17.37 Year-to-year comparison. Rather than comparing the two groups in each year, we could compare the behavior of each group in the first and second years. The study report says: "Our main prediction was that females receiving addi tional food in the nestling period should riot change laying date the next year, whereas controls, which (in our area) breed too late in their first year, were expected to advance their laying date in the second year."

Comparing days behind the caterpillar peak in Years 1 and 2 gave t = 0.63 for the control group and t = —2.63 for the sup plemented group. Are these paired or two-sample t statistics?

What are the degrees of freedom for each t ? Show that these t-values do not agtee with the prediction.

The remaining exercises ask you to answer questions from data with out having the details outlined for you. The exercise statements give you the State step of the four-step process. Follow the Plan, Solve, and Conclude steps as illustrated in Examples 17.3 and 17.4 for tests and Example 17.5 for confidence intervals. Remember that ex amining the data and discussing the conditions for inference are part of the Solve step.