Lab 1: Protocol

In today's lab you will work with your new group to develop a team contract that will serve you throughout this course. You will also begin your first research project by creating a research proposal and running some initial analyses.
​Exercise I. Develop a Group Contract
Exercise II. Create a Research Proposal
​Exercise III. Begin Analysis 

Review pre lab

preview post lab

Lab Objectives: Following today's lab, you should be able to...

Exercise I. Create Your Group Contract

Depending on your past experiences and your own personality, group work may be exciting or dreaded.

Procedure

1. Please review THESE TIPS, individually, then discuss them with your group.
2. Review the description of group roles below.
3. Download and complete the group contract form.
4. Ensure all members have an electronic copy for reference and be ready to upload the contract in the post-lab.  
5. ​Lastly, review the peer evaluation form you will complete at the end of this project. This form will be completed by every group member at the end of each unit. 

Activity Roles for Group Work. Click to enlarge.

Group Contract Template. Click to download.

Exercise II. Construct Your Research Proposal

Your first research Project: The ultimate goal is for you to create a technical report that, in some way, expands our understanding of the diabetes epidemic in the U.S., using our diabetes data set. You must first get your project idea approved, and therefore need to construct a research proposal.

​​In the pre-lab, you learned what a research proposal is and how it is linked to the steps of the scientific method. This initial proposal is very basic and we will build on it in subsequent labs. Remember, you are creating two separate research questions for this project, one that deals with the difference between two groups and one that deals with the relationship between two variable. You will be asked to generate a null and alternative hypothesis for each research questions your pose. ​

• A null hypothesis (H0, or HNULL) represents the default position that there is nothing interesting going on in your data and your findings exist as result of chance and randomness (e.g., there is no difference between groups or there is no relationship between  variables). We attempt to reject this idea and find support for our alternative hypothesis. 
• ​An alternative (or working) hypothesis (HA, H1, H2) states your claim or prediction. It is what you think might be happening. This hypothesis cannot be proven or accepted. If the H0 can be rejected based on our data, we may be able to show support for this idea.  ​

​Remember: A quality hypothesis must be objective, measurable and testable. It must also include a prediction and potential rationale (examples). Think about how your questions may connect with what we already know about the diabetes epidemic and related factors.

​Procedure 

1. Be sure you have downloaded the diabetes datasheet in excel.
2. Download the research proposal template.
3. Review the diabetes data, discuss your ideas and brainstorm which variables you like to use for each question. Each research question should address the prevalence of diabetes in the U.S but does not necessarily need to use data from the states + DC.
4. Use our research library as needed to better understand diabetes and think about how your questions may connect with what we already know about the diabetes epidemic and related factors. 
5. Complete your research proposal. 
6. Ensure all members have an electronic copy for reference and be ready to upload the proposal in the post-lab.  

Diabetes Data. Click to download.

Research proposal template for Lab 1. Click to download.

Click to enter research library.

Exercise III. Begin Analysis

You will be working on analysis over the next two labs. At this point you know the names of each statistic you will use to test your hypotheses. But, depending on your background, you may not know much else about statistics. Why do we need them?

Different is different...why do we need to run a t-test? ​A relationship is obvious...why do we need to run a correlation?

We look for patterns to help us understand the natural world. As we do so, we are fighting our own human tendency to see patterns where none truly exist, and to take what we see in a specific context and try to apply it more broadly.

Let’s say you notice that within your friend circle, those who regularly eat breakfast did much better on their first BIOL 120 exam than those who skipped your morning meals together. You might then assume that somehow, eating breakfast is causing the better grades. But! Consider the following...

• How much of a difference in test averages should there be between breakfast-eaters and breakfast-skippers for that to be true? 90 vs. 60? 80 vs. 70? Reasonable people could disagree.
• How often do you need to see this relationship for that to be true? On exam 1 and 2? On all the BIOL 120 exams? Reasonable people could disagree.  
• Is it really the breakfast? Maybe students who are disciplined enough in their sleeping and eating habits are also more disciplined in their study habits? Reasonable people could disagree.  ​

Statistics solve this problem. Using the principles of probability, they help us parse what we observe from randomness (chance alone) vs. meaning (a real difference, or a real relationship). Statistics tell us how likely we would be to make the same observations we have made, if chance and randomness were the only drivers. If the probability is very low (<5%), we refer to these patterns as significant. ​

Procedure

1. Make sure everyone is confident in concepts explored in the table above.
2. A t-test tutorial is available in the links below. After watching the tutorial, you can go to the calculator and run your test. 
3. Following the test, be sure to copy all the relevant information from the t-test calculator into your "t-test" excel workbook.​
4. There is also a correlation tutorial available below. After watching the tutorial, you can go to the calculator and run your test. 
5. Following the test, be sure to copy all the relevant information from the correlation calculator into your "correlation" excel workbook.
6. Next week, we will pick up with these results and move forward with how to analyze and report your findings. We will also go over the technical report you will create based on your research.
7. Ensure ALL group members have copies of the data and test results and move on to complete the post-lab.​

Comparing important concepts for each type of test. Click to enlarge.

Review the statistics we will use and associated concepts in this table. Discuss the contents with your group and ensure everyone feels confident with material before moving on to the procedure.

T-test Online Calculator.

paired t-test Demo

unpaired t-test Demo

Pearson's Correlation Online Calculator.

correlation demo

Post-lab

Lab 1 BIOL 120 CONNECTIONS
Section 1.6: Doing Biology
Big Picture 1: How to Think Like a Scientist
BioSkillls 2: Reading & Making Graphs
BioSkillls 3: Interpreting Standard Error and Using Statistical Tests
BioSkillls 4: Working with Probabilities 
​