2.4 Conducting Experiments

We will briefly discuss how to design and evaluate an experimental study. We will go much more in depth during our unit on inferential statistics.

Designing an Experiment (slides)

What are some important things to keep in mind as we design statistical experiments? What are some issues that may arise?

⊕ null vs alternative hypothesis
⊕ independent vs dependent variables
⊕ treatment, control group
⊕ blind, double-blind

Practice

Choose one of the following situations:

• A carpenter is looking to expand their business to include electric standing desks. They’ve studied existing models and noticed that motors play a big part in the perceived quality of the desks. The carpenter wants to make sure that their standing desks use the best motors available.
• A chef is thinking about what to prepare for the next seasonal menu. There has been a trend in sourcing ingredients from local farms. People have been saying that it just makes the food taste better. The chef is inclined to agree.
• An audio technician wants to replace the speakers in the studio, since the speakers are old and producing a lot of static. The studio’s budget includes renting and testing different speakers, so the technician hopes to come to a clear conclusion before making any final decision.

Then, for your chosen situation, write a response to the following:

1. Identify a statistical question you could ask in that situation.
2. Identify the null and alternative hypotheses.
3. Identify the population and sampling method you would use.
4. Identify the variables and treatments that would provide information about your question. Explain your choice of treatments.
5. Identify the treatment groups, and how each subject would be assigned to a group.
6. Briefly describe how the experiment would be conducted.
7. What issues may arise, and how would you address them?