Issues : Data Sets


How to use this data set in a class:

      This is a good data set to use in a discussion of feeding "strategies," e.g., trade-offs for predators vs. herbivores. The data concern energy balances in animals that have different feedings behaviors (diurnal/nocturnal) and different diets (e.g., carnivore/detritivore) - and therefore with different feeding-behavior strategies and life history evolution. It is also useful for instruction about the use of Excel to make figures because the data set is not very large.

      In regard to sampling and statistics, these data can also lead to discussion about means, standard deviation and standard error, and statistical differences. Despite the wide range in the data, piscivores were statistically different than the rest because of the large sample size. As a specific Excel skill, students could also learn how to sort with the unsorted data.

There are four Excel files:

  1. Complete Sorted Data Set (fishdata1.xls): lists Order, Family, Species, Collection location, Noctiural/Diurnal feeding behavior, trophic category (insectivore, omnivore, algal/detritivore, piscivore), Number of individuals analyzed, and Percent with empty stomachs.

  2. Simplified Students' Data Set (fishdata2.xls); lists Nocturnal/Diurnal feeding behavior, Trophic category, Number of individuals analyzed, and percent with empty stomachs.

  3. Very Simplified Students' Data Set (fishdata3.xls): lists Trophic category and Percent with empty stomachs.

  4. Faculty Data Set Version (with figure; fishdatafaculty.xls); lists Trophic category and Percent with empty stomachs; includes table with mean percent for each of the 4 trophic categories, standard deviation and standard error plus figure showing means and standard deviations.

      If you want your students to learn how to sort data and calculate standard deviations and standard errors, give them the complete sorted data set (no. 1 above). Otherwise give them one of the sorted data sets (no. 2 or 3 above) or rework the file (e.g. calculate standard error) as you wish.

To introduce your students to standard deviation and/or standard error, the Resources section below lists websites that explain these statistics. You can use them to make a hand-out for your students or you can direct them to a site that you like. To save time in class or lab, have students learn to use Excel to calculate mean, stdev. and s.e. as homework.

Making and Using Figure:

      The Resource section below includes URLs for a variety of Excel tutorials. Because it took me a long time to figure out how to make standard deviations on a Mac, I include these directions here.

The steps to the completed figure (for a Mac) are:

1. Sort the data by trophic category

2. Calculate mean percent for each category and create new 4 row table on the spreadsheet (see C and D6-D9).

3. Use STDEV command to calculate standard error for each mean.

4. Calculate standard error as standard deviation/square root of N.

5. Select first trophic category and respective mean (e.g. 2 cells), then select Chart Wizard and Column subtype. On the new figure place the mouse at the top of the column to get the "Format Data Series" menu; select Y Error Bars. Then select "Custom" and type in the standard error values in the + and - boxes. Repeat for each trophic category one by one.

6. Label axes.

      You can have students discuss their figures in a variety of ways. For instance students can work together in small groups (2-3) with Excel to create the figure and then discuss its implications together. In a small-medium size class each group can present their ideas and questions to the class; for larger classes or to take less time you can call on a group randomly to start off the discussion.

Student figures.

      The Student Instructions show students how to make the simplest figure in which each individual data point is shown. For this they do not need to calculate mean or standard deviation. If you have little time, this figure will allow discussion of the main finding and also variability in the data and how to address this in general.

      To use the data for a class on mean, standard deviation, and standard error give students the table with these data. Alternatively they can do the sorting and the calculations if you give them the equations.

Questions for discussion

1. Arrington et al. say in their conclusion "…our results reveal a potential influence of feeding strategy and energy balance on life history evolution.". Explain.

2. What further questions do you have after examining this data set and what follow-up research would you suggest to address these questions?.

3. What are the advantages/disadvantages for fish of a diet of algae and detritus vs. carnivorous diet? .

4. Arrington et al. looked at over 35,000 individual fish in this study. How does this large sample size influence interpretation of the data and also statistical analysis? How might the conclusions be different if the sample size was much smaller?.


      To assess your students' learning you need to figure out what you most want them to learn (e.g. how to use Excel, the definition and utility of standard deviation, trade-offs for animals from different trophic levels.). A rich discussion on this topic appears elsewhere in TIEE in a paper on Assessment and Evaluation.

      Assessable tasks include written answers to one or more of the questions above or creation of new figures with Excel.

Resources to Help Students Analyze these Data Sets