If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

### Course: High school statistics>Unit 5

Lesson 4: Introduction to experimental design

# Principles of experiment design

A footwear company wants to test the effectiveness of its new insoles designed to prevent shin splints resulting from running. They hire a group of physical trainers and a statistician, who recruits $100$ healthy adults between the ages of $18$ and $24$ to participate in a study.
The statistician randomly assigns $50$ of the adults to follow a weekly running schedule with the new insoles and the other $50$ to the same running schedule with the existing insoles the company already sells. After $10$ weeks, the statistician records the number of runners from each group that have developed shin splints.
Problem 1
What is the primary purpose of having a group of $50$ runners use the existing insoles?

problem 2
What is the primary purpose of randomly assigning the runners to use either the new or existing insoles?

problem 3
The researchers considered having one group use the new insoles and the other group use no insoles at all.
Why might this design lead to confounding?

Problem 4
Neither the participants nor the researchers collecting the data knew which group had the new insoles and which group had the existing insoles.
What is one reason for using this type of double-blind design?

## Want to join the conversation?

• for question 3: how would having people not wear insoles at all cause a placebo effect, at least compared to having them wear the normal insoles?
• The fact that the Treatment group has received something that the Control group hasn't might psychologically impact the Treatment group. The Treatment group might become more confident having received extra equipment. Compared to the Control group that didn't get to experience that little burst of excitement we all get from being introduced to something new.
• >Problem 2: What is the primary purpose of randomly assigning the runners to use either the new or existing insoles?

Roughly similar groups also reduce the confounding effect, right? Why should the answer not be B
• Having 2 roughly similar groups doesn't guarantee that you might not have to deal with any confounding variables.
• The purpose of the experiment is to test the effectiveness of the new soles versus their existing soles or no soles at all? I would think using No soles at all for the control group would break the experiment completely, as it would be testing new soles versus no soles, and no wearing no soles in the shoes while running would likely increase chances of getting shin splints. So question 3 is misleading. This is not just a question of placebo effect.
• The purpose of the experiment is indeed to test the effectiveness of the new insoles compared to the existing insoles. The consideration of having one group use no insoles at all raises concerns about the potential for confounding due to factors such as changes in running mechanics or increased risk of injury without any insoles. This design would not be appropriate because it introduces additional variables that could affect the outcome (shin splints), making it difficult to determine whether any observed differences are due to the new insoles or to the absence of insoles altogether. Therefore, the presence of a control group using the existing insoles helps isolate the effect of the new insoles by providing a baseline for comparison.
• In problem 3, why does the placebo effect lead to confounding?
• Confounding is when some other variable impacts the experiment. Basically, the placebo can introduce psychological factors that impact the results of the experiment.

While you may not think this is a big deal, a lot of diseases/medical issues can be caused by psychological factors e.g., hives, insomnia.
• In question 3 option 3 is also reasonably correct as we can say that by not providing insole to one group the study will become sort of observational as we can correlate whether having insole is reducing the shin splints. I agree that option 1 is most appropriate but again why option 3 shall be incorrect that I would like to know..
• From Wikipedia:

an observational study draws inferences from a sample to a population where the independent variable is not under the control of the researcher because of ethical concerns or logistical constraints.

In this scenario we can control the independent random variable whether a person receives the new insole.

The article does great job in explaining when an observational study is required.

https://en.wikipedia.org/wiki/Observational_study