Maybe. I very much suspect that the variation among individuals is considerably greater that the differences between groups. In which case, all tests will be non significant. Perhaps the best thing to do, as it always is, is to plot the data. Make a mark for each group on the x-axis and above these, use the y-axis scale to mark the values for the individuals in the group. If you cannot see a SUBSTANTIAL difference in the centers of the clouds, no statistical test will find it for you.
Stan Brown wrote: > In article <[EMAIL PROTECTED]> in > sci.stat.edu, m v wrote: > >>My wife and I both have degrees in math and have taken undergrad >>statistics courses long ago but are having trouble helping our child >>determine the appropriate statistics to use for her middle school >>science project and how to translate the result into english for her >>conclusion. >> >>Her project was to find if short term memory was better for young >>adults than for younger or older age groups. She devised a memory >>test and tested 10 people in each of five different age groups. >> >>What would be an appropriate statistical test for this hypothesis and >>data (considering it is for a middle school project)? > > > Assuming Group 1 is the youngest, she could do a two-sample t test > of Group 1 (n = 10) against the aggregate of the other groups (n = > 40). While she is looking to see whether memory is _better_ (a one- > tailed test), it is probably better to test whether it is > _different_ (a two-tailed test). A one-tailed test says, in effect, > "equal memory or worse memory, it's all one to me". > > On the other hand, she could do 2-sample t tests of one group > against another. Since there are 10 possible pairs of groups > (5*4/2), this could be rather laborious. She might, however, test > Group 1 against each of the other 4 groups. Perhaps she might find a > significant difference between Groups 1 and 5 but not between Groups > 1 and 2. > > Finally, there's a one-shot procedure that tests whether the five > population means are equal or not: this is ANOVA. But that just > tests whether the five population means are equal, not whether they > occur in any sort of order. > -- Bob Wheeler --- http://www.bobwheeler.com/ ECHIP, Inc. --- Randomness comes in bunches. . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
