Foundations of Inference

Song of the Day

• Love Will Tear Us Apart -Joy Division

Main ideas

• Understand statistical process terminology
• Understand the types of conclusions that can be made given study design and statistical process

Coming Up

• Homework #03 due Thursday 3-04.
• Break this weekend.
  • No lab this Friday 3-05.
  • No class Tuesday 3-09.
  • No TA office hours on Tuesday 3-09 and Wednesday 3-10.

Polls and statistical terminology

Go to the Monmouth University Polling Institute website and select a poll of interest. Briefly read the poll results and methodology section at the end. Try to identify the following:

• Population of interest: U.S. adults

• Parameter of interest: The true proportion of U.S. adults who will get the vaccine as soon as allowed (\(p\)).

• Sample: Random sample of landline telephones using random digit dialing.

• Sample size: 809.

• Sample statistic: The sample proportion.

• Sample statistic’s value: \(0.50\).

Link to poll

Prostate Cancer

A 2001 study published by researchers at the University of Illinois at Urbana-Champaign examined 753 men with prostate cancer and compared them to 703 men of the same age without cancer. They found that the prostate cancer risk increased with the lifetime number of female sexual partners. Dr. Karin Rosenblatt hypothesized an infectious agent as a potential cause of prostate cancer.

1. Is this an observational study or a designed experiment?

This is an observational study (case-control study).

2. Can you conclude that a higher number of sexual partners causes prostate cancer?

No.

Identify the following as causal link, confounder, or neither.

3. Testosterone: Higher testosterone is associated with higher numbers of sexual partners and also increased cancer risk.

Confounder.

4. Genetics: Some individuals are predisposed to prostate cancer.

Neither.

5. Infectious Agent: A sexually transmitted infection causes prostate cancer and a higher number of sexual partners increases your risk of catching the infection.

Causal link.

An Apple a Day

Researchers followed 1,456 Australian women over the age of 70 for 15 years and assessed their apple intake with a periodic questionnaire. The results were idiomatic; individuals who consumed at least an apple a day had a significantly lower cancer mortality relative to those who did not consume an apple a day.

List three confounding factors that could be at play here.

1. Healthy lifestyle choices.

2. Wealth.

3. Location.

Given that the above study was observational, what type of conclusion can be made?

We can conclude that there is an association between apple consumption and lower cancer mortality.

References

1. https://www.understandinghealthresearch.org/useful-information/confounders-17.

2. Rosenblatt K., et al, Sexual Factors and the Risk of Prostate Cancer. American Journal of Epidemiology, 153 (12), 2001, 1152–1158.

3. Hodgson et al (2016). Apple intake is inversely associated with all-cause and disease-specific mortality in elderly women. British Journal of Nutrition, 115 (5).