Biometry
Stat 412

Goals: 

1) To conduct one-way analysis of variance
2) To check data for assumptions of normality and homogeneity of variance
3) To conduct multiple range tests

A. One-way ANOVA: Otter Data

1. Retrieve the data file otters.MPJ from the course website. The data represent skull widths of male and female sea otters taken from the Aleutian Islands in Alaska and off of the California coast. The goal of this study was to determine if otters in these two areas differed enough to be called separate subspecies. This question was important because the California population is small and endangered while the Alaska population is large and not endangered. If the characteristics of the two populations are similar than the California population is not unique. The sexes were kept separate due to high sexual dimorphism in this species.

2. Analyze the four samples with a one-way analysis of variance. First, check the data for the assumption of normal distribution. Transform the data if necessary.

3. Second, stack your data stack in a single column and add group codes at the same time. Use the STACK command to move all the data into a single column and add group subscripts:

STACK C1 C2 C3 C4 C5;
Subscripts C6;
UseNames.

Notice the subscripts subcommand; this creates labels for each group using the column names (data in C1 is AMCH_M, data in C2 is AMCH_F, etc.) and places them in a given column.

4. Using the stacked data check the data for the assumption of equal group variances. Transform the data if necessary

5. Now that the data are stacked with groups codes complete the one-way ANOVA with the Tukey’s multiple range test using the command:

ONEWAY C5 C6;
Tukey 5.

6. Using the results from the multiple range test what can you conclude about the differences among Alaskan vs. California sea otters? Males vs. females? Have the assumptions of ANOVA been met? What do these results mean biologically? How would you resolve the problem discussed in step 1?

 

This assignment is due next MONDAY in class, 10/29