Overview of What You Do
Choose a question (or set of questions) that interests you and for which you have collected, or can locate, an appropriate data set. You are strongly encouraged to use your own data, if possible, from your current thesis or dissertation research. Decide what the research question of interest is and use that to decide upon an appropriate statistical analysis and run the analysis using SAS. Compile your results and discuss their interpretation in a concise and complete written summary.
Back to top
Overview of What is Expected
We are looking for an independent analysis of a question or set of questions of research interest to you, the student. This is your opportunity to tailor the course materials to your own individual research needs and interest. In the labs, we provide the questions and the data, NOW it's YOUR turn! We are more interested in you taking on a single question and looking at it very thoroughly and correctly than doing many things in a superficial or wrong way. Remember quality rather than quantity!
There are two objectives for the final project assignment.
- Demonstrate an ability to use the computer, i.e., mastery of the basics of the SAS statistical software package.
- Show a solid understanding of the basic principles of study design and statistical analysis and their application to real world, natural resources research problems.
We will be looking for the integration of both of these objectives in your final project. Implicit in these objectives is that you will be able to write effectively and clearly to explain what you did, why you did it, and what you learned from it.
"Correctness of approach" counts. We want to see evidence that you learned how to choose an appropriate statistical approach. It's up to you to choose and justify your choice of method. We would like to know why you did what you did as much as why you chose NOT to do other things.
Back to top
Format Requirements
- Double space your final report
- Use 12 pt font or larger, nothing smaller.
- No longer than 12 typewritten pages including figures.
Note that exceptional papers have been written in less than 5 pages - you do not need to be lengthy to be excellent. You should have enough length so that we can understand what you did but a wordy report detracts from our understanding. Remember that effective and accurate written communication is one objective of the class.
Back to top
Content
The final project report should include the following:
- A statement of your objectives, i.e., the research question(s) you set out to answer and any scientific hypotheses that you are testing. Note that we don't mean "statistical hypothesis" here - we want you to describe the biological hypothesis that interests you. You should include a description of what you expect to find and a quantitative description of what constitutes a biologically or contextually meaningful response. You should also include a statement describing your intended scope of inference.
- Details (documentation) of data collection, i.e., what, where, when, how many, etc. Include, where appropriate, details such as designation of experimental units, sampling units, blocks, replicates, randomization schemes, etc. This description, particularly how the samples were selected, relates directly to your scope of inference so be sure that this description matches up with your statements of the scope of inference in the beginning of your report and later on when you are drawing conclusions.
- A description of the statistical strategy and/or analysis you use. In some cases this could be a fairly standard statement, e.g., analysis of variance based on a 2x3 factorial treatment structure in a randomized complete block model. In other cases your strategy may include a number of steps that are not so neatly stated. In any case, some accurate description of your strategy should be included so the reader could repeat what you did.
- Whether you are using your own data or an inherited data set, you should demonstrate the limitations of the existing data as well as make recommendations for improving the design in the future (i.e., "things I would do differently next time").
- Presentation of your results in suitable form along with your discussion of their interpretation. Include graphs where helpful. Distill your computer output but be sure to include relevant statistics but include only what you need to explain what you did and why you did it. For example, don't forget to include sample sizes, or degrees of freedom on test statistics. Including extraneous output usually confuses the reader (or the grader!) and detracts from the results you are trying to present.
- Scientific conclusions based on the results of your analysis. It should be clear what analysis results you used to make your conclusions and how much statistical evidence there is in your data to support the conclusion. If you use "hedging" statements, do so consciously and accurately. Be sure and relate the results to your intended scope of inference.
Back to top
Evaluation
Your final project will be evaluated according to the following criteria:
- Clarity of presentation - Can the reader understand what you did and why? Can the reader understand how the results from your analysis are related to the conclusion?
- An a priori description of what constitutes a biologically or contextually meaningful response.
- Correct use of technique(s)
- Correctly reporting results
- Assumptions checked appropriately B residuals, etc.
- Evidence of originality of thinking
- "Statistical" versus "Ecological" (or other relevant scientific domain) correctness of interpretation of conclusions
- Discussion of limitations and ways to improve next time
- Thoroughness in terms of method(s) chosen and overall effort
Keep in mind that the emphasis will be on the synthesis and integration of statistical methodology learned in this class as well as from classes you have taken in the past. It should be clear from your write up not only what you did but why you did what you did. Show a working knowledge of the principles and a statistically sound experimental approach. As we near the end of the quarter, we will have examples of past final projects made available to you.
You are encouraged to make appointments with Lisa and/or Manuela throughout the term if you would like to discuss statistical issues related to your final project. We are able to provide statistical consulting during the term for your project. Due to our other responsibilities, we are, unfortunately, not able to do so after the term is over.
Back to top