Applied Research Project One-Page Proposal
What information should my One-page proposal contain?
Describe the data files you will use.
I will be using the data collected in the 2007 National Household Education Survey, and I will be using the Parent and Family Involvement in Education Survey.
o How many observations are there in the data?
There are 10, 681 observations in this data set.
o What is the primary entity? (students, teachers, schools etc.)
The primary entity are students, more specifically the children enrolled in Kindergarten through the 12th grade with the data set describing various things about the child’s family and their involvement in the child’s education.
o Describe the structure of the data:
Observational vs. Experimental? Cross-Section, Panel, or Repeated Cross-Section?
The structure of the data is observational because there was no treatment assignment and participants in the survey simply respond to the questions asked. The ...view middle of the document...
What is(are) your outcome(s) and how will you measure it(them)?
My outcome for both research questions is the amount of media the child has watched.
What variables will you use as controls to test the robustness of the observed association between your question predictor and your outcome?
I will control for the age of the child which influences if the child can read or not, which is the variable AGE2006. I will also control for the grade of the child, which as the grade increases the number a books a child reads is expected to increase, which is the variable GRADE. I will also control for the language the child speaks at home, which is the variable CSPEAK, since a child whose primary language is not English could struggle to read or watch tv, or not have as many books because they do not speak English as well. I could also control for how much the parents of the child read; if a parent reads a lot, it is very likely that the child has al lot of books, and whether the child is overweight or not (HDWEIGHT2) and whether the child has a learning disability (HDLEARN), which could impact how many books the child has. For further analysis, I can control for kinds of TV channels watched such as Disney channel, Discovery channel, Cartoon Network, et. cetera, and if my research questions do not yield enough analysis, I can examine the relationship between the different kinds of television channels a child watches and the amount of books a child has.
Describe the statistical method(s) you will use to answer your research question.
I will be using multiple OLS regression for both questions, controlling for a range of variables when answering both research questions using methods outlined in the statistical interactions unit. I will check to see if all the conditions for OLS regression have been met, transforming variables and fixing homoscedasticity if need be.
Describe the biggest threats to internal validity for the observed relationship between your predictor and your outcome.
We are assuming that each child is completely isolated from the other, whereas in reality, children are not. We are also assuming that all the information we have about the predictors is true, whereas in reality, people have been known to lie during surveys, but this is something that we cannot really avoid.