Chapter 14 Flashcards
What does a correlational, ecological, or aggregate study do?
Uses population-level data to examine the relationship between exposure rates and disease rates.
Correlational studies that explore an environmental exposure may be called:
Ecological studies
Why are correlational studies also called aggregate studies?
Because they only look at aggregate, or grouped, population-level data, and they do not include any individual level data.
What type of study are correlational studies?
Secondary study
What is the key to success for correlational studies?
Identifying data sources that contain comparable information about the variables of interest.
What is a requirement of correlational studies?
The topic must not have been previously explored using individual level-data
What is the key statistical measure for correlational studies
Correlation
Most exposures and outcomes used in correlational studies are in the form of:
Population-level statistics
What should be done before conducting a statistical analysis of aggregate data?
The data from each population must be entered into a spreadsheet. Each population should have its own row, and each exposure and outcome should be assigned to its own column.
The correlational study analysis will only be valid if:
The data points are comparable.
What might interfere with the definition of an exposure or a disease from one population to another in a correlational study?
1) Multiple sources of data are used.
2) The data was collected over a lengthy period of time.
On a scatter plot used to illustrate correlation what does each point represent?
One population in the study
What is plotted on the x-axis and the y-axis of a scatterplot for a correlational study?
X-axis: Exposure
Y-axis: Outcome/disease
When all points fall neatly along or a very near a sloped line, the correlation is _______. A positive (upward) slope shows that _____. A negative (downward) slope shows that ____.
Very strong; higher levels of exposure are associated with higher rates of disease; higher levels of exposure are associated with lower rates of disease.
When the points are not exactly linear but a line for trends can be drawn through them, the correlation is ____.
Mild or moderate
When the points appear to be randomly placed and no obvious line can be drawn through them, or the best-fit line is horizontal, then the correlation is ____.
Weak or nonexistent
When do we use Pearson correlation coefficient (r)?
For continuous variables and other variables with responses that can be plotted on a number line.
When do we use Spearman rank-order correlation (rho)?
For variables that assign a rank to responses or that have ordered categories.
For both the Pearson and Spearman correlation tests, r=-1 means:
Very strong negative correlation
For both the Pearson and Spearman correlation tests, r=0 means:
No association between the exposure and outcome.
For both the Pearson and Spearman correlation tests, r=+1 means:
Very strong positive correlation
What is the coefficient of determination?
It shows how strong a correlation is without indicating the direction of the association. (0 = no correlation, 1= perfect)
If two populations differ in age what should be done?
Age adjustment
What does direct age adjustment require?
Knowing the exposure or disease rates by age group in each population
When do we use indirect age adjustment?
When comparing populations for which the population age distributions are known, but age-specific rates of exposure and/or disease are not known.
What is the ecological fallacy?
The incorrect attribution of population–level associations to individuals.
True or false:
An individual from a country with a high average BMI will be obese.
False; this is a population-level trend and it is incorrect to assume that every individual follows the trend.
Correlational studies are useful starting point for:
Generating hypotheses about associations, but they are not the final word on risk factors for disease.