A not-for-profit organization entitled "People for a Return to Family Values" (PERFAVAL) have produced a policy report wherein they claim that families who eat meals together on a regular basis are more likely to be successful in "all realms of life." Leaving aside the question of whether eating together helps people bond emotionally, you know of many people who do not share meals regularly with their families, and yet have high incomes. During the upcoming teaching break, you decide to do some research for fun. Unfortunately, you fail to find any Australian data that contain both information on economic success and information on eating together, but you finally find (on http://www.bls.census.gov/spd/) the 2002 iteration of an American survey called the Survey of Program Dynamics (SPD) that appears to have both sorts of information. You note that the survey design specifically over-samples low-income families. Your research question is as follows: Using this sample, is there empirical evidence to support the assertion that people who take meals together with their family have higher wages than those who do not?

avgwage - Average hourly wage of a person (across all jobs held, and weighted by hours worked in each job).

bfast - Captures how often the person's family eats breakfast together, taking a value of 1 if never; 2 if 1 to 2 times per week; 3 if 3 to 6 times per week; and 4 if every day.

dinner - Captures how often the person's family eats dinner (that is, the last substantial meal of the day, sometimes known in Australia as "tea") together, taking a value of 1 if never; 2 if 1 to 2 times per week; 3 if 3 to 6 times per week; and 4 if every day.

age - Age of the person, in years.

lang - Dummy indicator for whether the person speaks a language other than English, taking a value of 1 if yes, and 0 if no.

female - Dummy indicator for whether the person is female, taking a value of 1 if yes, and 0 if no.

numper - Number of persons resident in the person's household.
married - Dummy indicator for marital status, taking a value of 1 if married, and 0 otherwise (i.e., divorced, widowed, separated, or never married).

healthy - Dummy indicator for whether the person reports him or herself to be of "good" or "very good" health, taking a value of 1 if yes, and 0 otherwise.

bfast1-bfast4 - Dummy indicators for each value of bfast.

dinner1-dinner4 - Dummy indicators for each value of dinner.

Goals

Development of understanding about how a research question is converted into an empirical action plan.

Development of quantitative analytical skills.