Data Collection and Observational Studies Introduction to Data

Data Collection and Observational Studies Introduction to Data 1

Data Collection Methods � Anecdotal Evidence � Observational � Controlled Studies (Designed) Experiments

Anecdotal Evidence � Data collected informally or through personal testimony � In February 2010, some media pundits cited one large snow storm as valid evidence against global warming. � Jon Stewart pointed out, “It’s one storm, in one region, of one country. ”

Treatment and Control Group � Treatment Group that receives the treatment (or has the characteristic of interest) � Control Group that does not receive the treatment (or does not have the characteristic of interest)

Observational Study � Observes individuals and measures variables of interest but does not attempt to influence the responses. � Subjects in the study are put into the treatment group or control group either ◦ By their own actions ◦ By the decision of someone else who is not involved in the research study.

Observational Study Examples � Most (if not all) disease, addiction, or “bad outcome” study. ◦ Cannot assign people to the treatment and control group randomly. ◦ May not be ethical to control. � Some interesting associations. ◦ Number of TVs a person owns and life expectancy. ◦ Cause of Global Warming… ◦ Spurious correlations

Confounding (Lurking) Variables Confounding Variable A variable that has not been accounted for but which is causing a difference in the groups being studied. From Previous Example: Higher income would lead to the ability to buy more TVs and most likely better healthcare. Therefore this might lead to longer life expectancy.

Observational Studies � Advantages ◦ It can detect associations between variables where the values of the variable have already been determined. ◦ Can be used to study variables that are impossible or unethical to control � Disadvantages ◦ It cannot isolate causes to determine causations.

Controlled Experiments � Deliberately impose a treatment on individuals and record their responses. Influential factors can be controlled. � Use randomization and controlling practices to ensure that the treatment group and the control group are as similar as possible. � More on Design of experiments in the future….

Important Study Terminology � Time terms ◦ Retrospective – A study that looks into the past ◦ Prospective – A study that deals with current data ◦ Longitudinal- A study that follows the same individuals over a long period of time � Types of studies ◦ Cross sectional study – Surveys that collect data on a population at one point in time (often prospective) ◦ Case-Control- Often a retrospective study for rare conditions ◦ Cohort Study- A Longitudinal Study following a group of similar individuals 10

Study Examples � With the increased use of cell phones, there has been a growing concern over the potential health risks. � Cell phones emit electromagnetic radiation (generally from the antenna). The closer the antenna is to the user’s head, the greater the exposure to radiation. � Multiple studies have been conducted to explore this risk. 11

Study Examples � My friend’s mom’s cousin got the first cell phone that ever came out and got cancer… �A 2001 German study compared 118 patients with a rare form of eye cancer (uveal melanoma) to 475 healthy patients (without the cancer). The patients’ cell phone use was measured using a questionnaire. ◦ On average, the eye cancer patients used the cell phone more often. �A 2006 British study compared 996 patients with brain cancer to 1716 patients without brain cancer. The patients’ cell phone use was again measured with a questionnaire. ◦ The results showed that the cell phone use for the two groups was similar.

Study 4 study was published in The Journal of the American Medical Association (Volkow et. al. , 2011). As part of a randomized study, 47 participants were fitted with a cell phone device on each ear and underwent to PET scans to measure a specific brain activity. � Another � During one scan, both cell phones were turned off and during the other scan, an automated 50 -minute muted call was made to the phone on the right ear. The order of when the call was received (first or second scan) was randomized. � Comparison of the PET scans showed a significant increase in activity in the part of the brain closest to the antenna during the transmission of the automated call. 13