TC3 → Stan Brown → Statistics → AAO pt 21 Guide
revised Jan 13, 2004

Viewer’s Guide to Against All Odds Part 21, Inferences for One Mean

Copyright © 2003–2008 by Stan Brown, Oak Road Systems

Cuing: start 11 min in and play 8½ minutes.

________________________ studies are usually more reliable than single-sample investigations. One type is a _______________ _____________________ study, where pairs are matched up. A random member of each pair gets the _________________, and the __________________ between the two results in the pair is your data. Another kind is the classic ______________ and _____________ study on the ___________ ________________.

Example: Loss of  ___________________  over time
Taste of cola is evaluated right after it’s made, and again after one month. The difference will be analyzed with a _________ ____ test.

Panelists are first ___________________ by reference to fixed points on the ____________________ scale. Then they taste their samples and record their data. A month later they come back and do it again, with aged cola. This eliminates all sources of _______________________ besides the _______________ in the cola itself. Ha is that sweetness after one month is ______________ initial sweetness. It’s a __________-tailed test because researchers are only interested in whether cola ____________ sweetness over time.

Sweetness ratings are shown below left. We first found the _____________________ between the pairs of scores. This last column is a _____________ ___________ of observations to analyze.

PanelistBeforeAfterd = B−A
16.34.32.0
27.87.40.4
37.77.00.7
46.74.72.0
57.88.2-0.4
68.15.92.2
79.310.6-1.3
87.96.71.2
96.25.11.1
108.25.92.3

= _________, meaning that the mean taste score _______________ by 1 unit over a month’s time. s = ________.
Ho: μd ______ 0 (sweetness doesn’t deteriorate)
Ha: μd ______ 0 (sweetness deteriorates over time)
xbar minus mu, over quantity sigma over root n
df = ______, p (from table) is between _____ and _____. (From TI-83, p = .0104.)

_____________ Ho and ______________ Ha. Conclusion: Sweetness of the test soda __________ over time.

Three advantages of t tests:
1. Don’t need to know _________
2. __________ samples are okay
3. Test is ___________ (works fairly well for non-____________ distribution, if n > about _______)
Caution:  Must have a good __________ sample, and watch out for ___________.


This page is used in instruction at Tompkins Cortland Community College in Dryden, New York; it’s not an official statement of the College. Please visit www.tc3.edu/instruct/sbrown/ to report errors or ask to copy it.

For updates and new info, go to http://www.tc3.edu/instruct/sbrown/stat/