Question 16.34

16.34 It’s all in the thumbs.

New mobile phone software is advertised as resulting in faster text message creation (measured in characters per second) when compared with traditional mobile phone software. The traditional software offers suggested word completions based on a list of commonly used words, while the new software uses a smartword completion algorithm that takes into account the context of the text message. Ten individuals are involved in an experiment to test the marketing claim. The following data are the differences in their text message creation speeds: a positive value indicates more characters per second with the new mobile phone software.

txtmes

(new minus old) 5 3 0 1 −1 7 2 −2 4 0
  1. Plot these data on a Normal quantile plot. Would a procedure be appropriate? Explain your answer.
  2. Is there statistical evidence to support the marketing claim? Remember that zeros are dropped from the data before ranking, so that is the number of nonzero differences within pairs. Clearly state the relevant details from the signed rank test.