here

advertisement
Page 1 of 3
Worksheet 1 -Data Analysis
This worksheet is designed to help you to:
i) Practice using Excel to organize, analyze and present data.
ii) Understand the three main types of statistical tests that you will be required to use during this
course.
Answer/address all of the following questions. In all cases, we require that your work be typed and that
you make computer printouts of spreadsheets, graphics, and any statistical output.
Question 1
Hennington pond was the only pond on the FIU campus not treated with chemicals or subjected to
storm run-off. In March 2007, a sewage main ruptured and the lake became contaminated with
sewage effluent.
You have been provided with data on the phosphorus levels in Hennington pond for two years:
2004 (before perturbation) and 2007 (after perturbation). For both years, sampling was conducted
by collecting water samples from 20 random sites within the pond.
Based on observations of Hennington pond during 2004 and 2007, we could ask the following
question:
Has the sewage contamination caused changes in phosphorus levels in Hennington pond?
1. Based on the above question, formulate null and alternative hypotheses (Ho and Ha) that can be
statistically tested (2pt).
2. Using data given, produce a summary table (DO NOT INCLUDE RAW DATA) showing the
mean and standard deviation for the phosphorous levels in 2004 and 2007 (4pt).
3. Produce a bar graph that shows the mean phosphorus levels in 2004 and 2007. Remember to
include the error bars (of the standard deviation) on each bar graph (5pt).
4. Which is the most appropriate statistical test that can be used to test the proposed hypothesis and
why? (2pts)
5. What would be an appropriate α-value (2pt)?
6. In Excel, run your statistical analysis, present the table of results and highlight the p-value (2pt).
7. State whether the p-value from your statistical analysis is greater or less than your α-value (1pt).
Will reject or fail to reject your null hypothesis and why? (2pts)
8. Based on the statistical significance of your results, draw conclusions based on your initial
questions (2pt).
9. Briefly discuss whether or not your results differed from what you expected to find (3pt)?
Page 2 of 3
Question 2
On the FIU campus there are three large freshwater ponds: DM pond, OE pond and Hennington
pond. The FIU irrigation committee has decided to increase the drainage of water to one of the
three ponds to reduce flooding on campus. Therefore, they are interested in knowing which pond(s)
could withstand receiving more water without overflowing.
You have been provided with data on the water depth in the three ponds during both the dry and
wet seasons. For each pond, sampling was conducted during each period by measuring the water
depth at 10 random points within each pond.
Based on observations of these three ponds and the need to increase water flow into one of the
ponds, we could ask the following questions:
i) Is there a significant difference in average annual water levels across the three ponds?
ii) Is there a significant difference in average wet season water levels across the three ponds?
1. Based on the above questions, formulate two sets of null and alternate hypotheses (Ho and Ha)
that can be statistically tested (4pts).
2. Using the data given, produce a summary table that shows the mean and standard deviation for
the water depths in each pond both annually and during the wet season (8pts).
3. Produce a single bar graph that shows the mean water depth during the wet season for each pond.
Remember to include the error bars (of the standard deviation) on the graph (5pts).
4. Which is the most appropriate statistical test that can be used to test each of the two sets of
proposed hypotheses and why? (2pts)
5. What would be an appropriate α-value for each test (2pts)?
6. Using Excel, run your two statistical analyses. Present the results table and highlight the p-value
(4pts).
7. For both of your statistical analyses, state whether the p-value from your statistical analysis is
greater or less than your α-value (1pt each; 2 total pts). Will you reject or fail to reject your null
hypotheses and why? (2pts each; 4 total pts)
8. Based on the statistical significance of your results, draw conclusions based on your initial
questions (4pts).
9. Briefly discuss whether or not your results differed from what you expected to find (3pts)?
Page 3 of 3
Question 3
You have been provided with data on the percent cover and average height of Thalassia testudinum
(Turtle grass) along a 50m transect running from the shoreline out to sea at Crandon Park beach.
Sampling was conducted using the line transect sampling technique.
Based on observations at Crandon Park we could put forth the following questions:
i) Does the percent cover of Thalassia testudinum (Turtle grass) change with distance from
shore?
ii) Does the average height of Thalassia testudinum (Turtle grass) change with distance from
shore?
1. Based on the above questions, formulate two sets of null and alternative hypotheses (Ho and Ha)
that can be statistically tested (4pts).
2. Produce two scatter plots which show the change in (i) percentage cover of Thalassia testudinum
and (ii) mean height of Thalassia testudinum along the transect. Remember to include a linear
trend line on each graph (10pts).
3. What is the most appropriate statistical test that can be used to test each of the two sets of
proposed hypotheses and why? (2pts)
4. What would be an appropriate α-value for each test (2pts)?
5. Using Excel, run your two statistical analyses, present the table of results, and highlight the pvalue (4pts).
6. What other value apart from the p-value in your statistical output gives you information about
your analysis (2pts)?
7. For both of your statistical analyses, state whether the p-value from your statistical analysis is
greater or less than your α-value (1pt each; 2 total pts). Will you reject or fail to reject your null
hypotheses and why? (2pts each; 4 total pts)
8. Based on the statistical significance of your results, draw conclusions based on your initial
questions (4pts).
9. Briefly discuss whether or not your results differed from what you expected to find (3pts)?
Please note for all questions:
* Tables must have (i) a proper title and (ii) a heading for each column indicating what data are in
the column and the unit of measure (e.g. cm, grams, %).
* Graphs must have (i) a proper title, (ii) a heading for each axis indicating what data are being
shown and the unit of measure (e.g. cm, grams, %) and (iii) means plotted with standard deviations
shown as error bars.
* For graphs showing relationships between two variables, choose a dependent variable (i.e., species
composition or biomass) that will be shown on the Y-axis and an independent variable (e.g. distance
from the shore, water depth, soil depth, etc.) which will be shown on the X-axis.
Download