295.
D. We have sufficient evidence to say that the true mean is different than 0.295.