What is the Nurses' Health Study and how can we tell that it is correlational rather than causal research?

Question

What will be an ideal response

Answer 1

The Nurses' Health Study (NHS) is perhaps the most comprehensive descriptive investigation of health-related behavior ever conducted. Since its inception in 1976, over 238,000 nurses have provided information. The NHS reports findings based on statistical analyses of millions of data points. Some remarkable, unexpected, and important correlations were discovered. Measured expressions like "investigations †¦ suggested †¦", "†¦ is associated with reduced risk †¦", and "strong correlations †¦ support †¦" characterize the annual reports. These are indicators of the carefully scientific investigation of phenomena where the scientists are looking for possible correlations. If the reports were narrative stories describing what happened to one or two individuals, we would probably classify the NHS as a human interest project focusing on chance or random events. If the reports talked more factors, which explain the biological, genetic, or environmental interactions that explain exactly how illnesses are contracted or transmitted, or if the reports described experimental studies aimed at preventing or curing illnesses, then we would regard the study as focused on causal explanations. But, we can see in the examples a focus on correlations. In the "2009 Early Life Factors and Risk of Breast Cancer" example the researchers say they have "begun to examine the relation" of high school diet and activity patterns to subsequent risk of breast cancer. In the "2013 Adolescent Alcohol Intake and Benign Brest Disease" example the scientists report that higher levels of alcohol consumption between ages 18—22 "was associated with" increased risk. The NHS scientists are not running journalists telling random stories, nor are they experimental investigators trying to figure out how to treat or control disease. They are researchers gathering data to analyze for the purpose of finding correlational patterns of relationships. Their conclusions are warranted because the statistical analyses provide sufficient confidence to assert that the relationships on which they report are highly unlikely to have occurred by random chance.