1. For each of the following scenarios, state the following:

1. Sampling plan (Based on exposure, Based on outcome, random sample)
2. Main outcome of interest and variable type (continuous, categorical-binary, categorical-nominal, categorical-ordinal)
3. Main predictor of interest and variable type (continuous, categorical-binary, categorical-nominal, categorical-ordinal)
4. Whether linear or logistic regression is the appropriate analysis technique to answer the research question and why.

1. A study was performed relating baldness pattern to history of MI (heart attack) among men in the ARIC study based on the assumption that one of the genes responsible for balding may also cause an increase in cardiac issues. Baldness pattern (5 categories) and history of MI were determined at the same examination during the period 1996-1998. Is there a significant association between MI and balding pattern after controlling for age?

2. Based on all doctor visits during the first year of life, children with at least one older sibling were classified as having one or more episodes vs. no episodes of otitis media. They recruited an equal number of children with and without otitis media. Several risk factors were studied as possible predictors of otitis media. One such risk factor was a sibling history of ear infection, measured at birth. Assess whether a sibling history of ear infection is associated with otitis media incidence in the first year of life.

3. An important area of investigation is the development of strategies for altering cardiovascular risk factors in children. LDL cholesterol has consistently been shown to be related to cardiovascular disease and is often used as a surrogate marker for cardiovascular risk. A study was conducted to identify whether ponderal index, a measure of obesity, is related to LDL cholesterol in children. At the visit, ponderal index was calculated and blood was drawn to determine LDL cholesterol levels. Is there an association between ponderal index and LDL cholesterol?