Analyzing the grass shrimp data from Wednesday’s lab

Step one: make 3 graphs, each with “Number of cysts” on the x-axis

Step two: do pairwise t-tests or one-factor ANOVA to look for differences.

Making the graphs (step one)You have eighteen groups of data:

BEHAVIOR WATER #CYSTS% time moving Control vs. Killifish None, Few, Lots# rises Control vs. Killifish None, Few, Lots# tailflips Control vs. Killifish None, Few, Lots

Gier calculated the averages of each, and is pretty sure the following table of averages is correct:

# of cysts % moving # of rises # of tailflips

None 65.48 2.68 .41 to 5 70.58 3.37 .68>5 81.74 4.68 .63

# of cysts % moving # of rises # of tailflips

None 62.27 4.14 .551 to 5 68.12 3.05 .63>5 80.28 5.61 .39

At left: CONTROL

At right: KILLIFISH

On a blank worksheet, create mini-spreadsheets of the averaged data, analyzing just one behavior at a time.

Example: the “% time moving” data looks like this:

Highlight these cells and do a bar graph (“column chart”)

#cysts

%time moving, control

% time moving, killifish

none 65.48 62.271 to 5 70.58 68.12>5 81.74 80.28

none 1 to 5 >560

65

70

75

80

85

%time moving, control% time moving, killifishNext: two more graphs like

this. Then: decide which pairwise t-tests or single-factor ANOVAs to perform.

If you want to try single-factor ANOVA, here is how to organize the data:

% time moving, no cysts

% time moving, 1-5 cysts

% time moving, >5 cysts

# # ## # ## # ##... #... #...

Each # is the data from an individual shrimp. In your data analysis window, you can do a single-factor ANOVA, which is roughly the same as doing all possible t-tests.